Manuscript Details

Manuscript number PSS_2020_4_R1

Title Ice Giant Systems: The Scientific Potential of Orbital Missions to Uranus and Neptune

Article type Review Article

Abstract Uranus and Neptune, and their diverse satellite and ring systems, represent the least explored environments of our Solar System, and yet may provide the archetype for the most common outcome of planetary formation throughout our galaxy. Ice Giants will be the last remaining class of Solar System planet to have a dedicated orbital explorer, and international efforts are under way to realise such an ambitious mission in the coming decades. In 2019, the European Space Agency released a call for scientific themes for its strategic science planning process for the 2030s and 2040s, known as Voyage 2050. We used this opportunity to review our present-day knowledge of the Uranus and Neptune systems, producing a revised and updated set of scientific questions and motivations for their exploration. This review article describes how such a mission could explore their origins, ice-rich interiors, dynamic atmospheres, unique magnetospheres, and myriad icy satellites, to address questions at the heart of modern planetary science. These two worlds are superb examples of how planets with shared origins can exhibit remarkably different evolutionary paths: Neptune as the archetype for Ice Giants, whereas Uranus may be atypical. Exploring Uranus’ natural satellites and Neptune’s captured moon Triton could reveal how Ocean Worlds form and remain active, redefining the extent of the habitable zone in our Solar System. For these reasons and more, we advocate that an Ice Giant System explorer should become a strategic cornerstone mission within ESA’s Voyage 2050 programme, in partnership with international collaborators, and targeting launch opportunities in the early 2030s.

Keywords Giant Planets; Ice Giants; Robotic Missions; Orbiters; Probes

Corresponding Author Leigh Fletcher

Corresponding Author's University of Leicester Institution

Order of Authors Leigh Fletcher, Ravit Helled, Elias Roussos, Geraint Jones, sebastien charnoz, Nicolas Andre, David Andrews, Michele Bannister, Emma Bunce, Thibault Cavalié, Francesca Ferri, Jonathan Fortney, Davide Grassi, Léa Griton, Paul Hartogh, Ricardo Hueso, Yohai Kaspi, Laurent Lamy, Adam Masters, Henrik Melin, Julianne Moses, Olivier Mousis, Nadine Nettelmann, Christina Plainaki, Juergen Schmidt, Amy Simon, Gabriel Tobie, Paolo Tortora, Federico Tosi, Diego Turrini

Suggested reviewers Ian Cohen, Sushil Atreya, Athena Coustenis, Olivier Grasset

Submission Files Included in this PDF

File Name [File Type] response2reviews.pdf [Response to Reviewers]

2020fletcher_Voyage2050_IceGiantSystems_trackchanges.docx [Revised Manuscript with Changes Marked]

2020fletcher_Voyage2050_IceGiantSystems_final.docx [Manuscript File] fig1_montage.jpg [Figure] fig7_exoplanets.jpg [Figure] declaration-of-competing-interests.pdf [Conflict of Interest] Submission Files Not Included in this PDF

File Name [File Type] fig2_interior.png [Figure] fig3_atmos.png [Figure] fig4_mag.png [Figure] fig5_moons.png [Figure] fig6_rings.png [Figure] fig8_launch.png [Figure] fig9_questions.png [Figure]

To view all the submission files, including those not included in the PDF, click on the manuscript title on your EVISE Homepage, then click 'Download zip file'.

Research Data Related to this Submission

There are no linked research data sets for this submission. The following reason is given: No data was used for the research described in the article Response to Reviews

Ref: PSS_2020_4 Title: Ice Giant Systems: The Scientific Potential of Missions to Uranus and Neptune Journal: Planetary and Space Science

I have received comments from reviewers on your manuscript. Your paper should become acceptable for publication pending suitable minor revision and modification of the article in light of the appended reviewer comments. When resubmitting your manuscript, please carefully consider all issues mentioned in the reviewers' comments, outline every change made point by point, and provide suitable rebuttals for any comments not addressed. Thank you for the reviews of our manuscript – we have addressed each one below in bold, italic text, and appreciate the opportunity to publish with PSS.

Reviewer One:

This is an excellent, well organized, well written paper. It is timely and very important. I have few suggestions on content, but I do have a number of wording and grammatical suggestions, with several questions and comments that may illustrate possible points of confusion with the current wording. Many thanks for your thorough review – we have attempted to build all of your suggestions into our revised manuscript.

I think my primary and over-riding is that the title refers to the scientific potential of missions to the ice giants, and it really should indicate the scientific potential of remote sensing missions to the ice giants. The manuscript does point out that this paper really is primarily limited to remote sensing since other papers have addressed the science value of probes. However, probe missions and remote sensing missions are not orthogonal, and are in fact so dependent on each other in a synergistic, complementary way that it is something of a dis-service to try to treat them entirely independently. Whilst we quite agree that probes are an essential element of any future mission to these destinations, we felt that the probe science had been fully covered by Mousis et al. Furthermore, it is not fair to characterise this article as purely remote sensing, as there is a substantial quantity of in situ sensing of the magnetosphere (and the heliosphere during the cruise). Nevertheless, we have renamed this as “…Potential of ORBITAL Missions…” to make the distinction clear.

One of the keywords listed is “Probes”, and yet the paper focuses almost entirely on remote sensing studies of the ice giants with very little discussion of probes. We assumed the keyword to refer to “spaceprobes” rather than the more specific “atmospheric entry probes”, and thus have chosen to leave this keyword in place.

Page 6, line 142: "We focus on the science achievable from orbit..." (Mousis, 2018). However, the title doesn't indicate "The Scientific Potential of Remote Sensing Missions....". I think there either needs to be some discussion of probes as a separate but essential element of an Ice Giant mission, or the title should be changed to indicate this covers one element of an Ice Giant mission. The authors of this white paper consider the orbital component as the key need-to-have element of any future mission. We prefer not to emphasise “remote sensing” in the title, for the reasons outlined above. Nevertheless, we have added additional mentions of atmospheric probe science to the appropriate points in the manuscript (as you suggested below), referring back to Mousis et al. (2018) and to a more recent article by Simon et al. (2020).

Line 306: “A combination of global multi-wavelength remote sensing from an orbiter and in situ measurements from an entry probe would provide a transformative understanding…” Again, I think this paper would significantly benefit from at least a brief discussion of the synergistic value of a probe + orbiter mission, and that either by itself would significantly compromise a future ice giant mission. We strongly support the addition of a probe to a future Ice Giant orbital mission, as we have emphasised in the article. However, we do not support a probe *by itself* as suggested in this comment, so prefer not to shift the emphasis in this way.

Other comments: Line 49: Add mention of Europa Clipper? Added.

Line 53: Suggest changing “world” to “system” Changed

Line 85: “planet’s” “planets’ “ Changed,

Line 159: I may be unfamiliar with the nomenclature here, but I think of low-mass as terrestrial -type, and Uranus and Neptune as intermediate-mass. We have changed it to “intermediate-mass exoplanet ”.

Line 199: “heat flux” “internal heat flux” (this may be obvious, but for me this reads slightly better if “internal” is added. OK.

Page 9, Figure 2: “Sketches of the possible internal structures of the ice giants. It is unclear whether the planets are differentiated and whether the transition between the different layers are distinct or gradual:”. Since “are distinct or gradual” is plural, I think that “transition” should be “transitions”. OK.

Line 206: It is not clear how orbital migration would lead to Uranus's large obliquity. I glanced at Boue's and Laskar's paper, but this is so non-intuitive that I wonder if it deserves a few more words, something along the lines of "Boue and Laskar (2010) showed that under special circumstances, the high obliquity could arise from orbital migration and therefore not require a collision to explain….” We have added this suggestion.

Line 247: “Figure 2 presents sketches of four possible internal structures of the ice giants where the transitions between layers is distinct (via phase/thermal boundary) and/or gradual.” Missing word “is”. We changed to “are.”

Line 272: “Finally, the different intrinsic heat fluxes of Uranus and Neptune, together with their similar mass and radius values suggest that the planets formed in a similar way, but then followed different evolutionary histories.” There seems to be an assumption built in here that similar size and mass = similar formation. I would think that similar J2 and J4, similar magnetic field structure, etc. would be a better indication of similar formation. It seems like entirely different formation mechanisms could lead to similar sizes and masses. Additionally, why would “different intrinsic heat fluxes of Uranus and Neptune” suggest that the planets formed in a similar way?

We thank the referee for these comments. We have modified the text to:

”Finally, the different intrinsic heat fluxes of Uranus and Neptune imply that they followed different evolutionary histories. This could be a result of a different growth history or a result of giant impacts at their early evolution (e.g., Reinhardt et al., 2020). Moreover, thermal evolution models that rely on Voyager’s measurements of the albedo, brightness temperatures, and atmospheric pressure-temperature profiles that are used to model the evolution of the two planets cannot explain both planets with the same set of assumptions.”

Reinhardt, C., Chau, A. Stadel, J. & Helled, R. 2020. Bifurcation in the history of Uranus and Neptune: the role of giant impacts. Monthly Notices of the Royal Astronomical Society, Volume 492, Issue 4, p.5336-5353

Line 275: “…temperatures, and atmospheric pressure-temperature profiles inferred from the occultation data cannot…” Radio occultation measurements only provide atmospheric structure to one or very few bars. Is there enough difference between Uranus and Neptune to begin to draw conclusions on the formation scenarios? Please see above - we reorganised to emphasise the evolutionary histories, for which the atmospheric temperatures are a constraint.

Line 279: “…and precise observational constraints on the planets' gravity field, rotation rate, magnetic field, atmospheric composition, and atmospheric thermal structure.” This is one place where it is important to note that atmospheric composition and thermal structure requires a probe, and remote sensing can only provide a very limited measurement. We added the mention of the probe here, although its limited sampling (one single location for temperature and composition) means that we find that this complements the orbiter.

Line 288: “Universe” should be lower case: “universe” Changed.

Line 299: Another good reference is Banfield: Ban_eld, D., P. Gierasch, and R. Dissly 2005. \Planetary Descent Probes: Polarization Nephelometer and Hydrogen Ortho/Para Instruments," IEEE Aerospace Confererence paper #1294, p691-694. We are aware of this article, and it is indeed excellent, but deals primarily with the question of instrumentation - we have chosen to retain the original reference to Smith and Gierasch.

Line 306: “A combination of global multi-wavelength remote sensing from an orbiter and in situ measurements from an entry probe would provide a transformative understanding of these unique atmospheres…” Another chance for me to again mention that probes and remote sensing cannot be treated separately. They are very complementary, and although a probe mission would provide valuable science alone, it also provides the opportunity for ground truth that will significantly benefit remote sensing science. We fully agree that probe science complements and extends that which can be achieved from orbit, via ground-truth and accessing species that cannot be reached by remote sensing. However, we can and do treat them separately (as we seek to do in this article), as we make the case for the significant potential of an orbital mission with remote sensing.

Figure 3, page 14: This is a nice figure, but certainly doesn't say much about the composition or temperature profile that would be the primary measurements goals of a probe. Other than possibly clouds and hazes, this figure doesn't seem to suggest a probe is needed. We do not understand the reviewer’s point here, as the figure is intended to show the range of science that can be achieved from an orbiter mission, conducting remote sensing of the atmosphere and its connection to the interior via magnetic field and gravity mapping. This figure was not intended to make the case for in situ atmospheric probes (although probes can certainly complement the orbital science shown in this figure). We also note that temperature/composition is being determined as a function of latitude, longitude and altitude, via remote sensing in the infrared and microwave, stellar and solar occultations in the UV and near-IR, and radio occultations.

Line 314: “…upper tropospheres, tracking of discrete cloud features has revealed that both planets exhibit broad retrograde jets…” Suggest adding word “planets”, otherwise “both” could almost be read to apply to “discrete cloud features”. Done.

Line 371: “The strength of vertical mixing may vary with location, and disequilibrium tracers such as CO (abundant in the tropospheres of Uranus and Neptune, Cavalie et al., 2017), para-H2 (Conrath et al., 1998) and yet-to-be-detected phosphine (Moreno et al., 2009), arsine, germane, silane, or even tropospheric hydrocarbons (ethane, acetylene), all of which could be used to determine where mixing is most active.” This is a long sentence and I think that read carefully, the wording is a bit awkward: "The strength of vertical mixing may vary with location, and disequilibrium tracers such as ..... all of which could be used to determine where mixing is most active.” We have split this sentence up into smaller chunks.

Line 413: “Summary: Investigations of dynamics, chemistry, cloud formation, atmospheric circulation, and energy transport on Uranus and Neptune would sample a sizeable gap in our understanding of planetary atmospheres, in an underexplored regime of weak seasonal sunlight, low temperatures, and extremes of internal energy and vertical mixing.” Several of these topics not only could benefit from probes, but in one or two cases actually require probes. Just another reminder that the stated intent of this paper is to focus on remote sensing. In the list we provide in this sentence - dynamics, chemistry, cloud formation, atmospheric circulation, and energy transport - we have emphasised the significant potential for orbital remote sensing to address questions on this topic. Probes would certainly contribute useful ground-truth, but none of the themes in this list require a probe exclusively.

Line 423: “The solar wind that embeds these two magnetospheres attains Mach numbers…” I'm not familiar with "embeds" being used in this way. Does this mean the same thing as "impinges upon"? Yes - and we’ve changed it for clarity.

Line 445: “and whether these ever approach steady-state.” Was there evidence that they were not in steady state? The question of whether a “steady state” is ever reached primarily stems from the large dipole tilts at the ice giants and their ~16 and ~17-hour rotation periods, which mean that these systems are expected to be particularly dynamic. Evidence for this deviation away from steady state is found throughout the cited studies that report Voyager observations, showing ongoing magnetospheric reconfiguration during the flybys. We have modified this sentence accordingly.

Line 448: “…believed to be a major contributor…” Contributor to what? Previous statement was that magnetosphere was devoid of cold plasma populations, so not sure what hydrogen corona is contributing to. We thank the reviewer for picking up this issue, the hydrogen corona should not be mentioned in this context and so we have removed the relevant text from the manuscript.

Line 528 “What can a comparison of Uranus’ natural satellites and the captured “Ocean World” Triton reveal about…” This sounds like you are only comparing Triton and Uranus's natural satellites. I think the comparison between Triton and Uranus's natural satellites to other similar ocean worlds and bodies in the solar system is equally (if not more) important. We have added this clause to our sentence.

Line 532: “The Neptunian system is dominated by the cuckoo-like arrival of…” I’m not sure what “cuckoo-like” means. It’s a turn of phrase meaning that the arrival of Triton displace the primordial satellite system - we’ve added this explanation.

Line 599: “The large moon Triton, …” I would suggest starting this with “Neptune’s large moon Triton…”. I realize that everyone reading this paper will understand this, but it helps saves starting a sentence with “The”, and also for those who are not as familiar, it again places Triton in the Neptune system. Changed.

Line 663: “Summary: Exploring Uranus’ natural satellites and Neptune’s captured moon Triton will reveal how…” Would it be better to say “Uranus’s and Neptune's natural satellites and Neptune’s captured moon Triton"?” Changed.

Section 2.4 Ice Giant Satellites – Natural and Captured seems to really emphasize the Uranian natural satellites, and not so much Neptune’s. Within this section there are the following key questions: – What can the geological diversity of the large icy satellites of Uranus reveal about the formation and continued evolution of primordial satellite systems? – What was the influence of tidal interaction and internal melting on shaping the Uranian worlds, and could internal water oceans still exist? – What is the chemical composition of the surfaces of the Uranian moons? – Does Triton currently harbour a subsurface ocean and is there evidence for recent, or ongoing, active exchange with its surface? – Are seasonal changes in Triton’s tenuous atmosphere linked to specific sources and sinks on the surface, including its remarkable plume activity? – Are the smaller satellites of Neptune primordial? – How does an Ice Giant satellite system interact with the planets’ magnetospheres? I’m not expert enough on the satellite systems of the ice giants to judge whether this is appropriate or not, but if there is indeed a higher emphasis placed on the moons of Uranus, then the rationale for this might be worth a short explanation. The key questions that you have listed above break down into 3 for Uranus, 2 for Triton, 1 for Neptune’s, and one for the interaction with the magnetosphere. We believe the balance between the two systems is fair in this section, and have included a line in the summary to state that we believe both systems to be equally compelling for future orbital exploration.

Line 757: “Thus, there is no substitute for orbital exploration of one or both of these worlds, but…” But please don’t forget the value of probes! We have added the mention of proves here.

Line 857: “found that an Ice Giant orbiter, to either Uranus or Neptune, would provide an unprecedented leap in our understanding of these enigmatic worlds. An orbiter would maximise the time spent in the system to conduct science of interest to the entire planetary community.” Suggestion (and I know you’re tired of hearing me say this!): “…found that an Ice Giant orbiter and probe to either Uranus or Neptune would provide…” Added.

Line 884: “Bepi Colombo” “BepiColombo” (no space) Changed,

Line 893: “…would be needed for exploration of the satellites, rings and atmosphere from a low-inclination orbit .” Also note that the requirements for a probe data return trajectory must also be considered. Added.

Line 917: “The availability of Jupiter, as the largest planet, is key to optimal launch trajectories, and the early 2030s offer the best opportunities.” Not so much for Neptune. Neptune trajectories past Jupiter require 2029-2030 or so, I believe. This is explained in the next sentence or two, and there are some Neptune opportunities after 2030, just strongly constrained on mass delivered to the system. Studies using powerful launch vehicles are also assessing whether direct trajectories are possible, and we have added this caveat.

Line 965: “4.3 Mature and Developing Technologies” I think it would be useful to provide a few more references for some of the listed technologies for those who want to delve into a bit more detail. One example is a reference for ESA’s 241Am power sources (line 979). We have added a reference to a recent review paper by Ambrosi et al. 2019.

Line 987: Given the 6-13-year interplanetary transfer” Suggest either “6 to 13-year” or “6- to 13-year”. Changed.

Line 996: “European development or access to US 70-m Deep Space Network capabilities with state-of-the-art Ka-band technology would greatly improve the situation, but even here the bandwidth would remain small.” How small is small? What order of magnitude are we talking about? Please see our response to the next query, where we provide the ball-park estimates.

Line 1001: “Nevertheless, the 2018 ESA CDF study confirmed that the data volume of a realistic M-class mission scenario to Uranus…” I'm not sure what elements of the mission this includes. Orbiter only, or probe-orbiter? We have significantly modified this section, including the following information: “The available power, length of the downlink window, and the need to balance science data, engineering and housekeeping telemetry, determines the overall data rate. For example, the NASA/JPL mission study in 2017 suggested that use of a 35-W traveling wave tube amplifier (the current power limit for space-qualified TWTAs) and a 34-m ground-station could provide 15 kbps at Uranus, whereas this could increase to 30 kbps with a 70-W TWTA. The 2018 ESA CDF study suggested that the data rate of a future M-class orbital mission to Uranus could use a 100-W TWTA to deliver 94 kbps Ka-band downlink from Uranus, or 42 kbps from Neptune. Assuming 3.2 hours/day for communications, this equates to daily data volumes for science of 1.09 and 0.48 Gb, respectively.”

Line 1005: Electromagnetic cleanliness: A number of the proposed payload elements (magnetometers, plasma and radio instruments) impose strict EMC requirements on the spacecraft, and…” Why are EMC requirements more stringent than Cassini, requiring new technologies and cost optimization strategies? This sentence was a holdover from the original white paper and, on reflection, we decided to omit it from the final journal article.

Line 1009: “Launch Vehicles: The market for launch vehicles is changing dramatically both in Europe and in the US, but an Ice Giant System cornerstone mission could take advantage of the latest technologies, delivering as much mass to the system as possible within the constraints of a realistic flight time. Larger launch vehicles may also open up the prospect for multiple spacecraft to share the same faring.” This sentence really doesn't say much that isn’t somewhat obvious. Can anything be said about the delivered mass and timeline of current launch vehicle technology vs. the improvement expected from new launch vehicles? We are aware of detailed trade-off studies that explore possible direct trajectories in comparison to the Jupiter GA and available launchers, however we are not in the position to quantify them in a meaningful way for this article. We have updated this text to advocate for such studies: “The market for launch vehicles is changing dramatically both in Europe and in the US. Ice Giant mission concepts have traditionally considered Jupiter gravity assists to provide realistic flight times and sufficient payload delivered to each system. However, we advocate investigation of the direct-transfer trajectories enabled by the heavy-lift capacity of the next generation of launch vehicles which may (i) obviate the need for Jupiter/Saturn GA; and (ii) open up the possibilities for launching multiple spacecraft that share the same faring.”

Line 1015: “This article reviews and updates the scientific rationale for a mission to an Ice Giant system, advocating that an orbital mission (alongside an atmospheric entry probe) be considered as a cornerstone of ESA’s Voyage 2050 programme,….” (Almost my last comment on probes, promise!) This paper doesn't really seem to “advocate” for a probe mission as an important element. There are brief mentions of probes, but it is largely discussing the value of orbiter science. Whilst you are correct that this article deals primarily with the potential for orbital science, we choose not to change this statement. However, we have added references to Mousis et al. (2018) which does this job for probes.

Can Figure 9 be reformatted slightly so there is a line between first and second themes on right hand side? Done.

Line 1040: “Most importantly, the Ice Giant System mission will continue the breath-taking legacy of discovery of the Voyager, Galileo, Cassini, Juno and JUICE missions to the giant planets.” You can add Europa Clipper to this, although JUICE and Europa Clipper haven't provided the breath-taking legacies yet. This has been modified.

Line 1041: “A dedicated orbiter of an Ice Giant is the next logical step in our exploration of the Solar System, completing humankind’s first reconnaissance of …” I would suggest that a dedicated orbiter is “one element” of the next logical step. Yes - but we see this as the highest priority for exploration of the Ice Giant systems, and so retain the wording in the conclusion.

References The following references have a typo at the end (instead of ending with a period, they end with “,.”: This is a result of missing information using the NASA/ADS system of referencing, we’ll make sure they’re removed globally.

Bailey, 2015 Croft, 1991 Farrell, 1992 French, 1991 Kempf, 2018 Krasnopolsky, 1995 Moreno, 2009 Roman, 2019 Rymer, 2019 Safronov, 1966 Simon, 2018 Stevenson, 1986 Stryk, 2008

Other Notes Běhounkova, M., Not in alphabetical order (follows Buchavarova) Changed. Decker, 1994: Text says 1987. Changed. Kegerreis, 2019: Text says 2018. Changed. Majeed, 2006: Text says 2004. Changed. Millot, 2019: Text says 2018 Changed. Plescia, J. B. (1987), 2 entries. Text says 1987a and 1987b. Not clear which is which. Changed. Rymer, 2019. Text says 2018. Changed. Safronov, 1966 out of alphabetical order. Previous entry is Sanchez-Lavega. Changed Zarka, 1986: Text shows both Zarka, 1986 and Zarka, 1995. The latter reference has been added. Podolak, et al., 1995 does not appear in references. Added. Herbert, et al. 1996 does not appear in references. This has been removed from the manuscript. Desch et al., 1991 does not appear in references. Added. Broadftoot et al., 1989 misspelled in text Changed.

Following references could not be found in text Christophe, B., et al. (2012) - Removed. de Pater, I et al. (1991) - Added. Helled, R. (2018), The Interiors of Jupiter and Saturn, Oxford Research Encyclopedia of Planetary Science, 175, 10.1093/acrefore/9780190647926.013.175. - Removed. Helled, R., & Guillot, T. (2018), - Removed. Podolak, M., Helled, R., & Schubert, G. (2019), - Removed.

Reviewer 2:

This is a paper on future exploration of ice giants, based on the white paper submitted to ESA as part of the Voyages 2050 exercise to define the next science missions in Europe. The paper is clear, focussed, well written, and deserves a fast publication in PSS. Many thanks for your review of our article.

Fletcher et al., Ice Giant Systems

1 Ice Giant Systems:

2 The Scientific Potential of Orbital Missions to Uranus and Neptune

3

4 Leigh N. Fletcher (School of Physics and Astronomy, University of Leicester, University Road, Leicester, LE1

5 7RH, UK; [email protected]; Tel: +441162523585), Ravit Helled (University of Zurich, Switzerland), Elias

6 Roussos (Max Planck Institute for Solar System Research, Germany), Geraint Jones (Mullard Space Science

7 Laboratory, UK), Sébastien Charnoz (Institut de Physique du Globe de Paris, France), Nicolas André (Institut

8 de Recherche en Astrophysique et Planétologie, Toulouse, France), David Andrews (Swedish Institute of

9 Space Physics, Sweden), Michele Bannister (Queens University Belfast, UK), Emma Bunce (University of

10 Leicester, UK), Thibault Cavalié (Laboratoire d’Astrophysique de Bordeaux, Univ. Bordeaux, CNRS, Pessac,

11 France; LESIA, Observatoire de Paris, Université PSL, CNRS, Sorbonne Université, Univ. Paris Diderot,

12 Sorbonne Paris Cité, Meudon, FranceLaboratoire d’Astrophysique de Bordeaux, Univ. Bordeaux, CNRS, B18N,

13 allée Geoffroy Saint-Hilaire, 33615 Pessac, France; LESIA, Observatoire de Paris, Université PSL, CNRS,

14 Sorbonne Université, Univ. Paris Diderot, Sorbonne Paris Cité, 5 place Jules Janssen, 92195 Meudon,

15 FranceUniversité de Bordeaux, France), Francesca Ferri (Università degli Studi di Padova, Italy), Jonathan

16 Fortney (University of Santa-Cruz, USA), Davide Grassi (Istituto Nazionale di AstroFisica – Istituto di

17 Astrofisica e Planetologia Spaziali (INAF-IAPS), Rome, Italy), Léa Griton (Institut de Recherche en

18 Astrophysique et Planétologie, CNRS, Toulouse, France), Paul Hartogh (Max-Planck-Institut für

19 Sonnensystemforschung, Germany), Ricardo Hueso (Escuela de Ingeniería de Bilbao, UPV/EHU, Bilbao,

20 Spain), Yohai Kaspi (Weizmann Institute of Science, Israel), Laurent Lamy (Observatoire de Paris, France),

21 Adam Masters (Imperial College London, UK), Henrik Melin (University of Leicester, UK), Julianne Moses

22 (Space Science Institute, USA), Oliver Mousis (Aix Marseille Univ, CNRS, CNES, LAM, Marseille, France),

23 Nadine Nettleman (University of Rostock, Germany), Christina Plainaki (ASI - Italian Space Agency, Italy),

24 Jürgen Schmidt (University of Oulu, Finland), Amy Simon (NASA Goddard Space Flight Center, USA), Gabriel

25 Tobie (Université de Nantes, France), Paolo Tortora (Università di Bologna, Italy), Federico Tosi (Istituto

26 Nazionale di AstroFisica – Istituto di Astrofisica e Planetologia Spaziali (INAF-IAPS), Rome, Italy), Diego Turrini

27 (Istituto Nazionale di AstroFisica – Istituto di Astrofisica e Planetologia Spaziali (INAF-IAPS), Rome, Italy) Fletcher et al., Ice Giant Systems

28

29

30 Abstract:

31 Uranus and Neptune, and their diverse satellite and ring systems, represent the least explored environments

32 of our Solar System, and yet may provide the archetype for the most common outcome of planetary

33 formation throughout our galaxy. Ice Giants will be the last remaining class of Solar System planet to have a

34 dedicated orbital explorer, and international efforts are under way to realise such an ambitious mission in

35 the coming decades. In 2019, the European Space Agency released a call for scientific themes for its strategic

36 science planning process for the 2030s and 2040s, known as Voyage 2050. We used this opportunity to

37 review our present-day knowledge of the Uranus and Neptune systems, producing a revised and updated set

38 of scientific questions and motivations for their exploration. This review article describes how such a mission

39 could explore their origins, ice-rich interiors, dynamic atmospheres, unique magnetospheres, and myriad icy

40 satellites, to address questions at the heart of modern planetary science. These two worlds are superb

41 examples of how planets with shared origins can exhibit remarkably different evolutionary paths: Neptune

42 as the archetype for Ice Giants, whereas Uranus may be atypical. Exploring Uranus’ natural satellites and

43 Neptune’s captured moon Triton could reveal how Ocean Worlds form and remain active, redefining the

44 extent of the habitable zone in our Solar System. For these reasons and more, we advocate that an Ice Giant

45 System explorer should become a strategic cornerstone mission within ESA’s Voyage 2050 programme, in

46 partnership with international collaborators, and targeting launch opportunities in the early 2030s.

47

48 Keywords: Giant Planets; Ice Giants; Robotic Missions; Orbiters; Probes

49

50 1. Introduction: Why Explore Ice Giant Systems?

51 1.1 Motivations

52

2 Fletcher et al., Ice Giant Systems

Figure 1 Each Ice Giant exhibits a rich system of planetary environments to explore, from their mysterious interiors, atmospheres and magnetospheres, to the diverse satellites and rings. The inner systems are to scale, with arrows next to major moons indicating that they orbit at larger planetocentric distances. The magnetosphere and radiation belts would encompass the full area of the figure. Credit: L.N. Fletcher/M. Hedman/E. Karkoschka/Voyager-2. 53 The early 21st century has provided unprecedented leaps in our exploration of the Gas Giant systems, via the

54 completion of the Galileo and Cassini orbital missions at Jupiter and Saturn; NASA/Juno’s ongoing exploration

55 of Jupiter’s deep interior, atmosphere, and magnetic field; and ESA’s development of the JUICE mission

56 (Jupiter Icy Moons Explorer) and NASA’s Europa Clipper to explore the Galilean satellites. The past decade

57 has also provided our first glimpses of the diversity of planetary environments in the outer solar system, via

58 the New Horizons mission to Pluto. Conversely, the realm of the Ice Giants, from Uranus (20 AU) to Neptune

59 (30 AU), remains largely unexplored, each world system having been visited only once by a flyby spacecraft

60 – Voyager 2 – in 1986 and 1989 respectively. More than three decades have passed since our first close-up

61 glimpses of these worlds, with cameras, spectrometers and sensors based on 1960s and 70s technologies.

62 Voyager’s systems were not optimised for the Ice Giants, which were considered to be extended mission

63 targets. Uranus and Neptune have therefore never had a dedicated mission, despite the rich and diverse

64 systems displayed in Figure 1. A return to the Ice Giants with an orbiter is the next logical step in humankind’s

65 exploration of our Solar System.

66

67 The Ice Giants may be our closest and best representatives of a whole class of astrophysical objects, as

68 Neptune-sized worlds have emerged as the dominant category in our expanding census of exoplanets (Fulton

69 et al., 2018), intermediate between the smaller terrestrial worlds and the larger hydrogen-rich gas giants 3 Fletcher et al., Ice Giant Systems

70 (Section 3.3). Our own Ice Giants offer an opportunity to explore physical and chemical processes within

71 these planetary systems as the archetype for these distant exoplanets (Rymer et al., 20198). Furthermore,

72 the formation and evolution of Uranus and Neptune (Section 2.1) pose a critical test of our understanding of

73 planet formation, and of the architecture of our Solar System. Their small size, compared to Jupiter, places

74 strong constraints on the timing of planet formation. Their bulk internal composition (i.e., the fraction of

75 rock, ices, and gases) and the differentiation with depth (i.e., molecular weight gradients, phase transitions

76 to form global water oceans and icy mantles) are poorly known, but help determine the conditions and

77 dynamics in the outer planetary nebula at the time of planet formation.

78

79 Uranus and Neptune also provide two intriguing endmembers for the Ice Giant class. Neptune may be

80 considered the archetype for a seasonal Ice Giant, whereas the cataclysmic collision responsible for Uranus’

81 extreme tilt renders it unique in our Solar System. Contrasting the conditions on these two worlds provides

82 insights into differential evolution from shared origins. The atmospheres of Uranus and Neptune (Section

83 2.2) exemplify the contrasts between these worlds. Uranus’ negligible internal heat renders its atmosphere

84 extremely sluggish, with consequences for storms, meteorology, and atmospheric chemistry. Conversely,

85 Neptune’s powerful winds and rapidly-evolving storms demonstrates how internal energy can drive powerful

86 weather, despite weak sunlight at 30 AU. Both of these worlds exhibit planetary banding, although the

87 atmospheric circulation responsible for these bands (and their associated winds, temperatures, composition

88 and clouds) remain unclear, and the connection to atmospheric flows below the topmost clouds remains a

89 mystery.

90

91 Conditions within the Ice Giant magnetospheres (Section 2.3) are unlike those found anywhere else, with

92 substantial offsets between their magnetic dipole axes and the planet’s’ rotational axes implying a system

93 with an extremely unusual interaction with the solar wind and internal plasma processes, varying on both

94 rotational cycles as the planet spins, and on orbital cycles.

95

4 Fletcher et al., Ice Giant Systems

96 The diverse Ice Giant satellites (Section 2.4) and narrow, incomplete ring systems (Section 2.5) provide an

97 intriguing counterpoint to the better-studied Jovian and Saturnian systems. Uranus may feature a natural,

98 primordial satellite system with evidence of extreme and violent events (e.g., Miranda). Neptune hosts a

99 captured Kuiper Belt Object, Triton, which may itself harbour a sub-surface ocean giving rise to active surface

100 geology (e.g., south polar plumes and cryovolcanism).

101

102 Advancing our knowledge of the Ice Giants and their diverse satellite systems requires humankind’s first

103 dedicated explorer for this distant realm. Such a spacecraft should combine interior science via gravity and

104 magnetic measurements, in situ measurements of their plasma and magnetic field environments, in situ

105 sampling of their chemical composition, and close-proximity multi-wavelength remote sensing of the planets,

106 their rings, and moons. This review article will summarise our present understanding of these worlds, and

107 propose a revised set of scientific questions to guide our preparation for such a mission. This article is

108 motivated by ESA’s recent call for scientific themes as part of its strategic space mission planning in the period

109 from 2035 to 2050,1 and is therefore biased to European perspectives on an Ice Giant mission, as explored in

110 the next section.

111

112 1.2 Ice Giants in ESA’s Cosmic Vision

113

114 The exploration of the Ice Giants addresses themes at the heart of ESA’s existing Cosmic Vision2 programme,

115 namely (1) exploring the conditions for planet formation and the emergence of life; (2) understanding how

116 our solar system works; and (3) exploring the fundamental physical laws of the universe. European-led

117 concepts for Ice Giant exploration have been submitted to several ESA Cosmic Vision competitions. The

118 Uranus Pathfinder mission, an orbiting spacecraft based on heritage from Mars Express and Rosetta, was

119 proposed as a medium-class (~€0.5bn) mission in both the M3 (2010) and M4 (2014) competitions (Arridge

1 https://www.cosmos.esa.int/web/voyage-2050 2 http://sci.esa.int/cosmic-vision/38542-esa-br-247-cosmic-vision-space-science-for-europe-2015-2025/

5 Fletcher et al., Ice Giant Systems

120 et al., 2012). However, the long duration of the mission, limited power available, and the programmatic

121 implications of having NASA provide the launch vehicle and radioisotope thermoelectric generators (RTGs),

122 meant that the Pathfinder concept did not proceed to the much-needed Phase A study.

123

124 The importance of Ice Giant science was reinforced by multiple submissions to ESA's call for large-class

125 mission themes in 2013: a Uranus orbiter with atmospheric probe (Arridge et al., 2014), an orbiter to explore

126 Neptune and Triton (Masters et al., 2014); and a concept for dual orbiters of both worlds (Turrini et al., 2014).

127 Once again, an ice giant mission failed to proceed to the formal study phase, but ESA's Senior Survey

128 Committee (SSC3) commented that ``the exploration of the icy giants appears to be a timely milestone, fully

129 appropriate for an L class mission. The whole planetology community would be involved in the various aspects

130 of this mission... the SSC recommends that every effort is made to pursue this theme through other means,

131 such as cooperation on missions led by partner agencies."

132

133 This prioritisation led to collaboration between ESA and NASA in the formation of a science definition team

134 (2016-17), which looked more closely at a number of different mission architectures for a future mission to

135 the Ice Giants (Hofstadter et al., 2019). In addition, ESA's own efforts to develop nuclear power sources for

136 space applications have been progressing, with prototypes now developed to utilise the heat from the decay

137 of 241Am as their power source (see Section 4.3), provided that the challenge of their low energy density can

138 be overcome. Such an advance might make smaller, European-led missions to the Ice Giants more realistic,

139 and addresses many of the challenges faced by the original Uranus Pathfinder concepts.

140

141 At the start of the 2020s, NASA and ESA are continuing to explore the potential for an international mission

142 to the Ice Giants. A palette of potential contributions (M-class in scale) to a US-led mission have been

143 identified by ESA4, and US scientists are currently undertaking detailed design and costing exercises for

3 http://sci.esa.int/cosmic-vision/53261-report-on-science-themes-for-the-l2-and-l3-missions/ 4 http://sci.esa.int/future-missions-department/61307-cdf-study-report-ice-giants/

6 Fletcher et al., Ice Giant Systems

144 missions to be assessed in the upcoming US Planetary Decadal Survey (~2022). Each of these emphasise

145 launch opportunities in the early 2030s (Section 4.2), with arrival in the early 2040s (timelines for Ice Giant

146 missions will be described in Section 4.3). This review article summarises those studies, whilst taking the

147 opportunity to update and thoroughly revise the scientific rationale for Ice Giant missions compared to

148 Arridge et al. (2012, 2014), Masters et al. (2014), Turrini et al. (2014) and Hofstadter et al. (2019). We focus

149 on the science achievable from orbit, as the science potential of in situ entry probes has been discussed

150 elsewhere (Mousis et al., 2018). Section 2 reviews our present-day knowledge of Ice Giant Systems; Section

151 3 places the Ice Giants into their wider exoplanetary context; Section 4 briefly reviews the recent mission

152 concept studies and outstanding technological challenges; and Section 5 summarises our scientific goals at

153 the Ice Giants at the start of the 2020s.

154

155 2. Science Themes for Ice Giant Exploration

156

157 In this section we explore the five multi-disciplinary scientific themes that could be accomplished via

158 orbital exploration of the Ice Giants, and show how in-depth studies of fundamental processes at Uranus and

159 Neptune would have far-reaching implications in our Solar System and beyond. Each sub-section is organised

160 via a series of high-level questions that could form the basis of a mission traceability matrix.

161

162 2.1 Ice Giant Origins and Interiors

163

164 What does the origin, structure, and composition of the two Ice Giants reveal about the formation of

165 planetary systems? Understanding the origins and internal structures of Uranus and Neptune will

166 substantially enhance our understanding of our own Solar System and lowintermediate-mass exoplanets.

7 Fletcher et al., Ice Giant Systems

167 Their bulk composition provides crucial constraints on the conditions in the solar nebula during planetary

168 formation and evolution.

169

170 How did the Ice Giants first form, and what constraints can be placed on the mechanisms for planetary

171 accretion? The formation of Uranus and Neptune has been a long-standing problem for planet formation

172 theory (e.g., Pollack et al., 1996, Dodson-Robinson & Bodenheimer, 2010, Helled & Bodenheimer, 2014). Yet,

173 the large number of detected exoplanets with sizes comparable (or smaller) to that of Uranus and Neptune

174 suggests that such planetary objects are very common (e.g., Batalha et al. 2013), a fact that is in conflict with

175 theoretical calculations.

176

177 The challenge for formation models is to prevent Uranus and Neptune from accreting large amounts of

178 hydrogen-helium (H-He) gas, like the Gas Giants Jupiter and Saturn, to provide the correct final mass and gas-

179 to-solids ratios as inferred by structure models. In the standard planet formation model, core accretion (see

180 Helled et al., 2014 for review and the references therein), a slow planetary growth is expected to occur at

181 large radial distances where the solid surface density is lower, and the accretion rate (of planetesimals) is

182 significantly smaller. For the current locations of Uranus and Neptune, the formation timescale can be

183 comparable to the lifetimes of protoplanetary disks. Due to long accretion times at large radial distances, the

184 formation process is too slow to reach the phase of runaway gas accretion, before the gas disk disappears,

185 leaving behind an intermediate-mass planet (a failed giant planet), which consists mostly of heavy elements

186 and a small fraction of H-He gas.

187

188 However, since the total mass of H-He in both Uranus and Neptune is estimated to be 2-3 Earth masses (M),

189 it implies that gas accretion had already begun, and this requires that the gas disk disappears at a very specific

190 time, to prevent further gas accretion onto the planets. This is known as the fine-tuning problem in

191 Uranus/Neptune formation (e.g., Venturini & Helled, 2017). Another possibility is that Uranus and Neptune

192 formed in situ within a few Mys by pebble accretion. In this formation scenario, the core’s growth is more

8 Fletcher et al., Ice Giant Systems

193 efficient than in the planetesimal accretion case, and the pebble isolation mass is above 20 M. As a result,

194 the forming planets could be heavy-element dominated with H-He envelopes that are metal-rich due to

195 sublimation of icy pebbles (e.g., Lambrechts et al., 2014).

196

197 Measuring the elemental abundances in the atmospheres of Uranus and Neptune can provide information

198 on their formation history by setting limits on their formation locations and/or the type of solids

199 (pebbles/planetesimals) that were accreted. Measurements of the elemental abundances of well-mixed

200 noble gases, which are only accessible via in situ entry probes, would be particularly informative (e.g. Mousis

201 et al., 2018). In addition, determining the atmospheric metallicity provides valuable constraints for structure

202 models, as discussed below.

203

204 What is the role of giant impacts in explaining the differences between Uranus and Neptune? Uranus and

205 Neptune are somewhat similar in terms of mass and radius, but they also have significant differences such

206 as their tilt, internal heat flux, and satellite system. It is possible that these observed differences are a result

207 of giant impacts that occurred after the formation of the planets (e.g., Safronov 1966, Stevenson 1986). An

208 oblique impact of a massive impactor can explain Uranus’ spin and lead to the formation of a disk where the

Figure 2 Sketches of the possible internal structures of the ice giants. It is unclear whether the planets are differentiated and whether the transitions between the different layers are distinct or gradual: (a) separation between the ices and rocks and the ice and H-He atmosphere (b) separation (phase boundary) between the H-He atmosphere and ices and a gradual transition between ice and rock, (c) gradual transition between the H-He atmosphere and ice layer, and a distinct separation between the ice and rock layers, and (d) gradual transition both between the H-He atmosphere and ice and the ice and rocks suggesting a global composition gradient with the planets (see text for discussion). 9 Fletcher et al., Ice Giant Systems

209 regular satellites form. Neptune could have also suffered a head-on impact that could have reached the deep

210 interior, providing sufficient energy (and mass) to explain the higher heat flux, and possibly the higher mass

211 and moment of inertia value (e.g., Stevenson 1986, Podolak & Helled, 2012). Giant impact simulations by

212 various groups confirmed that Uranus’ tilt and rotation could be explained by a giant impact (Slattery et al.,

213 1992, Kegerreis et al., 20198). Nevertheless, alternative explanations have been proposed: Boue and Laskar

214 (2010) showed that under special circumstances, the high obliquity could potentially arise from orbital

215 migration without the need for a collision. such as orbital migration cannot be excluded (e.g., Boue & Laskar

216 2010). Understanding the cause of Uranus’ tilt and the mechanisms that led to the observed differences

217 between the planets are key questions in planetary science.

218

219 What is the bulk composition and internal structure of Uranus and Neptune? There are still substantial

220 uncertainties regarding the bulk compositions and internal structures of Uranus and Neptune (e.g., Podolak

221 et al., 1995, Podolak & Helled, 2012, Nettelmann et al., 2013). The available measurements of their physical

222 properties such as mass, radius, rotation rate, atmospheric temperature, and gravitational and magnetic

223 fields are used to constrain models of their interiors. For the Ice Giants, only the gravitational harmonic

224 coefficients J2 and J4 are known and their error bars are relatively large (Jacobson, 2009; 2014), nevertheless,

225 various studies have aimed to constrain the planets’ internal structures.

226

227 Standard structure models of the planets consist of three layers: a rocky core, an 'icy' shell (water, ammonia,

228 methane, etc.), and a gaseous envelope composed of H-He and heavier elements. The middle layer is not

229 made of “ice” in regard to the physical state of the material (i.e., solid), but is referred to as an icy layer since

230 it is composed of volatile materials such as water, ammonia and methane. The masses and compositions of

231 the layers are modified until the model fits the observed parameters using a physical equation of state (EOS)

232 to represent the materials. Three-layer models predict very high ice-to-rock ratios, where the ice fraction is

233 found to be higher than the rock fraction by 19-35 times for Uranus, and 4-15 times for Neptune, with the

234 total H-He mass typically being 2 and 3 M for Uranus and Neptune, respectively. The exact estimates are

10 Fletcher et al., Ice Giant Systems

235 highly model-dependent, and are sensitive to the assumed composition, thermal structure and rotation rate

236 of the planets (Helled et al., 2011, Nettelmann et al., 2013).

237

238 The interiors of Uranus and Neptune could also be more complex with the different elements being mixed,

239 and could also include thermal boundary layers and composition gradients. Indeed, alternative structure

240 models of the planets suggested that Uranus and Neptune could have a density profile without

241 discontinuities (e.g., Helled et al., 2011), and that the planets do not need to contain large fractions of water

242 to fit their observed properties. This alternative model implies that Uranus and Neptune may not be as water-

243 rich as typically thought, but instead are rock-dominated like Pluto (e.g., McKinnon et al., 2017) and could be

244 dominated by composition gradients. It is therefore possible that the “ice giants” are in fact not ice-

245 dominated (see Helled et al., 2011 for details). The very large ice-to-rock ratios found from structure models

246 also suggest a more complex interior structure. At the moment, we have no way to discriminate among the

247 different ice-to-rock ratios inferred from structure models. As a result, further constraints on the gravity and

248 magnetic data (Section 2.3), as well as atmospheric composition and isotopic ratios (e.g., D/H, Atreya et al.,

249 2020) are required.

250

251 How can Ice Giant observations be used to explore the states of matter (e.g., water) and mixtures (e.g.,

252 rocks, water, H-He) under the extreme conditions of planetary interiors? In order to predict the mixing

253 within the planets, knowledge from equation-of-state (EOS) calculations is required. Internal structure

254 models must be consistent with the phase diagram of the assumed materials and their mixtures. This is a

255 challenging task and progress in that direction is ongoing. EOS calculations can be used to guide model

256 assumptions. For example, it is possible that Uranus and Neptune have deep water oceans that begin where

257H 2 and H2O become insoluble (e.g., Bailey & Stevenson, 2015, Bali et al., 2013). Figure 2 presents sketches

258 of four possible internal structures of the ice giants where the transitions between layers are distinct (via

259 phase/thermal boundariesy) and/or gradual.

260

11 Fletcher et al., Ice Giant Systems

261 Current observational constraints, foremost J2 and J4, clearly indicate that the deep interior is more enriched

262 in heavy elements than the outer part. Understanding the nature and origin of the compositional gradient

263 zone would yield important information on the formation process and subsequent evolution including

264 possible processes such as outgassing, immiscibility, and sedimentation of ices; processes that play a major

265 role on terrestrial planets and their habitability.

266

267 What physical and chemical processes during the planetary formation and evolution shape the magnetic

268 field, thermal profile, and other observable quantities? Structure models must be consistent with the

269 observed multi-polar magnetic fields (see Section 2.3), implying that the outermost ~20% of the planets is

270 convective and consists of conducting materials (e.g., Stanley & Bloxham, 2004, 2006). Currently, the best

271 candidate for the generation of the dynamo is the existence of partially dissociated fluid water in the

272 outermost layers (e.g., Redmer et al.,2011), located above solid and non-convecting superionic water ice

273 ‘mantle’ (Millot et al., 20198). Dynamo models that fit the Voyager magnetic field data suggest the deep

274 interior is stably stratified (Stanley & Bloxham 2004, 2006) or, alternatively, in a state of thermal-buoyancy

275 driven turbulent convection (Soderlund et al., 2013). Improved measurements of the magnetic fields of

276 Uranus and Neptune will also help to constrain the planetary rotation rate. Since Voyager’s measurements

277 of the periodicities in the radio emissions and magnetic fields have not been confirmed by another

278 spacecraft, it is unclear whether the Voyager rotation rate reflects the rotation of the deep interior (Helled

279 et al., 2010), with major consequences for the inferred planetary structure and the question of similar or

280 dissimilar interiors (Nettelmann et al., 2013).

281

282 Finally, the different intrinsic heat fluxes of Uranus and Neptune imply that they followed different

283 evolutionary histories. This could be a result of a different growth history or a result of giant impacts during

284 their early evolution (e.g., Reinhardt et al., 2020). Moreover, thermal evolution models that rely on Voyager’s

285 measurements of the albedo, brightness temperatures, and atmospheric pressure-temperature profiles that

286 are used to model the evolution of the two planets cannot explain both planets with the same set of

12 Fletcher et al., Ice Giant Systems

287 assumptions.

288 Finally, the different intrinsic heat fluxes of Uranus and Neptune, together with their similar mass and radius

289 values suggest that the planets formed in a similar way, but then followed different evolutionary histories.

290 Moreover, thermal evolution models that rely on Voyager’s measurements of the albedo, brightness

291 temperatures, and atmospheric pressure-temperature profiles inferred from the occultation data cannot

292 explain both planets with the same set of assumptions.

293

294

295

296 Summary: A better understanding of the origin, evolution and structure of Ice Giant planets requires new

297 and precise observational constraints on the planets' gravity field, rotation rate, magnetic field, atmospheric

298 composition, and atmospheric thermal structure, both from orbital observations and in situ sampling from

299 an atmospheric probe. The insights into origins, structures, dynamo operation and bulk composition

300 provided by an Ice Giant mission would not only shed light on the planet-forming processes at work in our

301 Solar System, but could also help to explain the most common planetary class throughout our observable

302 universe.

303

304 2.2 Ice Giant Atmospheres

305

306 Why do atmospheric processes differ between Uranus, Neptune, and the Gas Giants, and what are the

307 implications for Neptune-mass worlds throughout our Universeuniverse? Ice Giant atmospheres occupy a

308 wholly different region of parameter space compared to their Gas Giant cousins. Their dynamics and

309 chemistry are driven by extremes of internal energy (negligible on Uranus, but powerful on Neptune) and

310 extremes of solar insolation (most severe on Uranus due to its 98o axial tilt) that are not seen anywhere else

311 in the Solar System (Figure 3). Their smaller planetary radii, compared to Jupiter and Saturn, affects the width

312 of zonal bands and the drift behaviour of storms and vortices. Their zonal winds are dominated by broad

13 Fletcher et al., Ice Giant Systems

313 retrograde equatorial jets and do not exhibit the fine-scale banding found on Jupiter, which means that

314 features like bright storms and dark vortices are able to drift with latitude during their lifetimes. Their

315 hydrogen-helium atmospheres are highly enriched in volatiles like CH4 and H2S that show strong equator-to-

316 pole gradients, changing the atmospheric density and hence the vertical shear on the winds (Sun et al., 1991).

317 Their temperatures are so low that the energy released from interconversion between different states of

318 hydrogen (ortho and para spin isomers) can play a role in shaping atmospheric dynamics (Smith & Gierasch,

319 1995). Their middle and upper atmospheres are both much hotter than can be explained by weak solar

320 heating alone, implying a decisive role for additional energy from internal (e.g., waves) or external sources

321 (e.g., currents induced by complex coupling to the magnetic field). As the atmospheres are the windows

322 through which we interpret the bulk properties of planets, these defining properties of Ice Giants can

323 provide insights into atmospheric processes on intermediate-sized giant planets beyond our Solar System.

324

325 A combination of global multi-wavelength remote sensing from an orbiter and in situ measurements from an

326 entry probe would provide a transformative understanding of these unique atmospheres, focussing on the

327 following key questions (summarised in Figure 3):

328

329 What are the dynamical, meteorological, and chemical impacts of the extremes of planetary luminosity?

330 Despite the substantial differences in self-luminosity (Pearl et al., 1990, 1991), seasonal influences,

331 atmospheric activity (Hueso and Sanchez-Lavega et al., 20182019), and the strength of vertical mixing

332 (resulting in differences in atmospheric chemistry, Moses et al., 2018), there are many similarities between

333 these two worlds. In their upper tropospheres, tracking of discrete cloud features has revealed that both

334 planets exhibit broad retrograde jets at their equators and prograde jets nearer the poles (Sanchez-Lavega

335 et al., 2018), but unlike Jupiter, these are seemingly disconnected from the fine-scale banding revealed in

336 the visible and near-infrared (Sromovsky et al., 2015). Is this simply an observational bias, or are winds on

337 the ice giants truly different from those on Jupiter and Saturn? What sets the scales of the bands? On the

338 Gas Giants, small-scale eddies (from atmospheric instabilities and convective storms) appear to feed energy

14 Fletcher et al., Ice Giant Systems

Figure 3 Ice Giant atmospheres are shaped by dynamical, chemical and radiative processes that are not found elsewhere in our Solar System. Images A & C are false-colour representations of Voyager 2 observations of Uranus and Neptune, respectively. Images B and D were acquired by the Hubble Space Telescope in 2018.

339 into the large-scale winds, but we have never been able to investigate similar processes on Uranus and

340 Neptune. Indeed, convective processes themselves could be substantially different – moist convection

341 driven by the condensation of water will likely play a very limited role in the observable atmosphere, as water

342 is restricted to pressures that exceed tens or hundreds of bars. Instead, convection in the observable

343 tropospheres may be driven by methane condensation in the 0.1-1.5 bar range (Stoker & Toon, 1989), or by

344 heat release by ortho-para-H2 conversion (Smith & Gierasch, 1995). These sources of energy operate in a

345 very different way are much less efficient thancompared to those available on Jupiter and Saturn. Firstly,,

346 and the high enrichment in volatiles methane in the Ice Giants could inhibit vertical motions due to a vertical

347 gradients of the atmospheric molecular weight (Guillot et al., 1995; Leconte et al., 2017). This phenomenon

348 may also be at work in the deep and inaccessible water clouds of both the Gas and Ice giants, but the

349 observable methane clouds of Uranus and Neptune provide an excellent opportunity to study it (Guillot et

350 al., 2019). Secondly, heat release by ortho-para H2 conversion is much more efficient in providing energy

351 within the cold atmospheres of Uranus and Neptune. Thus, convection may occur in vertically-thin layers

352 (Gierasch et al., 1987), rather than extending vertically over tremendous heights, or in a complex and

353 inhomogeneous weather layer (Hueso et al., 2020). Multi-wavelength remote sensing of the temperatures,

354 clouds, winds, and gaseous composition is required to investigate how these meteorological processes differ

15 Fletcher et al., Ice Giant Systems

355 from the Gas Giants, how they derive their energy from the internal heating or weak sunlight, and their

356 relation to the large-scale banded patterns and winds (Fletcher et al., 2020). Spatially-resolved reflectivity

357 and thermal-emission mapping will allow precise constraints on the Ice Giant energy balance to constrain

358 their self-luminosities. And mapping the distribution and depth of Ice Giant lightning, previously detected

359 via radio emissions on both worlds (Zarka et al., 1986; Gurnett et al., 1990), could determine the frequency

360 and intensity of water-powered convection on the Ice Giants, elucidating its impact on their atmospheric

361 dynamics.

362

363 What is the large-scale circulation of Ice Giant atmospheres, and how deep does it go? Atmospheric

364 circulation, driven by both internal energy and solar heating, controls the thermal structure, radiative energy

365 balance, condensate cloud and photochemical haze characteristics, and meteorology. Unlike Jupiter and

366 Saturn, observations from Voyager, space telescopes, and ground-based observatories have revealed mid-

367 latitude upwelling (where most of the vigorous storms and coolest temperatures are found) and sinking

368 motions at the equator and poles (e.g., Conrath et al., 1998, Fletcher et al., 2020). This is superimposed onto

369 polar depletions in several key cloud-forming volatiles: methane (from reflection spectroscopy, Sromovsky

370 et al., 2014; Karkoschka et al., 2011); hydrogen sulphide (from near-IR and microwave spectroscopy, de Pater

371 et al., 1991; Hofstadter & Butler, 2003;, Irwin et al., 2018); and potentially ammonia (from microwave

372 imaging). Do these contrasts imply circulation patterns extending to great depths (de Pater et al., 2014), or

373 are they restricted in vertically-thin layers (Gierasch et al., 1987)? Recent re-analysis of the gravity fields

374 measured by Voyager (Kaspi et al., 2013) suggests that zonal flows are restricted to the outermost ~1000 km

375 of their radii, indicating relatively shallow weather layers overlying the deep and mysterious water-rich

376 interiors. The circulation of the stratosphere is almost entirely unknown on both worlds, due to the challenge

377 of observing weak emissions from hydrocarbons in the mid-infrared (e.g., Orton et al., 2014; Roman et al.,

378 20192020). Either way, observations of Uranus and Neptune will have stark implications for atmospheric

379 circulation on intermediate-sized planets with strong chemical enrichments and latitudinal gradients.

380

16 Fletcher et al., Ice Giant Systems

381 How does atmospheric chemistry and haze formation respond to extreme variations in sunlight and vertical

382 mixing, and the influence of external material? Methane can be transported into the stratosphere, where

383 photolysis initiates rich chemical pathways to produce a plethora of hydrocarbons (Moses et al., 2018). The

384 sluggish mixing of Uranus indicates that its photochemistry occurs in a different physical regime (higher

385 pressures) than on any other world. Furthermore, oxygen compounds from external sources (from cometary

386 impacts, infalling dust, satellite and ring material, Feuchtgruber et al., 1997) will play different photochemical

387 roles on Uranus, where the methane homopause is lower, than on Neptune (Moses & Poppe, 2017). This

388 exogenic influence can further complicate inferences of planetary formation mechanisms from

389 measurements of bulk abundances (particularly for CO). This rich atmospheric chemistry will be substantially

390 different from that on the Gas Giants, due to the weaker sunlight, the colder temperatures (changing reaction

391 rates and condensation processes, Moses et al., 2018), and the unusual ion-neutral chemistry resulting from

392 the complex magnetic field tilt and auroral processes (Dobrijevic et al. 2020). Condensation of these chemical

393 products (and water ice) can form thin haze layers observed in high-phase imaging (Rages et al., 1991), which

394 may add to the radiative warming of the stratosphere, be modulated by vertically-propagating waves, and

395 could sediment downwards to serve as condensation nuclei or aerosol coatings in the troposphere.

396 Furthermore, Uranus’ axial tilt presents an extreme test of coupled chemical and transport models, given

397 that each pole can spend decades in the shroud of winter darkness. The strength of vertical mixing may vary

398 with location, and disequilibrium tracers can be used to assess where mixing is strongest (Fletcher et al.,

399 2020; Cavalie et al., 2020). Such tracers include such as CO (abundant in the tropospheres of Uranus and

400 Neptune, Cavalie et al., 2017), para-H2 (Conrath et al., 1998) and yet-to-be-detected phosphine (Moreno et

401 al., 2009), arsine, germane, silane, or even tropospheric hydrocarbons (ethane, acetylene), all of which could

402 be used to determine where mixing is most active). Fluorescence spectroscopy, infrared emissions and sub-

403 mm sounding will reveal the vertical, horizontal, and temporal variability of the chemical networks in the

404 unique low-temperature regimes of Uranus and Neptune.

405

17 Fletcher et al., Ice Giant Systems

406 What are the sources of energy responsible for heating the middle- and upper-atmospheres? Weak sunlight

407 alone cannot explain the high temperatures of the stratosphere (Li et al., 2018) and thermosphere (Herbert

408 et al., 1987), and this severe deficit is known as the energy crisis. Exploration of Uranus and Neptune may

409 provide a solution to this problem, potentially revealing how waves transport energy vertically from the

410 convective troposphere into the middle atmosphere, and how the currents induced by the asymmetric and

411 time-variable magnetic fields provide energy to the upper atmosphere via Joule heating. For example, the

+ 412 long-term cooling of Uranus’ thermosphere, observed via emission from H3 in the ionosphere (Melin et al.,

413 2019), appears to follow Uranus’ magnetic season, which may hint at the importance of particle precipitation

414 modulated by the magnetosphere in resolving the energy crisis. Solving this issue at Uranus or Neptune, via

415 wave observations and exploring magnetosphere-ionosphere-atmosphere coupling processes (e.g., via

416 aurora detected in the UV and infrared), will provide insights into the energetics of all planetary atmospheres.

417

418 How do planetary ionospheres enable the energy transfer that couples the atmosphere and

419 magnetosphere? In-situ radio occultations remain the only source for the vertical distribution of electron

420 density in the ionosphere (Majeed et al., 2004), a critical parameter for determining the strength of the

421 coupling between the atmosphere and the magnetosphere. The Voyager 2 occultations of both Uranus and

422 Neptune (Lindal et al, 1986, 1992) provided only two profiles for each planet, providing very poor constraints

423 on what drives the complex shape of the electron density profiles in the ionosphere, including the influx of

424 meteoritic material (Moses et al., 2017).

425

426 How do Ice Giant atmospheres change with time? In the decades since the Voyager encounters, Uranus has

427 displayed seasonal polar caps of reflective aerosols with changing winds (Sromovsky et al., 2015) and long-

428 term upper atmospheric changes (Melin et al., 2019); Neptune’s large dark anticyclones – and their

429 associated orographic clouds – have grown, drifted, and dissipated (Lebeau et al., 1998, Stratman et al., 2001,

430 Wong et al., 2018, Simon et al., 2019); and a warm south polar vortex developed and strengthened during

431 the Neptunian summer (Fletcher et al., 2014). What are the drivers for these atmospheric changes, and how

18 Fletcher et al., Ice Giant Systems

432 do they compare to the other planets? There have been suggestions that storm activity has occurred

433 episodically, potentially with a seasonal connection (de Pater et al., 2015; Sromovsky et al., 2015), but is this

434 simply driven by observational bias to their sunlit hemispheres? Mission scenarios for the early 2040s would

435 result in observations separated from those obtained by the Voyager 2 by 0.5 Uranian years and 0.25

436 Neptunian years. Orbital remote sensing over long time periods, sampling both summer and winter

437 hemispheres, could reveal the causes of atmospheric changes in a regime of extremely weak solar forcing, in

438 contrast to Jupiter and Saturn.

439

440 Summary: Investigations of dynamics, chemistry, cloud formation, atmospheric circulation, and energy

441 transport on Uranus and Neptune would sample a sizeable gap in our understanding of planetary

442 atmospheres, in an underexplored regime of weak seasonal sunlight, low temperatures, and extremes of

443 internal energy and vertical mixing.

444

445 2.3 Ice Giant Magnetospheres

446

447 What can we learn about astrophysical plasma processes by studying the highly-asymmetric, fast-rotating

448 Ice Giant magnetospheres? The off-centered, oblique and fast rotating planetary magnetic fields of Uranus

449 and Neptune give rise to magnetospheres that are governed by large scale asymmetries and rapidly evolving

450 configurations (e.g. Griton et al. 2018), with no other parallels in our Solar System. The solar wind that

451 embeds impinges upon these two magnetospheres attains Mach numbers significantly larger than those

452 found at Earth, Jupiter, and Saturn, adding further to their uniqueness (Masters et al. 2018). Magnetospheric

453 observations should thus be a high priority in the exploration of the Ice Giants because they extend the

454 sampling of the vast parameter space that controls the structure and evolution of magnetospheres, thus

455 allowing us to achieve a more universal understanding of how these systems work. Insights would also be

456 provided to astrophysical plasma processes on similar systems that are remote to us both in space and time.

457 Such may be the magnetospheres of exoplanets or even that of the Earth at times of geomagnetic field

19 Fletcher et al., Ice Giant Systems

458 reversals, when the higher order moments of the terrestrial magnetic field become significant, as currently

+ 459 seen at the Ice Giants’ (Merrill and Mcfadden, 1999). Evidence for H3 ionospheric temperature modulations

460 at Uranus due to charged particle precipitation (Melin et al. 2011; 2013) is one of many indications reminding

461 us how strong a coupling between a planet and its magnetosphere can be, and why the study of the latter

462 would be essential also for achieving a system-level understanding of the Ice Giants.

463

464 A synergy between close proximity, remote sensing, and in-situ magnetospheric measurements at the Ice

465 Giants would redefine the state-of-the-art, currently determined by the single Voyager-2 flyby

466 measurements, and limited Earth-based auroral observations. Key questions that would guide such

467 observations are listed below:

468

469 Is there an equilibrium state of the Ice Giant magnetospheres? Voyager-2 spent only a few planetary

470 rotations within Uranus’ and Neptune’s magnetopauses, such that it was challenging to establish a nominal

471 configuration of their magnetospheres, their constituent particle populations, supporting current systems.

472 Furthermore, it is unclear whether the observations of this dynamic system represent any kind of steady

473 state, as and whether these ever approach steady-stateongoing magnetospheric reconfigurations were

474 observed throughout each flyby, owing to the large dipole tilts at the Ice Giants and their ~16 and ~17-hour

475 rotation periods. The extent to which the two magnetospheres are modified by internal plasma sources is

476 also poorly constrained; Uranus’ magnetosphere for instance was observed to be devoid of any appreciable

477 cold plasma populations (McNutt et al., 1987), although its hydrogen corona is believed to be a major

478 contributor (Herbert et al. 1996). The presence of strong electron radiation belts (Mauk et al. 2010), seems

479 contradictory to the absence of a dense, seed population at plasma energies, or could hint an efficient local

480 acceleration process by intense wave-fields (Scarf et al. 1987). A strong proton radiation belt driven by

481 Galactic Cosmic Ray (GCR)-planet collisions may reside closer to Uranus or Neptune than Voyager-2 reached

482 (e.g. Stone et al. 1986), given that Earth and Saturn, which are known to sustain such belts, are exposed to a

483 considerably lower GCR influx than the Ice Giants (Buchvarova and Belinov, 2009). Ion composition and UV

20 Fletcher et al., Ice Giant Systems

484 aurora measurements hint that Triton could be a major source of plasma in Neptune’s magnetosphere

485 (Broadftoot et al., 1989, Richardson & McNutt, 1990), although questions remain as to the effects of coupling

486 between the magnetosphere and the moon’s atmosphere and ionosphere in establishing the plasma source

487 (Hoogeveen & Cloutier, 1996). The magnetotails of both planets are expected to have very different

488 structures to those seen at other magnetized planets (Figure 4), with strong helical magnetic field

489 components (Cowley, 2013; Griton and Pantellini 2020), that may lead to a similarly helical topology of

490 reconnection sites across the magnetospheric current sheet (Griton et al. 2018). Whether the overall

491 magnetospheric configuration is controlled more by current sheet reconnection or a viscous interaction

492 along the magnetopause (Masters et al. 2018) is also unknown. Finally, measuring average escape rates of

493 ionospheric plasma would offer further insights on whether planetary magnetic fields protect planetary

494 atmospheres from solar wind erosion (Wei et al. 2014, Gunell et al., 2018). For such dynamic systems, long-

495 term observations that average out rotational effects and transients (e.g. Selesnick, 1988) are essential for

496 achieving closure to all the aforementioned questions.

497

498 How do the Ice Giant magnetospheres evolve dynamically? The large tilts of the Ice Giant magnetic fields

499 relative to their planetary spin-axes hint that large-scale reconfigurations at diurnal time scales are

500 dominating short-term dynamics, a view supported by MHD simulations (Griton and Pantellini 2020; Griton

501 et al. 2018; Cao and Paty 2017; Mejnertsen et al. 2016). The rate of magnetic reconnection, for instance, is

502 predicted to vary strongly with rotation, and so is the rate of matter and energy transfer into the

503 magnetosphere and eventually upper atmosphere (Masters et al., 2014). Simulations do not capture how

504 such variations impact regions as the radiation belts, which would typically require a stable environment to

505 accumulate the observed, high fluxes of energetic particles (Mauk et al. 2010). A strong diurnal variability

506 may also affect the space weather conditions at the orbits of the Ice Giant moons, regulating the interactions

507 between the charged particle populations, their surfaces and exospheres (Plainaki et al., 2016; Arridge et al.,

508 2014) through processes like surface sputtering, ion pick up, and charge exchange, which may also feed the

509 magnetosphere with neutrals and low energy, heavy ion plasma (Lammer 1995). In the time-frame

21 Fletcher et al., Ice Giant Systems

510 considered here, the exploration of Neptune could provide an opportunity to study a “pole-on”

511 magnetosphere at certain rotational phases.

512

513 Additional sources of variability can be solar wind driven, such as Corotating Interaction Regions (CIRs) and

514 Interplanetary Coronal Mass Ejections (ICMEs) (Lamy et al. 2017). By the time they reach Uranus and

515 Neptune, ICMEs expand and coalesce and attain a quasi-steady radial width of ~2-3 AU (Wang and

516 Richardson, 2004) that could result in active magnetospheric episodes with week-long durations. With Triton

517 being a potential active source, Neptune’s magnetosphere may show variations at the moon’s orbital period

518 (Decker and Cheng 1994987).

519

520 On longer time scales, seasonal changes of the planetary field’s orientation are most important, especially at

521 Uranus, because of its spin axis which almost lies on the ecliptic. There may be additional implications for the

522 long-term variability of the magnetosphere if magnetic field measurements reveal a secular drift of the

523 planetary field at any of the Ice Giants, in addition to providing constraints on the magnetic fields’ origin and

524 the planets’ interiors (Section 2.1).

525

526 How can we probe the Ice Giant magnetospheres through their aurorae? Aurorae form a unique diagnostic

527 of magnetospheric processes by probing the spatial distribution of active magnetospheric regions and their

528 dynamics at various timescales. Auroral emissions of Uranus and Neptune were detected by Voyager 2 at

529 ultraviolet and radio wavelengths. The Uranian UV aurora has been occasionally re-detected by the Hubble

530 Space Telescope since then. At NIR wavelengths auroral signatures remain elusive (e.g. Melin et al., 2019).

531

532 UV aurora are collisionally-excited H Ly-α and H 2 band emissions from the upper atmosphere, driven by

533 precipitating energetic particles. The radical differences of the Uranian UV aurora observed across different

534 seasons were assigned to seasonal variations of the magnetosphere/solar wind interaction (Broadfoot et al.,

535 1986; Lamy et al., 2012, 2017; Cowley, 2013; Barthélémy et al., 2014). The tentative detection of Neptunian

22 Fletcher et al., Ice Giant Systems

Figure 4 Typical magnetic field configuration in a terrestrial-like, solar wind driven magnetosphere (top) where the magnetic and spin axis are almost aligned, compared to a Uranus-like magnetospheric configuration (bottom), when the magnetic axis is 90° away from the spin axis, during equinox. The helicoidal magnetotail structure develops due to the planet’s fast rotation.

536 UV aurora did not reveal any clear planet-satellite interaction (e.g. with Triton) (Broadfoot et al., 1989).

537 Repeated UV spectro-imaging observations are essential to probe the diversity of Ice Giant aurora, assess

538 their underlying driver and constrain magnetic field models (e.g. Herbert, 2009).

539

540 Uranus and Neptune produce powerful auroral radio emissions above the ionosphere most likely driven by

541 the Cyclotron Maser Instability as at Earth, Jupiter, and Saturn. The Uranian and Neptunian Kilometric

542 Radiation are very similar (Desch et al., 1991, Zarka et al., 1995). They include (i) bursty emissions (lasting for

543 <10min) reminiscent of those from other planetary magnetospheres. A yet-to-be-identified time-stationary

544 source of free energy, able to operate in strongly variable magnetospheres, is thought to drive (ii) smoother

545 emissions (lasting for hours) that are unique in our solar system (Farrell, 1992). Long-term remote and in-situ

546 radio measurements are crucial to understand the generation of all types of Ice Giant radio emissions, to

547 complete the baseline for the search of exoplanetary radio emissions. Long-term monitoring of auroral

548 emissions will also be essential to precisely determine the rotation periods of these worlds.

549

550 Summary: The Ice Giant magnetospheres comprise unique plasma physics laboratories, the study of which

551 would allow us to observe and put to test a variety of astrophysical plasma processes that cannot be resolved

23 Fletcher et al., Ice Giant Systems

552 under the conditions that prevail at the terrestrial planets and at the Gas Giants. The strong planet-

553 magnetosphere links that exist further attest to the exploration of the magnetospheres as a key ingredient

554 of the Ice Giant systems.

555

556 2.4 Ice Giant Satellites – Natural & Captured

557

558 What can a comparison of Uranus’ natural satellites, and the captured “Ocean World” Triton, and the myriad

559 ice-rich bodies in the Solar System, reveal about the drivers of active geology and potential habitability

560 throughout the outer Solar Systemon icy satellites? The satellite systems of Uranus and Neptune offer very

561 different insights into moon formation and evolution; they are microcosms of the larger formation and

562 evolution of planetary systems. The Neptunian system is dominated by the ‘cuckoo-like’ arrival of Triton (i.e.,

563 severely disrupting any primordial Neptunian satellite system), which would be the largest known dwarf

564 planet in the Kuiper belt if it were still only orbiting the Sun. Neptune’s remaining satellites may not be

565 primordial, given the degree of system disruption generated by Triton’s arrival. In contrast, Uranus’s satellite

566 system appears to be relatively undisturbed since its formation — a highly surprising situation given that

567 Uranus has the most severe axial tilt of any planet, implying a dramatic collisional event in its past. Thus,

568 these two satellite systems offer laboratories for understanding the key planetary processes of formation,

569 capture and collision.

570

571 What can the geological diversity of the large icy satellites of Uranus reveal about the formation and

572 continued evolution of primordial satellite systems? The five largest moons of Uranus (Miranda, Ariel,

573 Umbriel, Titania, Oberon) are comparable in sizes and orbital configurations to the medium-sized moons of

574 Saturn. However, they have higher mean densities, about 1.5 g/cm3 on average, and have different insolation

575 patterns: their poles are directed towards the Sun for decades at a time, due to the large axial tilt of Uranus.

576 The surfaces of Uranus’s five mid-sized moons exhibit extreme geologic diversity, demonstrating a complex

577 and varied history (Figure 1). On Ariel and Miranda, signs of endogenic resurfacing associated with tectonic

24 Fletcher et al., Ice Giant Systems

578 stress, and possible cryovolcanic processes, are apparent: these moons appear to have the youngest

579 surfaces. Geological interpretation has suffered greatly from the incomplete Voyager 2 image coverage of

580 only the southern hemisphere, and extremely limited coverage by Uranus-shine in part of the north (Stryk

581 and Stooke, 2008). Apart from a very limited set of images with good resolution at Miranda, revealing

582 fascinatingly complex tectonic history and possible re-formation of the moon (Figure 5), most images were

583 acquired at low to medium resolution, only allowing characterisation of the main geological units (e.g., Croft

584 and Soderblom, 1991) and strongly limiting any surface dating from the crater-size frequency distribution

585 (e.g. Plescia, 1987a, Plescia, 1987b). High-resolution images of these moons, combined with spectral data,

586 will reveal essential information on the tectonic and cryovolcanic processes and the relative ages of the

587 different geological units, via crater statistics and sputtering processes. Comparison with Saturn’s inner

588 moons system will allow us to identify key drivers in the formation and evolution of compact multiplanetary

589 systems.

590

591 What was the influence of tidal interaction and internal melting on shaping the Uranian worlds, and could

592 internal water oceans still exist? As in the Jovian and Saturnian systems, tidal interaction is likely to have

593 played a key role in the evolution of the Uranian satellite system. Intense tidal heating during sporadic

594 passages through resonances is expected to have induced internal melting in some of the icy moons

595 (Tittemore and Wisdom 1990). Such tidally-induced melting events, comparable to those expected on

596 Enceladus (e.g. Běhounková et al. 2012), may have triggered the geological activity that led to the late

597 resurfacing of Ariel and possibly transient hydrothermal activity. The two largest (>1500 km diameter)

598 moons, Titania and Oberon, may still harbour liquid water oceans between their outer ice shells and inner

599 rocky cores – remnants of past melting events. Comparative study of the static and time-variable gravity field

600 of the Uranian and Saturnian moons, once well-characterized, will constrain the likelihood and duration of

601 internal melting events, essential to characterize their astrobiological potential. Complete spacecraft

602 mapping of their surfaces could reveal recent endogenic activity.

603

25 Fletcher et al., Ice Giant Systems

604 Accurate radio tracking and astrometric measurements can also be used to quantify the influence of tidal

605 interactions in the system at present, providing fundamental constraints on the dissipation factor of Uranus

606 itself (Lainey, 2008). Gravity and magnetic measurements, combined with global shape data, will greatly

607 improve the models of the satellites' interiors, providing fundamental constraints on their bulk composition

608 (density) and evolution (mean moment of inertia). Understanding their ice-to-rock ratio and internal

609 structure will enable us to understand if Uranus' natural satellite system are the original population of bodies

610 that formed around the planet, or if they were disrupted.

611

612 What is the chemical composition of the surfaces of the Uranian moons? The albedos of the major Uranian

613 moons, considerably lower than those of Saturn’s moons (except Phoebe and Iapetus’s dark hemisphere),

614 reveal that their surfaces are characterized by a mixture of H2O ice and a visually dark and spectrally bland

615 material that is possibly carbonaceous in origin (Brown and Cruikshank, 1983). Pure CO2 ice is concentrated

616 on the trailing hemispheres of Ariel, Umbriel and Titania (Grundy et al., 2006), and it decreases in abundance

617 with increasing semimajor axis (Grundy et al., 2006; Cartwright et al., 2015), as opposed to what is observed

618 in the Saturnian system. At Uranus’ distance from the Sun, CO2 ice should be removed on timescales shorter

619 than the age of the Solar System, so the detected CO2 ice may be actively produced (Cartwright et al., 2015).

620

621 The pattern of spectrally red material on the major moons will reveal their interaction with dust from the

622 decaying orbits of the irregular satellites. Spectrally red material has been detected primarily on the leading

623 hemispheres of Titania and Oberon. H2O ice bands are stronger on the leading hemispheres of the classical

624 satellites, and the leading/trailing asymmetry in H2O ice band strengths decreases with distance from Uranus.

625 Spectral mapping of the distribution of red material and trends in H2O ice band strengths across the satellites

626 and rings can map out infalling dust from Uranus's inward-migrating irregular satellites (Cartwright et al.,

627 2018), similar to what is observed in the Saturnian system on Phoebe/Iapetus (e.g., Tosi et al., 2010), and

628 could reveal how coupling with the Uranus plasma and dust environment influence their surface evolution.

629

26 Fletcher et al., Ice Giant Systems

Figure 5 Best-resolution (roughly ~1 km/px) imagery of terrains seen in the moons of Uranus (top row) and Neptune (lower row) by Voyager 2. At Uranus, Umbriel and Titania are highly cratered with some major faults; Miranda displays spectacular and massive tectonic features; Ariel’s filled-fissured surface is suggestive of late cryovolcanic activity. At Neptune, Proteus has a surface suggestive of Saturn’s dust-rich moon Helene. The dwarf-planet-sized Triton has both the sublimation-related “cantaloupe terrain” (left) and active nitrogen gas geysers in the south polar terrains (right), with deposited dust visible as dark streaks. Credit: NASA/JPL-Caltech/Ted Stryk/collage M. Bannister. 630 Does Triton currently harbour a subsurface ocean and is there evidence for recent, or ongoing, active

631 exchange with its surface? The Neptune’s large moon Triton, one of a rare class of Solar System bodies with

632 a substantial atmosphere and active geology, offers a unique opportunity to study a body comparable to the

633 dwarf planets of the rest of the trans-Neptunian region, but much closer. Triton shares many similarities in

634 surface and atmosphere with Pluto (Grundy et al. 2016), and both may harbour current oceans. Triton’s

635 retrograde orbit indicates it was captured, causing substantial early heating. Triton, in comparison with

636 Enceladus and Europa, will let us understand the role of tidally-induced activity on the habitability of ice-

637 covered oceans. The major discovery of plumes emanating from the southern polar cap of Triton (Soderblom

638 et al. 1990; see Figure 5) by Voyager 2, the most distant activity in the Solar System, is yet to be fully

639 understood (Smith et al. 1989).

640

641 Like Europa, Triton has a relatively young surface age of ~100 Ma (Stern and McKinnon 2000), inferred from

642 its few visible impact craters. Triton also displays a variety of distinctive curvilinear ridges and associated

643 troughs, comparable to those on Europa (Prockter et al. 2005) and especially apparent in the “cantaloupe

27 Fletcher et al., Ice Giant Systems

644 terrain”. This suggests that tidal stresses and dissipation have played an essential role in Triton’s geological

645 activity, and may be ongoing (Nimmo and Spencer 2015). Intense tidal heating following its capture could

646 have easily melted its icy interior (McKinnon et al. 1995). Its young surface suggests that Triton experienced

647 an ocean crystallization stage relatively recently (Hussmann et al. 2006), associated with enhanced surface

648 heat flux (Martin-Herrero et al. 2018). Combined magnetic, gravimetric and shape measurements from

649 multiple flybys or in orbit will allow us to detect if Triton possesses an ocean and to constrain the present-

650 day thickness of the ice shell. Correlating the derived shell structure and geological units will show if there is

651 exchange with the ocean. It is entirely unclear if the source(s) for Triton’s plumes reaches the ocean, as at

652 Enceladus.

653

654 Are seasonal changes in Triton’s tenuous atmosphere linked to specific sources and sinks on the surface,

655 including its remarkable plume activity? Triton’s surface has a range of volatile ices seen in Earth-based

656 near-infrared spectra, including N2, H2O, CO2, and CH4 (Quirico et al., 1999; Cruikshank et al., 2000; Tegler et

657 al, 2012). A 2.239 µm absorption feature suggests that CO and N2 molecules are intimately mixed in the ice

658 rather than existing as separate regions of pure CO and pure N2 deposits (Tegler et al., 2019). Mapping the

659 spatial variation of this absorption feature will constrain how the surface-atmosphere interaction affects the

660 surface composition, and more generally its climate and geologic evolution. Triton’s surface may also contain

661 complex organics from atmospheric photochemistry, like those of Pluto or Saturn’s moon Titan (e.g.

662 Krasnopolsky and Cruikshank 1995), as suggested by its yellowish areas (Thompson and Sagan 1990).

663 Identifying the organic compounds, and mapping out their correlation with recently active terrains and

664 geysers, will strongly raise the astrobiological potential of this exotic icy world.

665

666 Triton’s tenuous atmosphere is mainly molecular nitrogen, with a trace of methane and CO near the surface

667 (Broadfoot et al. 1989, Lellouch et al. 2010). Despite Triton's distance from the Sun and its cold temperatures,

668 the weak sunlight is enough to drive strong seasonal changes on its surface and atmosphere. Because CO and

669N 2 are the most volatile species on Triton, they are expected to dominate seasonal volatile transport across

28 Fletcher et al., Ice Giant Systems

670 its surface. Observation of increased CH4 partial pressure between 1989 and 2010 (Lellouch et al. 2010)

671 confirmed that Triton’s atmosphere is seasonably variable. The plumes of nitrogen gas and dust could be a

672 seasonal solar-driven process like the CO2 `spiders’ of the south polar regions of Mars, although an endogenic

673 origin is possible.

674

675 Are the smaller satellites of Neptune primordial? Voyager 2’s flyby led to the discovery of six small moons

676 inside Triton’s orbit: Naiad, Thalassa, Despina, Galatea, Larissa and Proteus. A seventh inner moon,

677 Hippocamp, has been recently discovered by HST observations orbiting between the two largest, Larissa and

678 Proteus. Almost nothing is known about these faint moons, which may post-date Triton’s capture rather than

679 being primordial. Only a new space mission could unveil basic features such as shape and surface

680 composition, shedding light on their origin.

681

682 How does an Ice Giant satellite system interact with the planets’ magnetospheres? Most of the major

683 moons of Uranus and Neptune orbit within the planets’ extensive magnetospheres (Figure 1). The tilt and

684 offset of both planets’ magnetic dipoles compared to their rotation axes mean that, unlike at Saturn, the

685 major moons at both Ice Giants experience continually-changing external magnetic fields. The potential

686 subsurface oceans of Titania, Oberon and Triton would be detectable by a spacecraft that can monitor for an

687 induced magnetic field. The moons in both systems orbit in relatively benign radiation environments, but

688 radiation belt particles could still drive sputtering processes at the inner moons’ surfaces and Triton’s

689 atmosphere. Triton could be a significant potential source of a neutral gas torus and magnetospheric plasma

690 at Neptune. Triton's ionosphere's transonic, sub-Alfvenic regime (Neubauer 1990; Strobel et al. 1990) may

691 generate an auroral spot in Neptune's upper atmosphere. No such interaction is anticipated at Uranus. Red

692 aurorae may also be present on Triton from N2 emission, providing valuable insights into Triton’s interaction

693 with its space environment.

694

29 Fletcher et al., Ice Giant Systems

695 Summary: Exploring Uranus’ and Neptune’s natural satellites, as well as and Neptune’s captured moon

696 Triton, will reveal how ocean worlds may form and remain active, defining the extent of the habitable zone

697 in the Solar System. Both satellite systems are equally compelling for future orbital exploration.

698

699 2.5 Ice Giant Ring Systems

700

701 What processes shape the narrow, dusty rings of Ice Giants, and why do they differ from the extensive rings

702 of Saturn? Uranus and Neptune both possess a complex system of rings and satellites (Figure 6, see also de

703 Pater et al., 2018, Nicholson et al., 2018). The rings exhibit narrow and dense ringlets, as well as fainter but

704 broader dust components. The moons can confine the rings gravitationally, and may also serve as sources

705 and sinks for ring material. Observations indicate rapid variability in the Uranian and Neptunian rings within

706 decades. A mission to the Ice Giants, with a dedicated suite of instruments, can answer fundamental

707 questions on the formation and evolution of the ring systems and the planets themselves:

708

709 What is the origin of the solar system ring systems, and why are they so different? The origin of the giant

710 planets’ rings is one of the unsolved mysteries in planetary science. Whereas all four giant planets do have

711 rings, their diversity of structure and composition argues for different formation scenarios (Charnoz et al.,

712 2018). It was hypothesized that the very massive Saturnian rings formed more than 3 Gyrs ago through tidal

713 destruction of a moon, or of a body on a path traversing the system. However, recent Cassini measurements

714 (Zhang et al., 2017, Kempf et al., 2018, Waite et al., 2018, Iess et al., 2019) argue for a younger ring age. In

715 contrast, Uranus’ and Neptune’s rings are far less massive and they have a different structure. Compared to

716 Saturn, their rings’ albedo is much lower, favouring a parent body that was a mixture of ice and dark material

717 (silicates and possibly organics). For instance, it was suggested (Colwell and Esposito, 1992, 1993) that these

718 two ring systems could result from the periodic destruction of moonlets though meteoroid bombardment,

719 in which case most of the ringlets would be only transient structures, currently in the process of re-accretion

720 to satellites.because of re-accretion processes. Among the Uranian dust rings the µ ring has a distinct blue

30 Fletcher et al., Ice Giant Systems

721 spectral slope (de Pater et al. 2006). In that regard it is similar to Saturn’s E ring, for which the blue slope

722 results from the narrow size distribution of its grains, formed in the cryo-volcanic activity of the moon

723 Enceladus. Although the moon Mab is embedded in the µ ring (Showalter et al, 2006) it appears much too

724 small (12km radius) to be volcanically active and create in this way the dust that forms the ring. Other dust

725 rings of Uranus exhibit a red spectral slope (de Pater et al. 2006), suggestive of dusty material released in

726 micrometeoroid impacts on atmosphere-less moons and the origin for different appearance of the µ ring is

727 unknown. Recent ground-based observations of thermal emission from the Uranian ring system (Molter et

728 al., 2019) are consistent with the idea that the rings are made up of larger particles, without micron-sized

729 dust, and that their temperatures result from low thermal inertia and/or slow rotation of the particles.

730 Clearly, more data is needed to ultimately settle the question of the origin and nature of Uranus’ and

731 Neptune’s rings.

732

733 How do the ring-moon systems evolve? A variety of processes govern the evolution of planetary rings, many

734 of which are also fundamental to other cosmic disks. Important for the rings are viscous transport, ring-

735 satellite interactions, the self-gravity, as well as meteoroid bombardment and cometary impacts. These

736 processes may induce rapid evolution on timescales much shorter than the rings’ age. For instance, the

737 Neptune ring arcs were initially interpreted to be confined by resonances with the moon Galatea. However,

738 it was found that the arcs actually move away from the corresponding corotation sites (Dumas et al., 1999,

739 Namouni & Porco 2002) and that they evolve rapidly (de Pater et al., 2005). Also, for the Uranian rings

740 significant changes since the Voyager epoch are observed (de Pater et al., 2006, 2007). In the Uranus and

741 Neptune systems the role the moons play in sculpting the rings is even stronger than for Saturn’s rings.

742 Moreover, depending on their composition, the Uranus and Neptune ring systems may even extend beyond

743 the planet’s Roche Limit (Tiscareno et al., 2013). This implies that their path of evolution is different from the

744 Saturnian rings, inducing changes on more rapid timescales. Some of the edges of the Uranian rings are

745 clearly confined by known satellites and others might be confined by yet undetected moons. Alternatively,

746 there might be different processes of confinement at work, in a similar manner as for narrow rings of Saturn,

31 Fletcher et al., Ice Giant Systems

Figure 6 Panel (a): Schematic diagram of the ring moon systems of Uranus, Neptune, and Saturn for comparison. The Roche radius (for icy ring particles) is marked as a dashed line. Panel (b) shows a combination of two Voyager images of the Uranus rings, taken at low (upper part) and high (lower part) phase angle (figure from Nicholson et al., 2018). At high phase angle the dusty components of the ring system stand out (IMAGES: NASA/JPL). Panel (c) shows a Voyager image of the rings of Neptune (CreditIMAGES: NASA/JPL).

747 for which shepherding moons are absent. Some of the dense rings of Uranus show sub-structure the origin

748 of which is unclear. Spacecraft investigation will answer the question if the structure is induced by resonant

749 interaction with moons, or if it represents intrinsic modes arising from instability and self-gravity of the rings.

750

751 What is the ring composition? The rings (as the moons) very likely consist of material that was present at

752 the location in the solar system where the planets themselves have formed. Therefore, the investigation of

753 the composition of the rings, while being interesting in its own right, may also shed light on models of planet

754 formation and migration in the solar system.

755

756 Imaging at high and intermediate phase angles will constrain the shapes and properties of known dust rings

757 and has the potential to discover new rings. Multicolour imaging at a range of phase angles will determine

758 the size distribution of the grains that form these rings. Imaging at low phase angles will probe the dense

32 Fletcher et al., Ice Giant Systems

759 rings and allow for a comprehensive search and discovery of yet unseen satellites that serve as sources for

760 ring material and that interact dynamically with the rings. Stellar occultations performed with a high-speed

761 photometer, or radio occultations, will determine the precise optical depths of the denser rings and resolve

762 their fine sub-structure (French et al., 1991). An IR spectrometer will determine the composition of the dense

763 rings. In a complementary manner a dust detector will directly determine the composition of the grains

764 forming the low optical depth dust rings (Postberg et al., 2009) and of particles lifted from the dense rings

765 (Hsu et al., 2018). A dust detector will also measure the flux and composition of interplanetary particles in

766 the outer solar system, which is a poorly known quantity of high importance for the origin and evolution of

767 the rings.

768

769 Summary: Ice Giant rings appear to be fundamentally different to those of Saturn, such that their origin,

770 evolution, composition and gravitational relationships with the icy satellites should provide key new insights

771 into the forces shaping ring systems surrounding planetary bodies.

772 3. Ice Giant Science in Context

773

774 The scientific themes highlighted in Section 2 span multiple disciplines within planetary science, and address

775 questions that touch on issues across astronomy. In this section we review how an Ice Giant mission must

776 be considered in the context of other fields and technologies that will be developing in the coming decades.

777

778 3.1 Astronomical Observatories

779

780 An Ice Giant System mission would be operating in the context of world-leading new facilities on or near

781 Earth. The 2020s will see the launch of the James Webb Space Telescope, able to provide spectral maps of

782 both Ice Giants from 1-30 µm but at a moderate spatial resolution and with limited temporal coverage. Earth-

783 based observatories in the 8-10 m class provide better spatial resolution at the expense of telluric

784 obscuration. A successor to the Hubble Space Telescope, which has been the key provider of visible and UV 33 Fletcher et al., Ice Giant Systems

785 observations of the Ice Giants, has yet to become a reality, but could be operating in the 2030s. And although

786 the next generation of Earth-based observatories in the 30+m class (the ELT, TMT, GMT) will provide exquisite

787 spatial resolution, this will remain limited to atmospheric and ionospheric investigations of the Earth-facing

788 hemisphere (with some disc-averaged spectroscopy of the satellites), leaving a multitude of fundamental

789 questions unanswered.

790

791 Additionally, distant remote sensing can only study phenomena that alter the emergent spectrum –

792 meteorology, seasonal variations, and ionospheric emissions (auroral and non-auroral). This limits our

793 understanding of the underlying mechanisms and means that ground- and space-based telescopes only serve

794 a narrow subset of the Ice Giant community, and cannot address the wide-ranging goals described in Section

795 2. Thus, there is no substitute for orbital exploration of one or both of these worldsIce Giant systems

796 (alongside in situ sampling of their atmospheres), but we envisage that these space missions will work in

797 synergy with the ground-based astronomy community, following successful examples of Galileo, Cassini, Juno

798 and, ultimately, JUICE and Europa Clipper. Support from Earth-based observatories, either on the ground or

799 in space, will be used to establish a temporal baseline for atmospheric changes (e.g., tracking storms), provide

800 global context for close-in observations from the orbiters, and plug any gaps in spectral coverage or spectral

801 resolution in the orbiter payload.

802

803 3.2 Heliophysics Connection

804

805 Missions to explore the Ice Giant Systems also resonate with the heliophysics community, as detailed

806 exploration of an oblique rotator can inform a universal model of magnetospheres. The panel on solar-wind

807 magnetosphere interactions of the 2013 heliophysics decadal survey5 identified how the magnetospheres of

808 Uranus and Neptune are fundamentally different from others in our Solar System, and sought to ensure that

5 https://www.nap.edu/catalog/13060/solar-and-space-physics-a-science-for-a-technological-society

34 Fletcher et al., Ice Giant Systems

809 magnetic field instruments would be guaranteed a place on outer planet missions, with a strong

810 recommendation being that NASA’s heliophysics and planetary divisions partner on a Uranus orbital mission.

811 They describe how Uranus offers an example of solar wind/magnetosphere interactions under strongly

812 changing orientations over diurnal timescales. Depending on the season, the effects of solar wind-

813 magnetosphere interaction vary dramatically over the course of each day.

814

815 There is also a need to understand how the solar wind evolves beyond 10 AU (Saturn orbit), as the states of

816 solar structures travelling within the solar wind (solar wind pressure pulses) are largely unknown due to the

817 lack of observations at such large heliocentric distances (Witasse et al., 2017). The long cruise duration of a

818 mission to Uranus or Neptune provides an excellent opportunity for both heliophysics and Ice Giant

819 communities if a space weather-monitoring package that includes a magnetometer, a solar wind analyser,

820 and a radiation monitor operates continuously during the cruise phase, to understand how conditions vary

821 out to 20-30 AU over a prolonged lifetime. Moreover, the complexity of the interactions of Uranus’s

822 magnetosphere with the solar wind provides an ideal testbed for the theoretical understanding of planetary

823 interactions with the solar wind, significantly expanding the parameter range over which scientists can study

824 magnetospheric structure and dynamics. The potential discoveries from its dynamo generation and its

825 variability stand to open new chapters in comparative planetary magnetospheres and interiors.

826

827 3.3 Exoplanet & Brown Dwarf Connection

828

829 The Ice Giant System mission will occur during an explosion in our understanding of planets beyond our Solar

830 System, through ESA’s Cosmic Vision missions Plato, Euclid, and ARIEL; through missions with international

831 partners like JWST and TESS; and through next-generation observatories like WFIRST, LUVOIR, Origins, and

832 HabEx. The physical and chemical processes at work within our own Solar System serve as the key foundation

833 for our understanding of those distant and unresolved worlds. Our Solar System provides the only laboratory

834 in which we can perform in-situ experiments to understand exoplanet formation, composition, structure,

35 Fletcher et al., Ice Giant Systems

Figure 7 Known transiting exoplanets in 2016, from the Kepler mission, showing sub-Neptunes as the most common planetary radius in the current census. Credit: NASA/Ames

835 dynamos, systems and magnetospheres. After the several highly-successful missions to the Gas Giants (and

836 the upcoming ESA/JUICE mission), dedicated exploration of the Ice Giants would place those discoveries into

837 a broader, solar-system context.

838

839 Planet Statistics: Uranus and Neptune represent our closest and best examples of a class of planets

840 intermediate in mass and size between the larger, hydrogen-helium-enriched gas giants, and the rocky

841 terrestrial worlds. Indeed, Neptune- and sub-Neptune-size worlds have emerged as commonplace in our

842 ever-expanding census of exoplanets (Figure 7). Neptune-size planets are among the most common classes

843 of exoplanet in our galaxy (Fulton et al., 2018). Fressin et al. (2013) suggest that this category of planets can

844 be found around 3-31% of sun-like stars, and Petigura et al (2017) suggests that sub-Neptunes remains the

845 most common category within Kepler’s survey. Based on statistics from the Kepler mission it is predicted that

846 TESS will detect over 1500 sub-Neptunes (2-4 RE) over the mission (Barclay et al. 2018). Microlensing surveys

847 (e.g., WFIRST, Penny et al., 2019) will also be more sensitive to lower-mass planets on wide orbits, and could

848 reveal new insights into extrasolar Ice Giants ahead of a mission to Uranus or Neptune. Given these planetary

849 occurrence rates, the exploration of bulk composition and interiors of our Ice Giants would provide strong

850 constraints on the most common outcome for planetary formation. However, we emphasise that being

851 Neptune-sized does not necessarily imply being Neptune-like, as a plethora of additional parameters come

852 into play to shape the environmental conditions on these worlds.

853

36 Fletcher et al., Ice Giant Systems

854 Atmospheric Insights: Although we are currently unable to resolve spatial contrasts on exoplanets and brown

855 dwarfs, comparisons of dayside (eclipse) and nightside (transit) spectra suggest the presence of powerful

856 winds on some targets that are responsible for redistribution of energy. Brightness variations as Brown

857 Dwarfs rotate suggest patchy cloud structures, but their rapid evolution was only understood when

858 compared with long-cadence Neptune observations (Apai et al., 2013, Stauffer et al. 2016, Simon et al. 2016).

859 Changes in exoplanet transit spectra with temperature indicate complex cloud condensation processes

860 (Morley et al., 2012; Wakeford et al., 2017). Atmospheric dynamics, chemistry, and cloud formation all vary

861 as a function of planetary age, distance from the host star, and the bulk enrichments of chemical species.

862 The smaller radii of Uranus and Neptune, compared to the larger gas giants, implies atmospheric processes

863 (zonal banding, storms, vortices) at work in a different region of dynamical parameter space, and one which

864 is unavailable elsewhere in our Solar System. Detailed exploration of our Ice Giants, in comparison to the

865 existing studies of the Gas Giants, would allow us to unravel these competing and complex phenomena

866 shaping the atmospheres of exoplanets and brown dwarfs. Importantly, measurements of Ice Giant

867 composition and dynamics can be directly compared to exoplanet and brown dwarf observations, placing

868 their formation location, age, and evolution into context, and vice versa. However, this cannot be done

869 without a detailed comparative dataset from our Solar System Ice Giants.

870

871 Magnetospheric Insights: Rymer et al. (2018) explore the importance of Ice Giant interactions with the wider

872 magnetic environment as a testbed for understanding exoplanets. Eccentric and complex orbital

873 characteristics appear commonplace beyond our Solar System, and Uranus is one of the only places where

874 radio emission, magnetospheric transport and diffusion resulting from a complex magnetospheric

875 orientation can be explored. The stability and strength of the Uranian radiation belts could also guide the

876 search for exoplanets with radiation belts. Finally, understanding the dynamos of Uranus and Neptune would

877 drastically improve our predictions of magnetic field strengths and exoplanet dynamo morphologies. Each

878 of these insights will be vital as exoplanetary science moves into an era of characterisation of atmospheric

879 composition, dynamics, clouds, and auroral/radio emissions.

37 Fletcher et al., Ice Giant Systems

880 4. Ice Giant Missions

881 4.1 Architectures: The Case for Orbiters

882

883 The scientific themes of Ice Giant System missions (summarised in Figure 9) are broad and challenging to

884 capture within a single mission architecture, but recent efforts by both ESA (e.g., the 2018-19 studies with

885 the Concurrent Design Facility for an M*-class mission6) and NASA (e.g., the 2017 Science Study Team report,

886 Hofstadter et al., 2019)7 have explored strategies to achieve many of the goals in Section 2. The joint NASA-

887 ESA science study team provided a detailed investigation of various combinations of flyby missions, orbiters,

888 multiple sub-satellites from a core spacecraft, satellite landers, and atmospheric probes. Strategies to

889 explore both Ice Giants with dual spacecraft have also been proposed (Turrini et al., 2014; Simon et al., 2018).

890 It was widely recognised that a flyby mission like Voyager, without any additional components like an entry

891 probe, was deemed to provide the lowest science return for the Ice Giants themselves, despite their lower

892 cost point. Without in situ measurements, and by providing only brief snapshots of the evolving atmospheres

893 and magnetospheres, and limited coverage of the satellites and rings, a flyby could not deliver on the highest-

894 priority science goals for an Ice Giant mission. Targeting Triton as a flyby, or the inclusion of an entry probe,

895 would increase the scientific reach, but would still prove inadequate for whole-system science. The study

896 found that an Ice Giant orbiter, to for either the Uranus or Neptune systems, alongside an in situ atmospheric

897 probe, would provide an unprecedented leap in our understanding of these enigmatic worlds. An orbiter

898 would maximise the time spent in the system to conduct science of interest to the entire planetary

899 community.

900

901 In our 2019 submission to ESA’s call for ideas to shape the planning of space science missions in the coming

902 decades (known as Voyage 2050), we therefore proposed that an orbital mission to an Ice Giant should be

6 http://sci.esa.int/future-missions-department/61307-cdf-study-report-ice-giants/ 7 https://www.lpi.usra.edu/icegiants/

38 Fletcher et al., Ice Giant Systems

903 considered as a cornerstone of ESA’s Voyage 2050 programme, if not already initiated with our international

904 partners in the coming decade. An ESA orbital mission, powered by radioisotope thermoelectric generators,

905 should be studied as an L-class mission to capitalise on the wealth of European experience of the Cassini and

906 JUICE missions. Alternatively, an M-class Ice Giant budget would allow a crucial contribution to an orbital

907 mission led by our international partners. The mass and mission requirements associated with additional

908 components, such as satellite landers, in situ probes, or secondary small satellites, must be tensioned against

909 the capabilities of the core payload, and the capability of the launch vehicle and propulsion. In all of these

910 cases, a formal study of the requirements and capabilities is necessary to mature the concept.

911

912 Payload Considerations: The 2017 NASA study found that payload masses of 90-150 kg could deliver

913 significant scientific return for a flagship-class mission, whilst the 2018 ESA CDF study identified 100 kg as a

914 realistic payload mass for a European orbiter. Different studies have resulted in different prioritisations for

915 instrumentation, but produced suites of orbiter experiments in common categories. Multi-spectral remote

916 sensing is required, using both imaging and spectroscopy, spanning the UV, visible, near-IR (e.g.,

917 atmosphere/surface reflectivity, dynamics; auroral observations), mid-IR, sub-mm, to centimetre

918 wavelengths (e.g., thermal emission and energy balance, atmospheric circulation). Such remote sensing is

919 also a requirement for characterising any atmospheric probe entry sites, or satellite lander sites. Direct

920 sensing of the magnetospheric and plasma environment would be accomplished via magnetometers, dust

921 detectors, plasma instruments, radio wave detectors, and potentially mass spectrometers. Radio science

922 would provide opportunities for interior sounding and neutral/ionospheric occultation studies. The provision

923 of such instruments would capitalise on European heritage on Cassini, JUICE, Rosetta, Venus/Mars Express,

924 and Bepi Colombo, but at the same time recognising the need to develop smaller, lighter, and less

925 power/data-intensive instruments, raising and maturing their technological readiness.

926

927 Orbit Considerations: Orbital missions to both Uranus and Neptune depend crucially on the chemical fuel

928 required for orbit insertion, which determines the deliverable mass. The potential use of aerocapture, using

39 Fletcher et al., Ice Giant Systems

929 atmospheric drag to slow down the spacecraft, permits larger payloads and faster trip times at the expense

930 of increased risk, which needs significant further study. Mission requirements and orbital geometries will

931 determine the inclination of orbital insertion – high geographical latitudes would benefit some atmospheric,

932 rings, and magnetospheric science, but satellite gravity assists and subsequent trajectory corrections would

933 be needed for exploration of the satellites, rings and atmosphere from a low-inclination orbit. High

934 inclinations are easier to achieve at Uranus, although Triton can be used to drive a satellite tour at Neptune.

935 The delivery and telemetry for an atmospheric entry probe must also be considered in an orbital tour design

936 (e.g., Simon et al., 2020). We also propose that distinct phases of an orbital tour be considered, balancing

937 moderate orbital distances (for remote sensing, outer magnetosphere science, and a satellite tour) with

938 close-in final orbits (for gravity science and inner magnetosphere), following the example of Cassini and Juno.

939 Multiple close flybys of major satellites are desirable to map their interiors, surfaces, and atmospheres via a

940 variety of techniques. Finally, multi-year orbital tours (at least ~3 years) would maximise our time in the

941 system, permitting the study of atmospheric and magnetospheric changes over longer time periods. The

942 2018 ESA CDF study confirmed the feasibility of orbital tours satisfying these scientific requirements.

943

944 Ring Hazards: The 2017 NASA-ESA report highlighted potentially unknown ring-plane hazards as a topic for

945 future study. Orbit insertion should be as close to the planet as possible to reduce the required fuel, but the

946 properties of Ice Giant rings remain poorly constrained. Potential options to mitigate this risk include: having

947 the insertion be further out (requiring more fuel); fly through the ring plane at an altitude where atmospheric

948 drag is high enough to reduce the number of particles, but not enough to adversely affect the spacecraft; use

949 a pathfinder spacecraft to measure the particle density ahead of time; or use Earth-based observations to

950 constrain the upper atmosphere/ring hazard. Detailed calculations on the location of this safe zone are

951 required.

952

953 4.2 Timeliness and Launch Opportunities

954

40 Fletcher et al., Ice Giant Systems

955 Trajectories to reach the Ice Giants depend on a number of factors: the use of chemical and/or solar-electric

956 propulsion (SEP) technologies; the lift capacity of the launch vehicle; the use of aerocapture/aerobraking;

957 and the need for gravity assists. The availability of Jupiter, as the largest planet, is key to optimal launch

958 trajectories, and the early 2030s offer the best opportunities. The synodic periods of Uranus and Neptune

959 with respect to Jupiter are ~13.8 and ~12.8 years, respectively, meaning that optimal Jupiter gravity-assist

960 (GA) windows occur every 13-14 years. The NASA-ESA joint study team identified chemical-propulsion

961 opportunities with a Jupiter GA in 2029-30 for Neptune, and a wider window of 2030-34 for Uranus. Such

962 windows would repeat in the 2040s, and a wider trade space (including the potential to use Saturn GA or

963 direct trajectories using the next generation of launch vehicles) should be explored. We stress that a mission

964 to an Ice Giant is feasible using conventional chemical propulsion.

965

966 Furthermore, a mission launched to Uranus or Neptune could offer significant opportunities for flybys of

967 Solar System objects en route, especially Centaurs, which are small bodies that orbit in the giant planet

968 region. This population has yet to be explored by spacecraft, but represents an important evolutionary step

969 between Kuiper Belt objects and comets. Around 300 are currently known, with 10% of these observed to

970 show cometary activity. In addition, we can expect to discover at least an order of magnitude more Centaurs

971 by the 2030s, following discoveries by the Large Synoptic Survey Telescope, increasing the probability that a

972 suitable flyby target can be found near to the trajectory to Uranus/Neptune. Some of the largest of the

973 Centaurs (~100 km scale objects) have their own ring systems, the origin of which has yet to be explained,

974 while smaller ones (1-10 km scale) could add an important ‘pre-activity’ view to better interpret data from

975 Rosetta’s exploration of a comet. The payload options described in Section 4 would be well suited to

976 characterise a Centaur during a flyby.

977

978 The launch time necessarily influences the arrival time, as depicted in Figure 8. Uranus will reach northern

979 summer solstice in 2030, and northern autumnal equinox in 2050. Voyager 2 observed near northern winter

980 solstice, meaning that the north poles of the planet and satellites were shrouded in darkness. These

41 Fletcher et al., Ice Giant Systems

981 completely unexplored northern terrains will begin to disappear into darkness again in 2050, where they will

982 remain hidden for the following 42 years (half a Uranian year).

983

984 Neptune passed northern winter solstice in 2005, and will reach northern spring equinox in 2046. After this

985 time, the southern hemispheres of the planet and satellites (most notably the plumes of Triton at high

986 southern latitudes) will sink into winter darkness, meaning that the Triton plumes – if they are indeed

987 restricted to the south – would no longer be in sunlight after ~2046, and would remain hidden for the next

988 ~82 years (half a Neptunian year).

Figure 8 Potential timeline of missions to Uranus (left) and Neptune (right), compared to the seasons on each Ice Giant. The white arrows show the approximate timescales for launch opportunities in the early 2030s, with arrival in the 2040s. 989

990 The 2028-34 launch opportunities were assessed by the joint ESA-NASA study team. Saturn GA was

991 considered but did not appear optimal for this launch window. Interplanetary flight times are 6 to 12 years

42 Fletcher et al., Ice Giant Systems

992 to Uranus, 8 to 13 years to Neptune, depending on launch year, mission architecture, and launch vehicle.

993 The greater challenge of reaching the Neptune system was reflected in their choice of detailed architecture

994 studies: five missions to Uranus (orbiters with/without probes; with/without SEP; and with different payload

995 masses), and a single orbiter and probe for Neptune. Both Uranus and Neptune were deemed equally

996 valuable as targets – Uranus standing out for its uniqueness; Neptune for the prospects of exploring Triton.

997 The choice between these two enticing destinations will ultimately be driven by launch opportunities and

998 deliverable mass to the systems.

999

1000 We must capitalise on the current momentum within Europe, alongside our international partners, to make

1001 use of the launch opportunities in the ~2030s. Such a mission would arrive at Uranus while we can still see

1002 the totally-unexplored northern terrains, or at Neptune while we can still see the active geology of Triton.

1003 Operations in the 2040s would also allow ESA to maintain Outer Solar System expertise from current and

1004 future missions like JUICE. An Ice Giant explorer would therefore be active as a cornerstone of ESA’s Voyage

1005 2050 strategic planning for space missions.

1006

1007 4.3 Mature and Developing Technologies

1008

1009 An ambitious mission to an Ice Giant System would largely build on existing mature technologies (e.g., see

1010 the discussion of payload development in Section 4.1), but several challenges have been identified that, if

1011 overcome, would optimise and enhance our first dedicated orbital mission to these worlds. Note that we

1012 omit the need for ablative materials on atmospheric entry probes, which will be required for in situ science

1013 (Mousis et al., 2018, Simon et al., 2020). Key areas where technology maturations are required include:

1014

1015 Space nuclear power: With the prospect of flying solar-powered spacecraft to 20 AU being non-viable, an

1016 Ice Giant System mission must rely on radioisotope power sources, both for electricity and for spacecraft

1017 heating. In the US, existing MMRTGs (multi-mission radioisotope thermal generators), based on the decay of

43 Fletcher et al., Ice Giant Systems

1018 238Pu, will be re-designed to create eMMRTGs (“enhanced”) to increase the available specific power at the

1019 end of life, 4-5 of which were considered for the mission architectures studied in the 2017 NASA-ESA report.

1020 Previous M-class Uranus mission proposals have relied on US provision of these power sources for an ESA-

1021 led mission. However, ESA continues to pursue the development of independent power sources based on

1022 241Am (Ambrosi et al., 2019). Whilst the power density is lower than that of 238Pu, the half-life is much longer,

1023 and much of the material is available from the reprocessing of spent fuel from European nuclear reactors,

1024 extracted chemically from plutonium to a ceramic oxide form. Prototypes for both radioisotope heater units

1025 (warming the spacecraft) and thermoelectric generators (providing spacecraft power) have now been

1026 demonstrated, and development is continuing for operational use late in the next decade (Ambrosi et al.,

1027 201920). An Ice Giant System mission could benefit tremendously from this independent European power

1028 source.

1029

1030 Hardware longevity: Given the 6- to 13-year interplanetary transfer, coupled with the desire for a long

1031 orbital tour, Ice Giant orbiters must be designed to last for a long duration under a variable thermal load

1032 imposed by gravity assists from the inner to the outer solar system. This poses constraints on the reliability

1033 of parts and power sources, as well as the need to develop optimised operational plans for the long cruise

1034 phases, such as the use of hibernation modes following the example of New Horizons.

1035

1036 Telemetry/Communications: All missions to the giant planets are somewhat constrained by competition

1037 between advanced instrumentation and the reduced data rates at large distances from Earth, but the case

1038 at Uranus and Neptune is most severe. The use of Ka-band in the downlink, currently supported by both the

1039 US Deep Space Network and two out of three ESA Deep Space Antennas, allows for mitigating this issue by

1040 increasing the achievable daily data volumes, but a careful optimization of the science tour remains critical.

1041 The available power, length of the downlink window, and the need to balance science data, engineering and

1042 housekeeping telemetry, all contribute to determining the overall data rate. For example, the NASA/JPL

44 Fletcher et al., Ice Giant Systems

1043 mission study8 in 2017 suggested that use of a 35-W traveling wave tube amplifier (the current power limit

1044 for space-qualified TWTAs) and a 34-m ground-station could provide 15 kbps at Uranus, whereas this could

1045 increase to 30 kbps with a 70-W TWTA. The 2018 ESA CDF study9 suggested that the data rate of a future M-

1046 class orbital mission to Uranus could use a 100-W TWTA to deliver 94 kbps Ka-band downlink from Uranus,

1047 or 42 kbps from Neptune. Assuming 3.2 hours/day for communications, this equates to daily data volumes

1048 for science of 1.09 and 0.48 Gb, respectively. Using current 35-W TWTAs, the ESA CDF study also suggested

1049 that lower data rates of 31 and 14 kbps could be achievable at Uranus and Neptune respectively, requiring

1050 the use of longer downlink windows. Even under the most optimistic assumptions discussed above, it is clear

1051 that some data optimization strategy would still be required. We would welcome detailed studies of new

1052 communications technologies, such as optical communications, as a general enabling technology for solar

1053 system exploration. However, we recognise that achieving the required directionality of a downlink laser

1054 from beyond 5AU will be challenging.

1055

1056

1057 Electromagnetic cleanliness: A number of the proposed payload elements (magnetometers, plasma and radio

1058 instruments) impose strict EMC requirements on the spacecraft, and new technologies and cost optimisation

1059 strategies should be explored to reduce these EMC issues.

1060

1061 Launch Vehicles: The market for launch vehicles is changing dramatically both in Europe and in the US. Ice

1062 Giant mission concepts have traditionally considered Jupiter gravity assists to provide realistic flight times

1063 and sufficient payload delivered to each system. However, we also advocate investigation of the direct-

1064 transfer trajectories enabled by the heavy-lift capacity of the next generation of launch vehicles. This which

1065 may (i) obviate the need for Jupiter/Saturn GA, so allow more flexibility in launch dates; and (ii) open up the

1066 possibilities for launching multiple spacecraft that share the same faring.

8 https://www.lpi.usra.edu/icegiants/mission_study/Full-Report.pdf 9 https://sci.esa.int/web/future-missions-department/-/61307-cdf-study-report-ice-giants

45 Fletcher et al., Ice Giant Systems

1067

1068 , but an Ice Giant System cornerstone mission could take advantage of the latest technologies, delivering as

1069 much mass to the system as possible within the constraints of a realistic flight time. Larger launch vehicles

1070 may also open up the prospect for multiple spacecraft to share the same faring.

1071 5. Summary and Perspectives

1072

1073 This article reviews and updates the scientific rationale for a mission to an Ice Giant system, advocating that

1074 an orbital mission (alongside an atmospheric entry probe, Mousis et al., 2018) be considered as a cornerstone

1075 of ESA’s Voyage 2050 programme, working in collaboration with international partners to launch the first

1076 dedicated mission to either Uranus or Neptune. Using technologies both mature and in development, the

1077 Ice Giants community hopes to capitalise on launch opportunities in the 2030s to reach the Ice Giants. As

1078 shown in Figure 9, an Ice Giant System mission would engage a wide community, drawing expertise from a

1079 vast range of disciplines within planetary science, from surface geology to planetary interiors; from

1080 meteorology to ionospheric physics; from plasma scientists to heliophysicists. But this challenge is also

1081 interdisciplinary in nature, engaging those studying potentially similar Neptune-size objects beyond our Solar

1082 System, by revealing the properties of this underexplored class of planetary objects. As Neptune’s orbit

1083 shapes the dynamical properties of objects in the distant solar system, an Ice Giant System mission also draws

1084 in the small-bodies community investigating objects throughout the Outer Solar System, from Centaurs, to

1085 TNOs, to Kuiper Belt Objects, and contrasting these with the natural satellites of Uranus.

1086

1087 To launch a mission to Uranus and/or Neptune in the early 2030s, the international Ice Giant community

1088 needs to significantly raise the maturity of the mission concepts and instrument technological readiness. At

1089 the time of writing (late 2019), we await the outcomes of both ESA’s Voyage 2050 process, and the next US

1090 planetary decadal survey. Nevertheless, this should not delay the start of a formal science definition and

1091 study process (pre-phase A conceptual studies), so that we are ready to capitalise on any opportunities that

1092 these strategic planning surveys provide. Fully costed and technologically robust mission concepts need to 46 Fletcher et al., Ice Giant Systems

1093 be developed, studied, and ready for implementation by 2023-25 to have the potential to meet the upcoming

1094 window for Jupiter gravity assists between 2029-2034. We hope that the scientific themes and rationale

1095 identified in this review will help to guide that conceptual study process, to develop a paradigm-shifting

1096 mission that will help redefine planetary science for a generation of scientists and engineers.

1097

1098 Most importantly, the Ice Giant System mission will continue the breath-taking legacy of discovery of the

1099 Voyager, Galileo, Cassini, and Juno and JUICE missions to the giant planets (in addition to the forthcoming

1100 JUICE and Europa Clipper missions). A dedicated orbiter of an Ice Giant is the next logical step in our

1101 exploration of the Solar System, completing humankind’s first reconnaissance of the eight planets. It will be

1102 those discoveries that no one expected, those mysteries that we did not anticipate, and those views that no

1103 human has previously witnessed, which will enthuse the general public, and inspire the next generation of

1104 explorers to look to the worlds of our Solar System. We urge both ESA and NASA to take up this challenge.

1105

1106

47 Fletcher et al., Ice Giant Systems

Figure 9 Left: Our last views of Uranus and Neptune from a robotic spacecraft, taken by Voyager 2 three decades ago (Credit: NASA/JPL/E. Lakdawalla). Will we see these views again before a half-century has elapsed? Right: Seven science themes for Voyage 2050 that could be addressed by an Ice Giant System mission.

1107

1108 Acknowledgements

1109

1110 Fletcher was supported by a Royal Society Research Fellowship and European Research Council Consolidator

1111 Grant (under the European Union's Horizon 2020 research and innovation programme, grant agreement No

1112 723890) at the University of Leicester. Mousis acknowledges support from CNES. Hueso was supported by

1113 the Spanish MINECO project AYA2015-65041-P (MINECO/FEDER, UE) and by Grupos Gobierno Vasco IT1366-

1114 19 from Gobierno Vasco. UK authors acknowledge support from the Science and Technology Facilities

1115 Council (STFC). The authors are grateful to Beatriz Sanchez-Cano, Colin Snodgrass, Hannah Wakeford,

1116 Marius Millot, and Radek Poleski for their comments on early drafts of this article, and to two anonymous

1117 reviewers for their helpful suggestions. We are grateful to the members of ESA’s Voyage 2050 Senior

1118 Committee for allowing us to present this review at the Voyage 2050 workshop in Madrid in October 2019

1119 (see https://www.cosmos.esa.int/web/voyage-2050/workshop-programme).

48 Fletcher et al., Ice Giant Systems

1120

1121 References

1122 Ambrosi, R. M., Williams, H., Watkinson, E. J., et al. (2019), European Radioisotope Thermoelectric

1123 Generators (RTGs) and Radioisotope Heater Units (RHUs) for Space Science and Exploration, Space Science

1124 Reviews, 215, 55, 10.1007/s11214-019-0623-9.

1125

1126 Apai, D., Radigan, J., Buenzli, E., et al. (2013), HST Spectral Mapping of L/T Transition Brown Dwarfs Reveals

1127 Cloud Thickness Variations, The Astrophysical Journal, 768, 121, 10.1088/0004-637X/768/2/121.

1128

1129 Arridge, C. S., Achilleos, N., Agarwal, J., et al. (2014), The science case for an orbital mission to Uranus:

1130 Exploring the origins and evolution of ice giant planets, Planetary and Space Science, 104, 122,

1131 10.1016/j.pss.2014.08.009.

1132

1133 Arridge, C. S., Agnor, C. B., André, N., et al. (2012), Uranus Pathfinder: exploring the origins and evolution of

1134 Ice Giant planets, Experimental Astronomy, 33, 753, 10.1007/s10686-011-9251-4.

1135

1136 Atreya, S. K., Hofstadter, M. H., In, J. H., et al. (2020), Deep Atmosphere Composition, Structure, Origin, and

1137 Exploration, with Particular Focus on Critical in situ Science at the Icy Giants, Space Science Reviews, 216, 18,

1138 10.1007/s11214-020-0640-8.

1139

1140 Bailey, E., & Stevenson, D. J. (2015), Modeling Ice Giant Interiors Using Constraints on the H₂2-

1141H₂2O Critical Curve, AGU Fall Meeting Abstracts, 2015, P31G-03,..

1142

1143 Bali, E., Audétat, A., & Keppler, H. (2013), Water and hydrogen are immiscible in Earth's mantle, Nature, 495,

1144 220, 10.1038/nature11908.

1145 49 Fletcher et al., Ice Giant Systems

1146 Barclay, T., Pepper, J., & Quintana, E. V. (2018), A Revised Exoplanet Yield from the Transiting Exoplanet

1147 Survey Satellite (TESS), The Astrophysical Journal Supplement Series, 239, 2, 10.3847/1538-4365/aae3e9.

1148

1149 Barthélemy, M., Lamy, L., Menager, H., et al. (2014), Dayglow and auroral emissions of Uranus in

1150H₂2 FUV bands, Icarus, 239, 160, 10.1016/j.icarus.2014.05.035.

1151

1152 Batalha, N. M., Rowe, J. F., Bryson, S. T., et al. (2013), Planetary Candidates Observed by Kepler. III. Analysis

1153 of the First 16 Months of Data, The Astrophysical Journal Supplement Series, 204, 24, 10.1088/0067-

1154 0049/204/2/24.

1155

1156 Běhounková, M., Tobie, G., Choblet, G., et al. (2012), Tidally-induced melting events as the origin of south-

1157 pole activity on Enceladus, Icarus, 219, 655, 10.1016/j.icarus.2012.03.024.

1158

1159 Boué, G., & Laskar, J. (2010), A Collisionless Scenario for Uranus Tilting, The Astrophysical Journal, 712, L44,

1160 10.1088/2041-8205/712/1/L44.

1161

1162 Broadfoot, A. L., Herbert, F., Holberg, J. B., et al. (1986), Ultraviolet Spectrometer Observations of Uranus,

1163 Science, 233, 74, 10.1126/science.233.4759.74.

1164

1165 Broadfoot, A. L., Atreya, S. K., Bertaux, J. L., et al. (1989), Ultraviolet Spectrometer Observations of Neptune

1166 and Triton, Science, 246, 1459, 10.1126/science.246.4936.1459.

1167

1168 Brown, R. H., & Cruikshank, D. P. (1983), The Uranian satellites: Surface compositions and opposition

1169 brightness surges, Icarus, 55, 83, 10.1016/0019-1035(83)90052-0.

1170

50 Fletcher et al., Ice Giant Systems

1171 Buchvarova, M., & Velinov, P. (2009), Cosmic ray spectra in planetary atmospheres, Universal Heliophysical

1172 Processes, 257, 471, 10.1017/S1743921309029718.

1173

1174 Běhounková, M., Tobie, G., Choblet, G., et al. (2012), Tidally-induced melting events as the origin of south-

1175 pole activity on Enceladus, Icarus, 219, 655, 10.1016/j.icarus.2012.03.024.

1176

1177 Cao, X., & Paty, C. (2017), Diurnal and seasonal variability of Uranus's magnetosphere, Journal of Geophysical

1178 Research (Space Physics), 122, 6318, 10.1002/2017JA024063.

1179

1180 Cartwright, R. J., Emery, J. P., Rivkin, A. S., et al. (2015), Distribution of CO₂2 ice on the large

1181 moons of Uranus and evidence for compositional stratification of their near-surfaces, Icarus, 257, 428,

1182 10.1016/j.icarus.2015.05.020.

1183

1184 Cartwright, R. J., Emery, J. P., Pinilla-Alonso, N., et al. (2018), Red material on the large moons of Uranus:

1185 Dust from the irregular satellites?, Icarus, 314, 210, 10.1016/j.icarus.2018.06.004.

1186

1187 Cavalié, T., Venot, O., Selsis, F., et al. (2017), Thermochemistry and vertical mixing in the tropospheres of

1188 Uranus and Neptune: How convection inhibition can affect the derivation of deep oxygen abundances, Icarus,

1189 291, 1, 10.1016/j.icarus.2017.03.015.

1190

1191 Cavalié, T., Venot, O., Miguel, Y., et al. (2020), The deep composition of Uranus and Neptune from in situ

1192 exploration and thermochemical modeling, arXiv e-prints, arXiv:2004.13987,..Space Science Reviews,

1193 accepted, 10.1007/s11214-020-00677-8.

1194

1195 Charnoz, S., Canup, R. M., Crida, A., et al. (2018), The Origin of Planetary Ring Systems, Planetary Ring

1196 Systems. Properties, Structure, and Evolution, 517, 10.1017/9781316286791.018.

51 Fletcher et al., Ice Giant Systems

1197

1198 Colwell, J. E., & Esposito, L. W. (1993), Origins of the rings of Uranus and Neptune. 2. Initial conditions and

1199 ring moon populations, Journal of Geophysical Research, 98, 7387, 10.1029/93JE00329.

1200

1201 Colwell, J. E., & Esposito, L. W. (1992), Origins of the rings of Uranus and Neptune 1. Statistics of satellite

1202 disruptions, Journal of Geophysical Research, 97, 10227, 10.1029/92JE00788.

1203

1204 Conrath, B. J., Gierasch, P. J., & Ustinov, E. A. (1998), Thermal Structure and Para Hydrogen Fraction on the

1205 Outer Planets from Voyager IRIS Measurements, Icarus, 135, 501, 10.1006/icar.1998.6000.

1206

1207 Cowley, S. W. H. (2013), Response of Uranus' auroras to solar wind compressions at equinox, Journal of

1208 Geophysical Research (Space Physics), 118, 2897, 10.1002/jgra.50323.

1209

1210 Croft, S. K., & Soderblom, L. A. (1991), Geology of the Uranian satellites., Uranus, 561,..

1211

1212 Cruikshank, D. P., Schmitt, B., Roush, T. L., et al. (2000), Water Ice on Triton, Icarus, 147, 309,

1213 10.1006/icar.2000.6451.

1214

1215 de Pater, I., Romani, P. N., & Atreya, S. K. (1991), Possible microwave absorption by H ₂2S gas

1216 in Uranus' and Neptune's atmospheres, Icarus, 91, 220, 10.1016/0019-1035(91)90020-T.

1217

1218 de Pater, I., Hammel, H. B., Showalter, M. R., et al. (2007), The Dark Side of the Rings of Uranus, Science, 317,

1219 1888, 10.1126/science.1148103.

1220

1221 de Pater, I., Sromovsky, L. A., Fry, P. M., et al. (2015), Record-breaking storm activity on Uranus in 2014,

1222 Icarus, 252, 121, 10.1016/j.icarus.2014.12.037.

52 Fletcher et al., Ice Giant Systems

1223

1224 de Pater, I., Gibbard, S. G., & Hammel, H. B. (2006), Evolution of the dusty rings of Uranus, Icarus, 180, 186,

1225 10.1016/j.icarus.2005.08.011.

1226

1227 de Pater, I., Gibbard, S. G., Chiang, E., et al. (2005), The dynamic neptunian ring arcs: evidence for a gradual

1228 disappearance of Liberté and resonant jump of courage, Icarus, 174, 263, 10.1016/j.icarus.2004.10.020.

1229

1230 de Pater, I., Fletcher, L. N., Luszcz-Cook, S., et al. (2014), Neptune’s global circulation deduced from multi-

1231 wavelength observations, Icarus, 237, 211, 10.1016/j.icarus.2014.02.030.

1232

1233 de Pater, I., Renner, S., Showalter, M. R., et al. (2018), The Rings of Neptune, Planetary Ring Systems.

1234 Properties, Structure, and Evolution, 112, 10.1017/9781316286791.005.

1235

1236 Decker, R. B., & Cheng, A. F. (1994), A model of Triton's role in Neptune's magnetosphere, Journal of

1237 Geophysical Research, 99, 19027, 10.1029/94JE01867.

1238

1239 Desch, M. D., Kaiser, M. L., Zarka, P., et al. (1991), Uranus as a radio source., Uranus, 894,..

1240

1241 Dobrijevic, M., Loison, J. C., Hue, V., et al. (2020), 1D photochemical model of the ionosphere and the

1242 stratosphere of Neptune, Icarus, 335, 113375, 10.1016/j.icarus.2019.07.009.

1243

1244 Dodson-Robinson, S. E., & Bodenheimer, P. (2010), The formation of Uranus and Neptune in solid-rich feeding

1245 zones: Connecting chemistry and dynamics, Icarus, 207, 491, 10.1016/j.icarus.2009.11.021.

1246

1247 Dumas, C., Terrile, R. J., Smith, B. A., et al. (1999), Stability of Neptune's ring arcs in question, Nature, 400,

1248 733, 10.1038/23414.

53 Fletcher et al., Ice Giant Systems

1249

1250 Farrell, W. M. (1992), Nonthermal radio emissions from Uranus, Planetary Radio Emissions III, 241,..

1251

1252 Feuchtgruber, H., Lellouch, E., de Graauw, T., et al. (1997), External supply of oxygen to the atmospheres of

1253 the giant planets, Nature, 389, 159, 10.1038/38236.

1254

1255 Fletcher, L. N., de Pater, I., Orton, G. S., et al. (2020), Ice Giant Circulation Patterns: Implications for

1256 Atmospheric Probes, Space Science Reviews, 216, 21, 10.1007/s11214-020-00646-1.

1257

1258 Fletcher, L. N., de Pater, I., Orton, G. S., et al. (2014), Neptune at summer solstice: Zonal mean temperatures

1259 from ground-based observations, 2003-2007, Icarus, 231, 146, 10.1016/j.icarus.2013.11.035.

1260

1261 French, R. G., Nicholson, P. D., Porco, C. C., et al. (1991), Dynamics and structure of the Uranian rings., Uranus,

1262 327,..

1263

1264 Fressin, F., Torres, G., Charbonneau, D., et al. (2013), The False Positive Rate of Kepler and the Occurrence of

1265 Planets, The Astrophysical Journal, 766, 81, 10.1088/0004-637X/766/2/81.

1266

1267 Fulton, B. J., & Petigura, E. A. (2018), The California-Kepler Survey. VII. Precise Planet Radii Leveraging Gaia

1268 DR2 Reveal the Stellar Mass Dependence of the Planet Radius Gap, The Astronomical Journal, 156, 264,

1269 10.3847/1538-3881/aae828.

1270

1271 Gierasch, P. J., & Conrath, B. J. (1987), Vertical temperature gradients on Uranus: Implications for layered

1272 convection, Journal of Geophysical Research, 92, 15019, 10.1029/JA092iA13p15019.

1273

54 Fletcher et al., Ice Giant Systems

1274 Griton, L., Pantellini, F., & Meliani, Z. (2018), Three-Dimensional Magnetohydrodynamic Simulations of the

1275 Solar Wind Interaction With a Hyperfast-Rotating Uranus, Journal of Geophysical Research (Space Physics),

1276 123, 5394, 10.1029/2018JA025331.

1277

1278 Griton, L., & Pantellini, F. (2020), Magnetohydrodynamic simulations of a Uranus-at-equinox type rotating

1279 magnetosphere, Astronomy and Astrophysics, 633, A87, 10.1051/0004-6361/201936604.

1280

1281 Grundy, W. M., Binzel, R. P., Buratti, B. J., et al. (2016), Surface compositions across Pluto and Charon,

1282 Science, 351, aad9189, 10.1126/science.aad9189.

1283

1284 Grundy, W. M., Young, L. A., Spencer, J. R., et al. (2006), Distributions of H ₂2O and CO

1285 ₂2 ices on Ariel, Umbriel, Titania, and Oberon from IRTF/SpeX observations, Icarus, 184, 543,

1286 10.1016/j.icarus.2006.04.016.

1287

1288 Guillot, T. (1995), Condensation of Methane, Ammonia, and Water and the Inhibition of Convection in Giant

1289 Planets, Science, 269, 1697, 10.1126/science.7569896.

1290

1291 Guillot, T. (2019), Uranus and Neptune are key to understand planets with hydrogen atmospheres, arXiv e-

1292 prints, arXiv:1908.02092,..

1293

1294 Gunell, H., Maggiolo, R., Nilsson, H., et al. (2018), Why an intrinsic magnetic field does not protect a planet

1295 against atmospheric escape, Astronomy and Astrophysics, 614, L3, 10.1051/0004-6361/201832934.

1296

1297 Gurnett, D. A., Kurth, W. S., Cairns, I. H., et al. (1990), Whistlers in Neptune's magnetosphere: Evidence of

1298 atmospheric lightning, Journal of Geophysical Research, 95, 20967, 10.1029/JA095iA12p20967.

1299

55 Fletcher et al., Ice Giant Systems

1300 Helled, R., Bodenheimer, P., Podolak, M., et al. (2014), Giant Planet Formation, Evolution, and Internal

1301 Structure, Protostars and Planets VI, 643, 10.2458/azu_uapress_9780816531240-ch028.

1302

1303 Helled, R., & Bodenheimer, P. (2014), The Formation of Uranus and Neptune: Challenges and Implications for

1304 Intermediate-mass Exoplanets, The Astrophysical Journal, 789, 69, 10.1088/0004-637X/789/1/69.

1305

1306 Helled, R., Anderson, J. D., Podolak, M., et al. (2011), Interior Models of Uranus and Neptune, The

1307 Astrophysical Journal, 726, 15, 10.1088/0004-637X/726/1/15.

1308

1309 Helled, R., Anderson, J. D., & Schubert, G. (2010), Uranus and Neptune: Shape and rotation, Icarus, 210, 446,

1310 10.1016/j.icarus.2010.06.037.

1311

1312 Herbert, F. (2009), Aurora and magnetic field of Uranus, Journal of Geophysical Research (Space Physics),

1313 114, A11206, 10.1029/2009JA014394.

1314

1315 Herbert, F., Sandel, B. R., Yelle, R. V., et al. (1987), The upper atmosphere of Uranus: EUV occultations

1316 observed by Voyager 2, Journal of Geophysical Research, 92, 15093, 10.1029/JA092iA13p15093.

1317

1318 Herbert, F., & Hall, D. T. (1996), Atomic hydrogen corona of Uranus, Journal of Geophysical Research, 101,

1319 10877, 10.1029/96JA00427.

1320

1321 Hofstadter, M., Simon, A., Atreya, S., et al. (2019), Uranus and Neptune missions: A study in advance of the

1322 next Planetary Science Decadal Survey, Planetary and Space Science, 177, 104680,

1323 10.1016/j.pss.2019.06.004.

1324

56 Fletcher et al., Ice Giant Systems

1325 Hofstadter, M. D., & Butler, B. J. (2003), Seasonal change in the deep atmosphere of Uranus, Icarus, 165, 168,

1326 10.1016/S0019-1035(03)00174-X.

1327

1328 Hoogeveen, G. W., & Cloutier, P. A. (1996), The Triton-Neptune plasma interaction, Journal of Geophysical

1329 Research, 101, 19, 10.1029/95JA02761.

1330

1331 Hsu, H.-W., Schmidt, J., Kempf, S., et al. (2018), In situ collection of dust grains falling from Saturn's rings into

1332 its atmosphere, Science, 362, aat3185, 10.1126/science.aat3185.

1333

1334 Hueso, R., & Sánchez-Lavega, A. (2019), Atmospheric Dynamics and Vertical Structure of Uranus and

1335 Neptune's Weather Layers, Space Science Reviews, 215, 52, 10.1007/s11214-019-0618-6.

1336

1337 Hueso, R., Guillot, T., Sánchez-Lavega, A. (2020). Atmospheric dynamics and convective regimes in the non-

1338 homogeneous weather layers of the ice giants, Philosophical Transactions A. Submitted.

1339

1340 Hussmann, H., Sohl, F., & Spohn, T. (2006), Subsurface oceans and deep interiors of medium-sized outer

1341 planet satellites and large trans-neptunian objects, Icarus, 185, 258, 10.1016/j.icarus.2006.06.005.

1342

1343 Iess, L., Militzer, B., Kaspi, Y., et al. (2019), Measurement and implications of Saturn's gravity field and ring

1344 mass, Science, 364, aat2965, 10.1126/science.aat2965.

1345

1346 Irwin, P. G. J., Toledo, D., Garland, R., et al. (2018), Detection of hydrogen sulfide above the clouds in Uranus's

1347 atmosphere, Nature Astronomy, 2, 420, 10.1038/s41550-018-0432-1.

1348

1349 Jacobson, R. A. (2014), The Orbits of the Uranian Satellites and Rings, the Gravity Field of the Uranian System,

1350 and the Orientation of the Pole of Uranus, The Astronomical Journal, 148, 76, 10.1088/0004-6256/148/5/76.

57 Fletcher et al., Ice Giant Systems

1351

1352 Jacobson, R. A. (2009), The Orbits of the Neptunian Satellites and the Orientation of the Pole of Neptune,

1353 The Astronomical Journal, 137, 4322, 10.1088/0004-6256/137/5/4322.

1354

1355 Karkoschka, E., & Tomasko, M. G. (2011), The haze and methane distributions on Neptune from HST-STIS

1356 spectroscopy, Icarus, 211, 780, 10.1016/j.icarus.2010.08.013.

1357

1358 Kaspi, Y., Showman, A. P., Hubbard, W. B., et al. (2013), Atmospheric confinement of jet streams on Uranus

1359 and Neptune, Nature, 497, 344, 10.1038/nature12131.

1360

1361 Kegerreis, J. A., Eke, V. R., Gonnet, P., et al. (2019), Planetary giant impacts: convergence of high-resolution

1362 simulations using efficient spherical initial conditions and SWIFT, Monthly Notices of the Royal Astronomical

1363 Society, 487, 5029, 10.1093/mnras/stz1606.

1364

1365 Kempf, S., Altobelli, N., Srama, R., et al. (2018), The Age of Saturn's Rings Constrained by the Meteoroid Flux

1366 Into the System, EGU General Assembly Conference Abstracts, 10791,..

1367

1368 Krasnopolsky, V. A., & Cruikshank, D. P. (1995), Photochemistry of Triton's atmosphere and ionosphere.,

1369 Journal of Geophysical Research, 100, 21,271,286,..

1370

1371 Lainey, V. (2008), A new dynamical model for the Uranian satellites, Planetary and Space Science, 56, 1766,

1372 10.1016/j.pss.2008.02.015.

1373

1374 Lambrechts, M., Johansen, A., & Morbidelli, A. (2014), Separating gas-giant and ice-giant planets by halting

1375 pebble accretion, Astronomy and Astrophysics, 572, A35, 10.1051/0004-6361/201423814.

1376

58 Fletcher et al., Ice Giant Systems

1377 Lammer, H. (1995), Mass loss of N ₂2 molecules from Triton by magnetospheric plasma

1378 interaction, Planetary and Space Science, 43, 845, 10.1016/0032-0633(94)00214-C.

1379

1380 Lamy, L., Prangé, R., Hansen, K. C., et al. (2017), The aurorae of Uranus past equinox, Journal of Geophysical

1381 Research (Space Physics), 122, 3997, 10.1002/2017JA023918.

1382

1383 Lamy, L., Prangé, R., Hansen, K. C., et al. (2012), Earth-based detection of Uranus' aurorae, Geophysical

1384 Research Letters, 39, L07105, 10.1029/2012GL051312.

1385

1386 LeBeau, R. P., & Dowling, T. E. (1998), EPIC Simulations of Time-Dependent, Three-Dimensional Vortices with

1387 Application to Neptune's Great Dark SPOT, Icarus, 132, 239, 10.1006/icar.1998.5918.

1388

1389 Leconte, J., Selsis, F., Hersant, F., et al. (2017), Condensation-inhibited convection in hydrogen-rich

1390 atmospheres . Stability against double-diffusive processes and thermal profiles for Jupiter, Saturn, Uranus,

1391 and Neptune, Astronomy and Astrophysics, 598, A98, 10.1051/0004-6361/201629140.

1392

1393 Lellouch, E., de Bergh, C., Sicardy, B., et al. (2010), Detection of CO in Triton's atmosphere and the nature of

1394 surface-atmosphere interactions, Astronomy and Astrophysics, 512, L8, 10.1051/0004-6361/201014339.

1395

1396 Li, C., Le, T., Zhang, X., et al. (2018), A high-performance atmospheric radiation package: With applications to

1397 the radiative energy budgets of giant planets, Journal of Quantitative Spectroscopy and Radiative Transfer,

1398 217, 353, 10.1016/j.jqsrt.2018.06.002.

1399

1400 Lindal, G. F., Lyons, J. R., Sweetnam, D. N., et al. (1987), The atmosphere of Uranus: Results of radio

1401 occultation measurements with Voyager 2, Journal of Geophysical Research, 92, 14987,

1402 10.1029/JA092iA13p14987.

59 Fletcher et al., Ice Giant Systems

1403

1404 Lindal, G. F. (1992), The Atmosphere of Neptune: an Analysis of Radio Occultation Data Acquired with

1405 Voyager 2, The Astronomical Journal, 103, 967, 10.1086/116119.

1406

1407 Majeed, T., Waite, J. H., Bougher, S. W., et al. (2004), The ionospheres-thermospheres of the giant planets,

1408 Advances in Space Research, 33, 197, 10.1016/j.asr.2003.05.009.

1409

1410 Martin-Herrero, A., Romeo, I., & Ruiz, J. (2018), Heat flow in Triton: Implications for heat sources powering

1411 recent geologic activity, Planetary and Space Science, 160, 19, 10.1016/j.pss.2018.03.010.

1412

1413 Masters, A. (2014), Magnetic reconnection at Uranus' magnetopause, Journal of Geophysical Research

1414 (Space Physics), 119, 5520, 10.1002/2014JA020077.

1415

1416 Masters, A. (2018), A More Viscous-Like Solar Wind Interaction With All the Giant Planets, Geophysical

1417 Research Letters, 45, 7320, 10.1029/2018GL078416.

1418

1419 Masters, A., Achilleos, N., Agnor, C. B., et al. (2014), Neptune and Triton: Essential pieces of the Solar System

1420 puzzle, Planetary and Space Science, 104, 108, 10.1016/j.pss.2014.05.008.

1421

1422 Mauk, B. H., & Fox, N. J. (2010), Electron radiation belts of the solar system, Journal of Geophysical Research

1423 (Space Physics), 115, A12220, 10.1029/2010JA015660.

1424

1425 McKinnon, W. B., & Leith, A. C. (1995), Gas drag and the orbital evolution of a captured Triton., Icarus, 118,

1426 392, 10.1006/icar.1995.1199.

1427

60 Fletcher et al., Ice Giant Systems

1428 McKinnon, W. B., Stern, S. A., Weaver, H. A., et al. (2017), Origin of the Pluto-Charon system: Constraints

1429 from the New Horizons flyby, Icarus, 287, 2, 10.1016/j.icarus.2016.11.019.

1430

1431 McNutt, R. L., Selesnick, R. S., & Richardson, J. D. (1987), Low-energy plasma observations in the

1432 magnetosphere of Uranus, Journal of Geophysical Research, 92, 4399, 10.1029/JA092iA05p04399.

1433

1434 Mejnertsen, L., Eastwood, J. P., Chittenden, J. P., et al. (2016), Global MHD simulations of Neptune's

1435 magnetosphere, Journal of Geophysical Research (Space Physics), 121, 7497, 10.1002/2015JA022272.

1436

1437 Melin, H., Stallard, T., Miller, S., et al. (2011), Seasonal Variability in the Ionosphere of Uranus, The

1438 Astrophysical Journal, 729, 134, 10.1088/0004-637X/729/2/134.

1439

1440 Melin, H., Fletcher, L. N., Stallard, T. S., et al. (2019), The H₃3+⁺ ionosphere of

1441 Uranus: decades-long cooling and local-time morphology, Philosophical Transactions of the Royal Society of

1442 London Series A, 377, 20180408, 10.1098/rsta.2018.0408.

1443

1444 Melin, H., Stallard, T. S., Miller, S., et al. (2013), Post-equinoctial observations of the ionosphere of Uranus,

1445 Icarus, 223, 741, 10.1016/j.icarus.2013.01.012.

1446

1447 Merrill, R. T., & McFadden, P. L. (1999), Geomagnetic polarity transitions, Reviews of Geophysics, 37, 201,

1448 10.1029/1998RG900004.

1449

1450 Millot, M., Coppari, F., Rygg, J. R., et al. (2019), Nanosecond X-ray diffraction of shock-compressed superionic

1451 water ice, Nature, 569, 251, 10.1038/s41586-019-1114-6.

1452

61 Fletcher et al., Ice Giant Systems

1453 Molter, E. M., de Pater, I., Roman, M. T., et al. (2019), Thermal Emission from the Uranian Ring System, The

1454 Astronomical Journal, 158, 47, 10.3847/1538-3881/ab258c.

1455

1456 Moreno, R., Marten, A., & Lellouch, E. (2009), Search for PH3 in the Atmospheres of Uranus and Neptune at

1457 Millimeter Wavelength, AAS/Division for Planetary Sciences Meeting Abstracts #41, 28.02,..

1458

1459 Morley, C. V., Fortney, J. J., Marley, M. S., et al. (2012), Neglected Clouds in T and Y Dwarf Atmospheres, The

1460 Astrophysical Journal, 756, 172, 10.1088/0004-637X/756/2/172.

1461

1462 Moses, J. I., & Poppe, A. R. (2017), Dust ablation on the giant planets: Consequences for stratospheric

1463 photochemistry, Icarus, 297, 33, 10.1016/j.icarus.2017.06.002.

1464

1465 Moses, J. I., Fletcher, L. N., Greathouse, T. K., et al. (2018), Seasonal stratospheric photochemistry on Uranus

1466 and Neptune, Icarus, 307, 124, 10.1016/j.icarus.2018.02.004.

1467

1468 Mousis, O., Atkinson, D. H., Cavalié, T., et al. (2018), Scientific rationale for Uranus and Neptune in situ

1469 explorations, Planetary and Space Science, 155, 12, 10.1016/j.pss.2017.10.005.

1470

1471 Namouni, F., & Porco, C. (2002), The confinement of Neptune's ring arcs by the moon Galatea, Nature, 417,

1472 45, 10.1038/417045a.

1473

1474 Nettelmann, N., Helled, R., Fortney, J. J., et al. (2013), New indication for a dichotomy in the interior structure

1475 of Uranus and Neptune from the application of modified shape and rotation data, Planetary and Space

1476 Science, 77, 143, 10.1016/j.pss.2012.06.019.

1477

62 Fletcher et al., Ice Giant Systems

1478 Neubauer, F. M. (1990), Satellite plasma interactions, Advances in Space Research, 10, 25, 10.1016/0273-

1479 1177(90)90083-C.

1480

1481 Nicholson, P. D., De Pater, I., French, R. G., et al. (2018), The Rings of Uranus, Planetary Ring Systems.

1482 Properties, Structure, and Evolution, 93, 10.1017/9781316286791.004.

1483

1484 Nimmo, F., & Spencer, J. R. (2015), Powering Triton's recent geological activity by obliquity tides: Implications

1485 for Pluto geology, Icarus, 246, 2, 10.1016/j.icarus.2014.01.044.

1486

1487 Orton, G. S., Fletcher, L. N., Moses, J. I., et al. (2014), Mid-infrared spectroscopy of Uranus from the Spitzer

1488 Infrared Spectrometer: 1. Determination of the mean temperature structure of the upper troposphere and

1489 stratosphere, Icarus, 243, 494, 10.1016/j.icarus.2014.07.010.

1490

1491 Pearl, J. C., Conrath, B. J., Hanel, R. A., et al. (1990), The albedo, effective temperature, and energy balance

1492 of Uranus, as determined from Voyager IRIS data, Icarus, 84, 12, 10.1016/0019-1035(90)90155-3.

1493

1494 Pearl, J. C., & Conrath, B. J. (1991), The albedo, effective temperature, and energy balance of Neptune, as

1495 determined from Voyager data, Journal of Geophysical Research, 96, 18921, 10.1029/91JA01087.

1496

1497 Penny, M. T., Gaudi, B. S., Kerins, E., et al. (2019), Predictions of the WFIRST Microlensing Survey. I. Bound

1498 Planet Detection Rates, The Astrophysical Journal Supplement Series, 241, 3, 10.3847/1538-4365/aafb69.

1499

1500 Petigura, E. A., Howard, A. W., Marcy, G. W., et al. (2017), The California-Kepler Survey. I. High-resolution

1501 Spectroscopy of 1305 Stars Hosting Kepler Transiting Planets, The Astronomical Journal, 154, 107,

1502 10.3847/1538-3881/aa80de.

1503

63 Fletcher et al., Ice Giant Systems

1504 Plainaki, C., Lilensten, J., Radioti, A., et al. (2016), Planetary space weather: scientific aspects and future

1505 perspectives, Journal of Space Weather and Space Climate, 6, A31, 10.1051/swsc/2016024.

1506

1507 Plescia, J. B. (1987a), Cratering history of the Uranian satellites: Umbriel, Titania, and Oberon, Journal of

1508 Geophysical Research, 92, 14918, 10.1029/JA092iA13p14918.

1509

1510 Plescia, J. B. (1987b), Geological terrains and crater frequencies on Ariel, Nature, 327, 201,

1511 10.1038/327201a0.

1512

1513 Podolak, M., & Helled, R. (2012), What Do We Really Know about Uranus and Neptune?, The Astrophysical

1514 Journal, 759, L32, 10.1088/2041-8205/759/2/L32.

1515

1516 Podolak, M., Weizman, A., & Marley, M. (1995), Comparative models of Uranus and Neptune, Planetary and

1517 Space Science, 43, 1517, 10.1016/0032-0633(95)00061-5.

1518

1519 Pollack, J. B., Hubickyj, O., Bodenheimer, P., et al. (1996), Formation of the Giant Planets by Concurrent

1520 Accretion of Solids and Gas, Icarus, 124, 62, 10.1006/icar.1996.0190.

1521

1522 Postberg, F., Kempf, S., Schmidt, J., et al. (2009), Sodium salts in E-ring ice grains from an ocean below the

1523 surface of Enceladus, Nature, 459, 1098, 10.1038/nature08046.

1524

1525 Prockter, L. M., Nimmo, F., & Pappalardo, R. T. (2005), A shear heating origin for ridges on Triton, Geophysical

1526 Research Letters, 32, L14202, 10.1029/2005GL022832.

1527

1528 Quirico, E., Douté, S., Schmitt, B., et al. (1999), Composition, Physical State, and Distribution of Ices at the

1529 Surface of Triton, Icarus, 139, 159, 10.1006/icar.1999.6111.

64 Fletcher et al., Ice Giant Systems

1530

1531 Rages, K., Pollack, J. B., Tomasko, M. G., et al. (1991), Properties of scatterers in the troposphere and lower

1532 stratosphere of Uranus based on Voyager imaging data, Icarus, 89, 359, 10.1016/0019-1035(91)90183-T.

1533

1534 Redmer, R., Mattsson, T. R., Nettelmann, N., et al. (2011), The phase diagram of water and the magnetic

1535 fields of Uranus and Neptune, Icarus, 211, 798, 10.1016/j.icarus.2010.08.008.

1536

1537 Reinhardt, C., Chau, A., Stadel, J., et al. (2020), Bifurcation in the history of Uranus and Neptune: the role of

1538 giant impacts, Monthly Notices of the Royal Astronomical Society, 492, 5336, 10.1093/mnras/stz3271.

1539

1540 Richardson, J. D., & McNutt, R. L. (1990), Low-energy plasma in Neptune's magnetosphere, Geophysical

1541 Research Letters, 17, 1689, 10.1029/GL017i010p01689.

1542

1543 Roman, M. T., Fletcher, L. N., Orton, G. S., et al. (2020), Uranus in Northern Midspring: Persistent Atmospheric

1544 Temperatures and Circulations Inferred from Thermal Imaging, The Astronomical Journal, 159, 45,

1545 10.3847/1538-3881/ab5dc7.

1546

1547 Rymer, A., Mandt, K., Hurley, D., et al. (2019), Solar System Ice Giants: Exoplanets in our Backyard., Bulletin

1548 of the American Astronomical Society, 51, 176,..

1549

1550 Safronov, V. S. (1966), Sizes of the largest bodies falling onto the planets during their formation, Soviet

1551 Astronomy, 9, 987,..

1552

1553 Sánchez-Lavega, A., Sromovsky, L. A., et al. (2018), Gas Giants. In: Galperin, B., Read, P.L., (eds) Zonal Jets

1554 Phenomenology, Genesis and Physics. Cambridge (doi: 10.1017/9781107358225).

1555

65 Fletcher et al., Ice Giant Systems

1556 Scarf, F. L., Gurnett, D. A., Kurth, W. S., et al. (1987), Voyager 2 plasma wave observations at Uranus, Advances

1557 in Space Research, 7, 253, 10.1016/0273-1177(87)90226-2.

1558

1559 Selesnick, R. S. (1988), Magnetospheric convection in the nondipolar magnetic field of Uranus, Journal of

1560 Geophysical Research, 93, 9607, 10.1029/JA093iA09p09607.

1561

1562 Showalter, M. R., & Lissauer, J. J. (2006), The Second Ring-Moon System of Uranus: Discovery and Dynamics,

1563 Science, 311, 973, 10.1126/science.1122882.

1564

1565 Simon, A. A., Fletcher, L. N., Arridge, C., et al. (2020), A Review of the in Situ Probe Designs from Recent Ice

1566 Giant Mission Concept Studies, Space Science Reviews, 216, 17, 10.1007/s11214-020-0639-1.

1567

1568 Simon, A. A., Wong, M. H., & Hsu, A. I. (2019), Formation of a New Great Dark Spot on Neptune in 2018,

1569 Geophysical Research Letters, 46, 3108, 10.1029/2019GL081961.

1570

1571 Simon, A. A., Stern, S. A., & Hofstadter, M. (2018), Outer Solar System Exploration: A Compelling and Unified

1572 Dual Mission Decadal Strategy for Exploring Uranus, Neptune, Triton, Dwarf Planets, and Small KBOs and

1573 Centaurs, arXiv e-prints, arXiv:1807.08769,.

1574

1575 Simon, A. A., Rowe, J. F., Gaulme, P., et al. (2016), Neptune's Dynamic Atmosphere from Kepler K2

1576 Observations: Implications for Brown Dwarf Light Curve Analyses, The Astrophysical Journal, 817, 162,

1577 10.3847/0004-637X/817/2/162.

1578

1579 Slattery, W. L., Benz, W., & Cameron, A. G. W. (1992), Giant impacts on a primitive Uranus, Icarus, 99, 167,

1580 10.1016/0019-1035(92)90180-F.

1581

66 Fletcher et al., Ice Giant Systems

1582 Smith, B. A., Soderblom, L. A., Banfield, D., et al. (1989), Voyager 2 at Neptune: Imaging Science Results,

1583 Science, 246, 1422, 10.1126/science.246.4936.1422.

1584

1585 Smith, M. D., & Gierasch, P. J. (1995), Convection in the outer planet atmospheres including ortho-para

1586 hydrogen conversion., Icarus, 116, 159, 10.1006/icar.1995.1118.

1587

1588 Soderblom, L. A., Kieffer, S. W., Becker, T. L., et al. (1990), Triton's Geyser-Like Plumes: Discovery and Basic

1589 Characterization, Science, 250, 410, 10.1126/science.250.4979.410.

1590

1591 Soderlund, K. M., Heimpel, M. H., King, E. M., et al. (2013), Turbulent models of ice giant internal dynamics:

1592 Dynamos, heat transfer, and zonal flows, Icarus, 224, 97, 10.1016/j.icarus.2013.02.014.

1593

1594 Sromovsky, L. A., de Pater, I., Fry, P. M., et al. (2015), High S/N Keck and Gemini AO imaging of Uranus during

1595 2012-2014: New cloud patterns, increasing activity, and improved wind measurements, Icarus, 258, 192,

1596 10.1016/j.icarus.2015.05.029.

1597

1598 Sromovsky, L. A., Karkoschka, E., Fry, P. M., et al. (2014), Methane depletion in both polar regions of Uranus

1599 inferred from HST/STIS and Keck/NIRC2 observations, Icarus, 238, 137, 10.1016/j.icarus.2014.05.016.

1600

1601 Stanley, S., & Bloxham, J. (2004), Convective-region geometry as the cause of Uranus' and Neptune's unusual

1602 magnetic fields, Nature, 428, 151, 10.1038/nature02376.

1603

1604 Stanley, S., & Bloxham, J. (2006), Numerical dynamo models of Uranus' and Neptune's magnetic fields, Icarus,

1605 184, 556, 10.1016/j.icarus.2006.05.005.

1606

67 Fletcher et al., Ice Giant Systems

1607 Stauffer, J., Marley, M. S., Gizis, J. E., et al. (2016), Spitzer Space Telescope Mid-IR Light Curves of Neptune,

1608 The Astronomical Journal, 152, 142, 10.3847/0004-6256/152/5/142.

1609

1610 Stern, S. A., & McKinnon, W. B. (2000), Triton's Surface Age and Impactor Population Revisited in Light of

1611 Kuiper Belt Fluxes: Evidence for Small Kuiper Belt Objects and Recent Geological Activity, The Astronomical

1612 Journal, 119, 945, 10.1086/301207.

1613

1614 Stevenson, D. J. (1986), The Uranus-Neptune Dichotomy: the Role of Giant Impacts, Lunar and Planetary

1615 Science Conference, 1011,..

1616

1617 Stoker, C. R., & Toon, O. B. (1989), Moist convection on Neptune, Geophysical Research Letters, 16, 929,

1618 10.1029/GL016i008p00929.

1619

1620 Stone, E. C., Cooper, J. F., Cummings, A. C., et al. (1986), Energetic Charged Particles in the Uranian

1621 Magnetosphere, Science, 233, 93, 10.1126/science.233.4759.93.

1622

1623 Stratman, P. W., Showman, A. P., Dowling, T. E., et al. (2001), EPIC Simulations of Bright Companions to

1624 Neptune's Great Dark Spots, Icarus, 151, 275, 10.1006/icar.2001.6603.

1625

1626 Strobel, D. F., Cheng, A. F., Summers, M. E., et al. (1990), Magnetospheric interaction with Triton's

1627 ionosphere, Geophysical Research Letters, 17, 1661, 10.1029/GL017i010p01661.

1628

1629 Stryk, T., & Stooke, P. J. (2008), Voyager 2 Images of Uranian Satellites: Reprocessing and New

1630 Interpretations, Lunar and Planetary Science Conference, 1362,..

1631

68 Fletcher et al., Ice Giant Systems

1632 Sun, Z.-P., Schubert, G., & Stoker, C. R. (1991), Thermal and humidity winds in outer planet atmospheres,

1633 Icarus, 91, 154, 10.1016/0019-1035(91)90134-F.

1634

1635 Tegler, S. C., Grundy, W. M., Olkin, C. B., et al. (2012), Ice Mineralogy across and into the Surfaces of Pluto,

1636 Triton, and Eris, The Astrophysical Journal, 751, 76, 10.1088/0004-637X/751/1/76.

1637

1638 Tegler, S. C., Stufflebeam, T. D., Grundy, W. M., et al. (2019), A New Two-molecule Combination Band as a

1639 Diagnostic of Carbon Monoxide Diluted in Nitrogen Ice on Triton, The Astronomical Journal, 158, 17,

1640 10.3847/1538-3881/ab199f.

1641

1642 Thompson, W. R., & Sagan, C. (1990), Color and chemistry on Triton, Science, 250, 415,

1643 10.1126/science.11538073.

1644

1645 Tiscareno, M. S., Hedman, M. M., Burns, J. A., et al. (2013), Compositions and Origins of Outer Planet Systems:

1646 Insights from the Roche Critical Density, The Astrophysical Journal, 765, L28, 10.1088/2041-8205/765/2/L28.

1647

1648 Tittemore, W. C., & Wisdom, J. (1990), Tidal evolution of the Uranian satellites III. Evolution through the

1649 Miranda-Umbriel 3:1, Miranda-Ariel 5:3, and Ariel-Umbriel 2:1 mean-motion commensurabilities, Icarus, 85,

1650 394, 10.1016/0019-1035(90)90125-S.

1651

1652 Tosi, F., Turrini, D., Coradini, A., et al. (2010), Probing the origin of the dark material on Iapetus, Monthly

1653 Notices of the Royal Astronomical Society, 403, 1113, 10.1111/j.1365-2966.2010.16044.x.

1654

1655 Turrini, D., Politi, R., Peron, R., et al. (2014), The comparative exploration of the ice giant planets with twin

1656 spacecraft: Unveiling the history of our Solar System, Planetary and Space Science, 104, 93,

1657 10.1016/j.pss.2014.09.005.

69 Fletcher et al., Ice Giant Systems

1658

1659 Venturini, J., & Helled, R. (2017), The Formation of Mini-Neptunes, The Astrophysical Journal, 848, 95,

1660 10.3847/1538-4357/aa8cd0.

1661

1662 Waite, J. H., Perryman, R. S., Perry, M. E., et al. (2018), Chemical interactions between Saturn's atmosphere

1663 and its rings, Science, 362, aat2382, 10.1126/science.aat2382.

1664

1665 Wakeford, H. R., Visscher, C., Lewis, N. K., et al. (2017), High-temperature condensate clouds in super-hot

1666 Jupiter atmospheres, Monthly Notices of the Royal Astronomical Society, 464, 4247,

1667 10.1093/mnras/stw2639.

1668

1669 Wang, C., & Richardson, J. D. (2004), Interplanetary coronal mass ejections observed by Voyager 2 between

1670 1 and 30 AU, Journal of Geophysical Research (Space Physics), 109, A06104, 10.1029/2004JA010379.

1671

1672 Wei, Y., Pu, Z., Zong, Q., et al. (2014), Oxygen escape from the Earth during geomagnetic reversals:

1673 Implications to mass extinction, Earth and Planetary Science Letters, 394, 94, 10.1016/j.epsl.2014.03.018.

1674

1675 Witasse, O., Sánchez-Cano, B., Mays, M. L., et al. (2017), Interplanetary coronal mass ejection observed at

1676 STEREO-A, Mars, comet 67P/Churyumov-Gerasimenko, Saturn, and New Horizons en route to Pluto:

1677 Comparison of its Forbush decreases at 1.4, 3.1, and 9.9 AU, Journal of Geophysical Research (Space Physics),

1678 122, 7865, 10.1002/2017JA023884.

1679

1680 Wong, M. H., Tollefson, J., Hsu, A. I., et al. (2018), A New Dark Vortex on Neptune, The Astronomical Journal,

1681 155, 117, 10.3847/1538-3881/aaa6d6.

1682

70 Fletcher et al., Ice Giant Systems

1683 Zarka, P., & Pedersen, B. M. (1986), Radio detection of uranian lightning by Voyager 2, Nature, 323, 605,

1684 10.1038/323605a0.

1685

1686 Zarka, P., Pedersen, B. M., Lecacheux, A., et al. (1995), Radio emissions from Neptune., Neptune and Triton,

1687 341,..

1688

1689 Zhang, Z., Hayes, A. G., Janssen, M. A., et al. (2017), Exposure age of Saturn's A and B rings, and the Cassini

1690 Division as suggested by their non-icy material content, Icarus, 294, 14, 10.1016/j.icarus.2017.04.008.

1691

1692

71 Fletcher et al., Ice Giant Systems

1 Ice Giant Systems:

2 The Scientific Potential of Orbital Missions to Uranus and Neptune

3

4 Leigh N. Fletcher (School of Physics and Astronomy, University of Leicester, University Road, Leicester, LE1

5 7RH, UK; [email protected]; Tel: +441162523585), Ravit Helled (University of Zurich, Switzerland), Elias

6 Roussos (Max Planck Institute for Solar System Research, Germany), Geraint Jones (Mullard Space Science

7 Laboratory, UK), Sébastien Charnoz (Institut de Physique du Globe de Paris, France), Nicolas André (Institut

8 de Recherche en Astrophysique et Planétologie, Toulouse, France), David Andrews (Swedish Institute of

9 Space Physics, Sweden), Michele Bannister (Queens University Belfast, UK), Emma Bunce (University of

10 Leicester, UK), Thibault Cavalié (Laboratoire d’Astrophysique de Bordeaux, Univ. Bordeaux, CNRS, Pessac,

11 France; LESIA, Observatoire de Paris, Université PSL, CNRS, Sorbonne Université, Univ. Paris Diderot,

12 Sorbonne Paris Cité, Meudon, France), Francesca Ferri (Università degli Studi di Padova, Italy), Jonathan

13 Fortney (University of Santa-Cruz, USA), Davide Grassi (Istituto Nazionale di AstroFisica – Istituto di

14 Astrofisica e Planetologia Spaziali (INAF-IAPS), Rome, Italy), Léa Griton (Institut de Recherche en

15 Astrophysique et Planétologie, CNRS, Toulouse, France), Paul Hartogh (Max-Planck-Institut für

16 Sonnensystemforschung, Germany), Ricardo Hueso (Escuela de Ingeniería de Bilbao, UPV/EHU, Bilbao,

17 Spain), Yohai Kaspi (Weizmann Institute of Science, Israel), Laurent Lamy (Observatoire de Paris, France),

18 Adam Masters (Imperial College London, UK), Henrik Melin (University of Leicester, UK), Julianne Moses

19 (Space Science Institute, USA), Oliver Mousis (Aix Marseille Univ, CNRS, CNES, LAM, Marseille, France),

20 Nadine Nettleman (University of Rostock, Germany), Christina Plainaki (ASI - Italian Space Agency, Italy),

21 Jürgen Schmidt (University of Oulu, Finland), Amy Simon (NASA Goddard Space Flight Center, USA), Gabriel

22 Tobie (Université de Nantes, France), Paolo Tortora (Università di Bologna, Italy), Federico Tosi (Istituto

23 Nazionale di AstroFisica – Istituto di Astrofisica e Planetologia Spaziali (INAF-IAPS), Rome, Italy), Diego Turrini

24 (Istituto Nazionale di AstroFisica – Istituto di Astrofisica e Planetologia Spaziali (INAF-IAPS), Rome, Italy)

25

26

27 Abstract: Fletcher et al., Ice Giant Systems

28 Uranus and Neptune, and their diverse satellite and ring systems, represent the least explored environments

29 of our Solar System, and yet may provide the archetype for the most common outcome of planetary

30 formation throughout our galaxy. Ice Giants will be the last remaining class of Solar System planet to have a

31 dedicated orbital explorer, and international efforts are under way to realise such an ambitious mission in

32 the coming decades. In 2019, the European Space Agency released a call for scientific themes for its strategic

33 science planning process for the 2030s and 2040s, known as Voyage 2050. We used this opportunity to

34 review our present-day knowledge of the Uranus and Neptune systems, producing a revised and updated set

35 of scientific questions and motivations for their exploration. This review article describes how such a mission

36 could explore their origins, ice-rich interiors, dynamic atmospheres, unique magnetospheres, and myriad icy

37 satellites, to address questions at the heart of modern planetary science. These two worlds are superb

38 examples of how planets with shared origins can exhibit remarkably different evolutionary paths: Neptune

39 as the archetype for Ice Giants, whereas Uranus may be atypical. Exploring Uranus’ natural satellites and

40 Neptune’s captured moon Triton could reveal how Ocean Worlds form and remain active, redefining the

41 extent of the habitable zone in our Solar System. For these reasons and more, we advocate that an Ice Giant

42 System explorer should become a strategic cornerstone mission within ESA’s Voyage 2050 programme, in

43 partnership with international collaborators, and targeting launch opportunities in the early 2030s.

44

45 Keywords: Giant Planets; Ice Giants; Robotic Missions; Orbiters; Probes

46

47 1. Introduction: Why Explore Ice Giant Systems?

48 1.1 Motivations

49

50 The early 21st century has provided unprecedented leaps in our exploration of the Gas Giant systems, via the

51 completion of the Galileo and Cassini orbital missions at Jupiter and Saturn; NASA/Juno’s ongoing exploration

52 of Jupiter’s deep interior, atmosphere, and magnetic field; ESA’s development of the JUICE mission (Jupiter

53 Icy Moons Explorer) and NASA’s Europa Clipper to explore the Galilean satellites. The past decade has also 2 Fletcher et al., Ice Giant Systems

Figure 1 Each Ice Giant exhibits a rich system of planetary environments to explore, from their mysterious interiors, atmospheres and magnetospheres, to the diverse satellites and rings. The inner systems are to scale, with arrows next to major moons indicating that they orbit at larger planetocentric distances. The magnetosphere and radiation belts would encompass the full area of the figure. Credit: L.N. Fletcher/M. Hedman/E. Karkoschka/Voyager-2. 54 provided our first glimpses of the diversity of planetary environments in the outer solar system, via the New

55 Horizons mission to Pluto. Conversely, the realm of the Ice Giants, from Uranus (20 AU) to Neptune (30 AU),

56 remains largely unexplored, each system having been visited only once by a flyby spacecraft – Voyager 2 – in

57 1986 and 1989 respectively. More than three decades have passed since our first close-up glimpses of these

58 worlds, with cameras, spectrometers and sensors based on 1960s and 70s technologies. Voyager’s systems

59 were not optimised for the Ice Giants, which were considered to be extended mission targets. Uranus and

60 Neptune have therefore never had a dedicated mission, despite the rich and diverse systems displayed in

61 Figure 1. A return to the Ice Giants with an orbiter is the next logical step in humankind’s exploration of our

62 Solar System.

63

64 The Ice Giants may be our closest and best representatives of a whole class of astrophysical objects, as

65 Neptune-sized worlds have emerged as the dominant category in our expanding census of exoplanets (Fulton

66 et al., 2018), intermediate between the smaller terrestrial worlds and the larger hydrogen-rich gas giants

67 (Section 3.3). Our own Ice Giants offer an opportunity to explore physical and chemical processes within

68 these planetary systems as the archetype for these distant exoplanets (Rymer et al., 2019). Furthermore,

69 the formation and evolution of Uranus and Neptune (Section 2.1) pose a critical test of our understanding of

70 planet formation, and of the architecture of our Solar System. Their small size, compared to Jupiter, places 3 Fletcher et al., Ice Giant Systems

71 strong constraints on the timing of planet formation. Their bulk internal composition (i.e., the fraction of

72 rock, ices, and gases) and the differentiation with depth (i.e., molecular weight gradients, phase transitions

73 to form global water oceans and icy mantles) are poorly known, but help determine the conditions and

74 dynamics in the outer planetary nebula at the time of planet formation.

75

76 Uranus and Neptune also provide two intriguing endmembers for the Ice Giant class. Neptune may be

77 considered the archetype for a seasonal Ice Giant, whereas the cataclysmic collision responsible for Uranus’

78 extreme tilt renders it unique in our Solar System. Contrasting the conditions on these two worlds provides

79 insights into differential evolution from shared origins. The atmospheres of Uranus and Neptune (Section

80 2.2) exemplify the contrasts between these worlds. Uranus’ negligible internal heat renders its atmosphere

81 extremely sluggish, with consequences for storms, meteorology, and atmospheric chemistry. Conversely,

82 Neptune’s powerful winds and rapidly-evolving storms demonstrates how internal energy can drive powerful

83 weather, despite weak sunlight at 30 AU. Both of these worlds exhibit planetary banding, although the

84 atmospheric circulation responsible for these bands (and their associated winds, temperatures, composition

85 and clouds) remain unclear, and the connection to atmospheric flows below the topmost clouds remains a

86 mystery.

87

88 Conditions within the Ice Giant magnetospheres (Section 2.3) are unlike those found anywhere else, with

89 substantial offsets between their magnetic dipole axes and the planets’ rotational axes implying a system

90 with an extremely unusual interaction with the solar wind and internal plasma processes, varying on both

91 rotational cycles as the planet spins, and on orbital cycles.

92

93 The diverse Ice Giant satellites (Section 2.4) and narrow, incomplete ring systems (Section 2.5) provide an

94 intriguing counterpoint to the better-studied Jovian and Saturnian systems. Uranus may feature a natural,

95 primordial satellite system with evidence of extreme and violent events (e.g., Miranda). Neptune hosts a

4 Fletcher et al., Ice Giant Systems

96 captured Kuiper Belt Object, Triton, which may itself harbour a sub-surface ocean giving rise to active surface

97 geology (e.g., south polar plumes and cryovolcanism).

98

99 Advancing our knowledge of the Ice Giants and their diverse satellite systems requires humankind’s first

100 dedicated explorer for this distant realm. Such a spacecraft should combine interior science via gravity and

101 magnetic measurements, in situ measurements of their plasma and magnetic field environments, in situ

102 sampling of their chemical composition, and close-proximity multi-wavelength remote sensing of the planets,

103 their rings, and moons. This review article will summarise our present understanding of these worlds, and

104 propose a revised set of scientific questions to guide our preparation for such a mission. This article is

105 motivated by ESA’s recent call for scientific themes as part of its strategic space mission planning in the period

106 from 2035 to 2050,1 and is therefore biased to European perspectives on an Ice Giant mission, as explored in

107 the next section.

108

109 1.2 Ice Giants in ESA’s Cosmic Vision

110

111 The exploration of the Ice Giants addresses themes at the heart of ESA’s existing Cosmic Vision2 programme,

112 namely (1) exploring the conditions for planet formation and the emergence of life; (2) understanding how

113 our solar system works; and (3) exploring the fundamental physical laws of the universe. European-led

114 concepts for Ice Giant exploration have been submitted to several ESA Cosmic Vision competitions. The

115 Uranus Pathfinder mission, an orbiting spacecraft based on heritage from Mars Express and Rosetta, was

116 proposed as a medium-class (~€0.5bn) mission in both the M3 (2010) and M4 (2014) competitions (Arridge

117 et al., 2012). However, the long duration of the mission, limited power available, and the programmatic

118 implications of having NASA provide the launch vehicle and radioisotope thermoelectric generators (RTGs),

119 meant that the Pathfinder concept did not proceed to the much-needed Phase A study.

1 https://www.cosmos.esa.int/web/voyage-2050 2 http://sci.esa.int/cosmic-vision/38542-esa-br-247-cosmic-vision-space-science-for-europe-2015-2025/

5 Fletcher et al., Ice Giant Systems

120

121 The importance of Ice Giant science was reinforced by multiple submissions to ESA's call for large-class

122 mission themes in 2013: a Uranus orbiter with atmospheric probe (Arridge et al., 2014), an orbiter to explore

123 Neptune and Triton (Masters et al., 2014); and a concept for dual orbiters of both worlds (Turrini et al., 2014).

124 Once again, an ice giant mission failed to proceed to the formal study phase, but ESA's Senior Survey

125 Committee (SSC3) commented that ``the exploration of the icy giants appears to be a timely milestone, fully

126 appropriate for an L class mission. The whole planetology community would be involved in the various aspects

127 of this mission... the SSC recommends that every effort is made to pursue this theme through other means,

128 such as cooperation on missions led by partner agencies."

129

130 This prioritisation led to collaboration between ESA and NASA in the formation of a science definition team

131 (2016-17), which looked more closely at a number of different mission architectures for a future mission to

132 the Ice Giants (Hofstadter et al., 2019). In addition, ESA's own efforts to develop nuclear power sources for

133 space applications have been progressing, with prototypes now developed to utilise the heat from the decay

134 of 241Am as their power source (see Section 4.3), provided that the challenge of their low energy density can

135 be overcome. Such an advance might make smaller, European-led missions to the Ice Giants more realistic,

136 and addresses many of the challenges faced by the original Uranus Pathfinder concepts.

137

138 At the start of the 2020s, NASA and ESA are continuing to explore the potential for an international mission

139 to the Ice Giants. A palette of potential contributions (M-class in scale) to a US-led mission have been

140 identified by ESA4, and US scientists are currently undertaking detailed design and costing exercises for

141 missions to be assessed in the upcoming US Planetary Decadal Survey (~2022). Each of these emphasise

142 launch opportunities in the early 2030s (Section 4.2), with arrival in the early 2040s (timelines for Ice Giant

3 http://sci.esa.int/cosmic-vision/53261-report-on-science-themes-for-the-l2-and-l3-missions/ 4 http://sci.esa.int/future-missions-department/61307-cdf-study-report-ice-giants/

6 Fletcher et al., Ice Giant Systems

143 missions will be described in Section 4.3). This review article summarises those studies, whilst taking the

144 opportunity to update and thoroughly revise the scientific rationale for Ice Giant missions compared to

145 Arridge et al. (2012, 2014), Masters et al. (2014), Turrini et al. (2014) and Hofstadter et al. (2019). We focus

146 on the science achievable from orbit, as the science potential of in situ entry probes has been discussed

147 elsewhere (Mousis et al., 2018). Section 2 reviews our present-day knowledge of Ice Giant Systems; Section

148 3 places the Ice Giants into their wider exoplanetary context; Section 4 briefly reviews the recent mission

149 concept studies and outstanding technological challenges; and Section 5 summarises our scientific goals at

150 the Ice Giants at the start of the 2020s.

151

152 2. Science Themes for Ice Giant Exploration

153

154 In this section we explore the five multi-disciplinary scientific themes that could be accomplished via

155 orbital exploration of the Ice Giants, and show how in-depth studies of fundamental processes at Uranus and

156 Neptune would have far-reaching implications in our Solar System and beyond. Each sub-section is organised

157 via a series of high-level questions that could form the basis of a mission traceability matrix.

158

159 2.1 Ice Giant Origins and Interiors

160

161 What does the origin, structure, and composition of the two Ice Giants reveal about the formation of

162 planetary systems? Understanding the origins and internal structures of Uranus and Neptune will

163 substantially enhance our understanding of our own Solar System and intermediate-mass exoplanets. Their

164 bulk composition provides crucial constraints on the conditions in the solar nebula during planetary

165 formation and evolution.

166

167 How did the Ice Giants first form, and what constraints can be placed on the mechanisms for planetary

168 accretion? The formation of Uranus and Neptune has been a long-standing problem for planet formation 7 Fletcher et al., Ice Giant Systems

169 theory (e.g., Pollack et al., 1996, Dodson-Robinson & Bodenheimer, 2010, Helled & Bodenheimer, 2014). Yet,

170 the large number of detected exoplanets with sizes comparable (or smaller) to that of Uranus and Neptune

171 suggests that such planetary objects are very common (e.g., Batalha et al. 2013), a fact that is in conflict with

172 theoretical calculations.

173

174 The challenge for formation models is to prevent Uranus and Neptune from accreting large amounts of

175 hydrogen-helium (H-He) gas, like the Gas Giants Jupiter and Saturn, to provide the correct final mass and gas-

176 to-solids ratios as inferred by structure models. In the standard planet formation model, core accretion (see

177 Helled et al., 2014 for review and the references therein), a slow planetary growth is expected to occur at

178 large radial distances where the solid surface density is lower, and the accretion rate (of planetesimals) is

179 significantly smaller. For the current locations of Uranus and Neptune, the formation timescale can be

180 comparable to the lifetimes of protoplanetary disks. Due to long accretion times at large radial distances, the

181 formation process is too slow to reach the phase of runaway gas accretion, before the gas disk disappears,

182 leaving behind an intermediate-mass planet (a failed giant planet), which consists mostly of heavy elements

183 and a small fraction of H-He gas.

184

185 However, since the total mass of H-He in both Uranus and Neptune is estimated to be 2-3 Earth masses (M),

186 it implies that gas accretion had already begun, and this requires that the gas disk disappears at a very specific

187 time, to prevent further gas accretion onto the planets. This is known as the fine-tuning problem in

188 Uranus/Neptune formation (e.g., Venturini & Helled, 2017). Another possibility is that Uranus and Neptune

189 formed in situ within a few Mys by pebble accretion. In this formation scenario, the core’s growth is more

190 efficient than in the planetesimal accretion case, and the pebble isolation mass is above 20 M. As a result,

191 the forming planets could be heavy-element dominated with H-He envelopes that are metal-rich due to

192 sublimation of icy pebbles (e.g., Lambrechts et al., 2014).

193

8 Fletcher et al., Ice Giant Systems

194 Measuring the elemental abundances in the atmospheres of Uranus and Neptune can provide information

195 on their formation history by setting limits on their formation locations and/or the type of solids

196 (pebbles/planetesimals) that were accreted. Measurements of the elemental abundances of well-mixed

197 noble gases, which are only accessible via in situ entry probes, would be particularly informative (e.g. Mousis

198 et al., 2018). In addition, determining the atmospheric metallicity provides valuable constraints for structure

199 models, as discussed below.

200

201 What is the role of giant impacts in explaining the differences between Uranus and Neptune? Uranus and

202 Neptune are somewhat similar in terms of mass and radius, but they also have significant differences such

203 as their tilt, internal heat flux, and satellite system. It is possible that these observed differences are a result

204 of giant impacts that occurred after the formation of the planets (e.g., Safronov 1966, Stevenson 1986). An

205 oblique impact of a massive impactor can explain Uranus’ spin and lead to the formation of a disk where the

206 regular satellites form. Neptune could have also suffered a head-on impact that could have reached the deep

207 interior, providing sufficient energy (and mass) to explain the higher heat flux, and possibly the higher mass

208 and moment of inertia value (e.g., Stevenson 1986, Podolak & Helled, 2012). Giant impact simulations by

209 various groups confirmed that Uranus’ tilt and rotation could be explained by a giant impact (Slattery et al.,

Figure 2 Sketches of the possible internal structures of the ice giants. It is unclear whether the planets are differentiated and whether the transitions between the different layers are distinct or gradual: (a) separation between the ices and rocks and the ice and H-He atmosphere (b) separation (phase boundary) between the H-He atmosphere and ices and a gradual transition between ice and rock, (c) gradual transition between the H-He atmosphere and ice layer, and a distinct separation between the ice and rock layers, and (d) gradual transition both between the H-He atmosphere and ice and the ice and rocks suggesting a global composition gradient with the planets (see text for discussion). 9 Fletcher et al., Ice Giant Systems

210 1992, Kegerreis et al., 2019). Nevertheless, alternative explanations have been proposed: Boue and Laskar

211 (2010) showed that under special circumstances, the high obliquity could potentially arise from orbital

212 migration without the need for a collision. Understanding the cause of Uranus’ tilt and the mechanisms that

213 led to the observed differences between the planets are key questions in planetary science.

214

215 What is the bulk composition and internal structure of Uranus and Neptune? There are still substantial

216 uncertainties regarding the bulk compositions and internal structures of Uranus and Neptune (e.g., Podolak

217 et al., 1995, Podolak & Helled, 2012, Nettelmann et al., 2013). The available measurements of their physical

218 properties such as mass, radius, rotation rate, atmospheric temperature, and gravitational and magnetic

219 fields are used to constrain models of their interiors. For the Ice Giants, only the gravitational harmonic

220 coefficients J2 and J4 are known and their error bars are relatively large (Jacobson, 2009; 2014), nevertheless,

221 various studies have aimed to constrain the planets’ internal structures.

222

223 Standard structure models of the planets consist of three layers: a rocky core, an 'icy' shell (water, ammonia,

224 methane, etc.), and a gaseous envelope composed of H-He and heavier elements. The middle layer is not

225 made of “ice” in regard to the physical state of the material (i.e., solid), but is referred to as an icy layer since

226 it is composed of volatile materials such as water, ammonia and methane. The masses and compositions of

227 the layers are modified until the model fits the observed parameters using a physical equation of state (EOS)

228 to represent the materials. Three-layer models predict very high ice-to-rock ratios, where the ice fraction is

229 found to be higher than the rock fraction by 19-35 times for Uranus, and 4-15 times for Neptune, with the

230 total H-He mass typically being 2 and 3 M for Uranus and Neptune, respectively. The exact estimates are

231 highly model-dependent, and are sensitive to the assumed composition, thermal structure and rotation rate

232 of the planets (Helled et al., 2011, Nettelmann et al., 2013).

233

234 The interiors of Uranus and Neptune could also be more complex with the different elements being mixed,

235 and could also include thermal boundary layers and composition gradients. Indeed, alternative structure

10 Fletcher et al., Ice Giant Systems

236 models of the planets suggested that Uranus and Neptune could have a density profile without

237 discontinuities (e.g., Helled et al., 2011), and that the planets do not need to contain large fractions of water

238 to fit their observed properties. This alternative model implies that Uranus and Neptune may not be as water-

239 rich as typically thought, but instead are rock-dominated like Pluto (e.g., McKinnon et al., 2017) and could be

240 dominated by composition gradients. It is therefore possible that the “ice giants” are in fact not ice-

241 dominated (see Helled et al., 2011 for details). The very large ice-to-rock ratios found from structure models

242 also suggest a more complex interior structure. At the moment, we have no way to discriminate among the

243 different ice-to-rock ratios inferred from structure models. As a result, further constraints on the gravity and

244 magnetic data (Section 2.3), as well as atmospheric composition and isotopic ratios (e.g., D/H, Atreya et al.,

245 2020) are required.

246

247 How can Ice Giant observations be used to explore the states of matter (e.g., water) and mixtures (e.g.,

248 rocks, water, H-He) under the extreme conditions of planetary interiors? In order to predict the mixing

249 within the planets, knowledge from equation-of-state (EOS) calculations is required. Internal structure

250 models must be consistent with the phase diagram of the assumed materials and their mixtures. This is a

251 challenging task and progress in that direction is ongoing. EOS calculations can be used to guide model

252 assumptions. For example, it is possible that Uranus and Neptune have deep water oceans that begin where

253H 2 and H2O become insoluble (e.g., Bailey & Stevenson, 2015, Bali et al., 2013). Figure 2 presents sketches

254 of four possible internal structures of the ice giants where the transitions between layers are distinct (via

255 phase/thermal boundaries) and/or gradual.

256

257 Current observational constraints, foremost J2 and J4, clearly indicate that the deep interior is more enriched

258 in heavy elements than the outer part. Understanding the nature and origin of the compositional gradient

259 zone would yield important information on the formation process and subsequent evolution including

260 possible processes such as outgassing, immiscibility, and sedimentation of ices; processes that play a major

261 role on terrestrial planets and their habitability.

11 Fletcher et al., Ice Giant Systems

262

263 What physical and chemical processes during the planetary formation and evolution shape the magnetic

264 field, thermal profile, and other observable quantities? Structure models must be consistent with the

265 observed multi-polar magnetic fields (see Section 2.3), implying that the outermost ~20% of the planets is

266 convective and consists of conducting materials (e.g., Stanley & Bloxham, 2004, 2006). Currently, the best

267 candidate for the generation of the dynamo is the existence of partially dissociated fluid water in the

268 outermost layers (e.g., Redmer et al.,2011), located above solid and non-convecting superionic water ice

269 ‘mantle’ (Millot et al., 2019). Dynamo models that fit the Voyager magnetic field data suggest the deep

270 interior is stably stratified (Stanley & Bloxham 2004, 2006) or, alternatively, in a state of thermal-buoyancy

271 driven turbulent convection (Soderlund et al., 2013). Improved measurements of the magnetic fields of

272 Uranus and Neptune will also help to constrain the planetary rotation rate. Since Voyager’s measurements

273 of the periodicities in the radio emissions and magnetic fields have not been confirmed by another

274 spacecraft, it is unclear whether the Voyager rotation rate reflects the rotation of the deep interior (Helled

275 et al., 2010), with major consequences for the inferred planetary structure and the question of similar or

276 dissimilar interiors (Nettelmann et al., 2013).

277

278 Finally, the different intrinsic heat fluxes of Uranus and Neptune imply that they followed different

279 evolutionary histories. This could be a result of a different growth history or a result of giant impacts during

280 their early evolution (e.g., Reinhardt et al., 2020). Moreover, thermal evolution models that rely on Voyager’s

281 measurements of the albedo, brightness temperatures, and atmospheric pressure-temperature profiles that

282 are used to model the evolution of the two planets cannot explain both planets with the same set of

283 assumptions.

284

285 Summary: A better understanding of the origin, evolution and structure of Ice Giant planets requires new

286 and precise observational constraints on the planets' gravity field, rotation rate, magnetic field, atmospheric

287 composition, and atmospheric thermal structure, both from orbital observations and in situ sampling from

12 Fletcher et al., Ice Giant Systems

288 an atmospheric probe. The insights into origins, structures, dynamo operation and bulk composition

289 provided by an Ice Giant mission would not only shed light on the planet-forming processes at work in our

290 Solar System, but could also help to explain the most common planetary class throughout our observable

291 universe.

292

293 2.2 Ice Giant Atmospheres

294

295 Why do atmospheric processes differ between Uranus, Neptune, and the Gas Giants, and what are the

296 implications for Neptune-mass worlds throughout our universe? Ice Giant atmospheres occupy a wholly

297 different region of parameter space compared to their Gas Giant cousins. Their dynamics and chemistry are

298 driven by extremes of internal energy (negligible on Uranus, but powerful on Neptune) and extremes of solar

299 insolation (most severe on Uranus due to its 98o axial tilt) that are not seen anywhere else in the Solar System

300( Figure 3). Their smaller planetary radii, compared to Jupiter and Saturn, affects the width of zonal bands

301 and the drift behaviour of storms and vortices. Their zonal winds are dominated by broad retrograde

302 equatorial jets and do not exhibit the fine-scale banding found on Jupiter, which means that features like

303 bright storms and dark vortices are able to drift with latitude during their lifetimes. Their hydrogen-helium

304 atmospheres are highly enriched in volatiles like CH4 and H2S that show strong equator-to-pole gradients,

305 changing the atmospheric density and hence the vertical shear on the winds (Sun et al., 1991). Their

306 temperatures are so low that the energy released from interconversion between different states of hydrogen

307 (ortho and para spin isomers) can play a role in shaping atmospheric dynamics (Smith & Gierasch, 1995).

308 Their middle and upper atmospheres are both much hotter than can be explained by weak solar heating

309 alone, implying a decisive role for additional energy from internal (e.g., waves) or external sources (e.g.,

310 currents induced by complex coupling to the magnetic field). As the atmospheres are the windows through

311 which we interpret the bulk properties of planets, these defining properties of Ice Giants can provide insights

312 into atmospheric processes on intermediate-sized giant planets beyond our Solar System.

313

13 Fletcher et al., Ice Giant Systems

314 A combination of global multi-wavelength remote sensing from an orbiter and in situ measurements from an

315 entry probe would provide a transformative understanding of these unique atmospheres, focussing on the

316 following key questions (summarised in Figure 3):

317

14 Fletcher et al., Ice Giant Systems

Figure 3 Ice Giant atmospheres are shaped by dynamical, chemical and radiative processes that are not found elsewhere in our Solar System. Images A & C are false-colour representations of Voyager 2 observations of Uranus and Neptune, respectively. Images B and D were acquired by the Hubble Space Telescope in 2018.

318 What are the dynamical, meteorological, and chemical impacts of the extremes of planetary luminosity?

319 Despite the substantial differences in self-luminosity (Pearl et al., 1990, 1991), seasonal influences,

320 atmospheric activity (Hueso and Sanchez-Lavega et al., 2019), and the strength of vertical mixing (resulting

321 in differences in atmospheric chemistry, Moses et al., 2018), there are many similarities between these two

322 worlds. In their upper tropospheres, tracking of discrete cloud features has revealed that both planets exhibit

323 broad retrograde jets at their equators and prograde jets nearer the poles (Sanchez-Lavega et al., 2018), but

324 unlike Jupiter, these are seemingly disconnected from the fine-scale banding revealed in the visible and near-

325 infrared (Sromovsky et al., 2015). Is this simply an observational bias, or are winds on the ice giants truly

326 different from those on Jupiter and Saturn? What sets the scales of the bands? On the Gas Giants, small-

327 scale eddies (from atmospheric instabilities and convective storms) appear to feed energy into the large-

328 scale winds, but we have never been able to investigate similar processes on Uranus and Neptune. Indeed,

329 convective processes themselves could be substantially different – moist convection driven by the

330 condensation of water will likely play a very limited role in the observable atmosphere, as water is restricted

331 to pressures that exceed tens or hundreds of bars. Instead, convection in the observable tropospheres may

332 be driven by methane condensation in the 0.1-1.5 bar range (Stoker & Toon, 1989), or by heat release by

333 ortho-para-H2 conversion (Smith & Gierasch, 1995). These sources of energy operate in a very different way

15 Fletcher et al., Ice Giant Systems

334 compared to those available on Jupiter and Saturn. Firstly, the high enrichment in methane in the Ice Giants

335 could inhibit vertical motions due to a vertical gradient of the atmospheric molecular weight (Guillot et al.,

336 1995; Leconte et al., 2017). This phenomenon may also be at work in the deep and inaccessible water clouds

337 of both the Gas and Ice giants, but the observable methane clouds of Uranus and Neptune provide an

338 excellent opportunity to study it (Guillot et al., 2019). Secondly, heat release by ortho-para H2 conversion is

339 much more efficient in providing energy within the cold atmospheres of Uranus and Neptune. Thus,

340 convection may occur in vertically-thin layers (Gierasch et al., 1987), rather than extending vertically over

341 tremendous heights, or in a complex and inhomogeneous weather layer (Hueso et al., 2020). Multi-

342 wavelength remote sensing of the temperatures, clouds, winds, and gaseous composition is required to

343 investigate how these meteorological processes differ from the Gas Giants, how they derive their energy

344 from the internal heating or weak sunlight, and their relation to the large-scale banded patterns and winds

345 (Fletcher et al., 2020). Spatially-resolved reflectivity and thermal-emission mapping will allow precise

346 constraints on the Ice Giant energy balance to constrain their self-luminosities. And mapping the distribution

347 and depth of Ice Giant lightning, previously detected via radio emissions on both worlds (Zarka et al., 1986;

348 Gurnett et al., 1990), could determine the frequency and intensity of water-powered convection on the Ice

349 Giants, elucidating its impact on their atmospheric dynamics.

350

351 What is the large-scale circulation of Ice Giant atmospheres, and how deep does it go? Atmospheric

352 circulation, driven by both internal energy and solar heating, controls the thermal structure, radiative energy

353 balance, condensate cloud and photochemical haze characteristics, and meteorology. Unlike Jupiter and

354 Saturn, observations from Voyager, space telescopes, and ground-based observatories have revealed mid-

355 latitude upwelling (where most of the vigorous storms and coolest temperatures are found) and sinking

356 motions at the equator and poles (e.g., Conrath et al., 1998, Fletcher et al., 2020). This is superimposed onto

357 polar depletions in several key cloud-forming volatiles: methane (from reflection spectroscopy, Sromovsky

358 et al., 2014; Karkoschka et al., 2011); hydrogen sulphide (from near-IR and microwave spectroscopy, de Pater

359 et al., 1991; Hofstadter & Butler, 2003; Irwin et al., 2018); and potentially ammonia (from microwave

16 Fletcher et al., Ice Giant Systems

360 imaging). Do these contrasts imply circulation patterns extending to great depths (de Pater et al., 2014), or

361 are they restricted in vertically-thin layers (Gierasch et al., 1987)? Recent re-analysis of the gravity fields

362 measured by Voyager (Kaspi et al., 2013) suggests that zonal flows are restricted to the outermost ~1000 km

363 of their radii, indicating relatively shallow weather layers overlying the deep and mysterious water-rich

364 interiors. The circulation of the stratosphere is almost entirely unknown on both worlds, due to the challenge

365 of observing weak emissions from hydrocarbons in the mid-infrared (e.g., Orton et al., 2014; Roman et al.,

366 2020). Either way, observations of Uranus and Neptune will have stark implications for atmospheric

367 circulation on intermediate-sized planets with strong chemical enrichments and latitudinal gradients.

368

369 How does atmospheric chemistry and haze formation respond to extreme variations in sunlight and vertical

370 mixing, and the influence of external material? Methane can be transported into the stratosphere, where

371 photolysis initiates rich chemical pathways to produce a plethora of hydrocarbons (Moses et al., 2018). The

372 sluggish mixing of Uranus indicates that its photochemistry occurs in a different physical regime (higher

373 pressures) than on any other world. Furthermore, oxygen compounds from external sources (from cometary

374 impacts, infalling dust, satellite and ring material, Feuchtgruber et al., 1997) will play different photochemical

375 roles on Uranus, where the methane homopause is lower, than on Neptune (Moses & Poppe, 2017). This

376 exogenic influence can further complicate inferences of planetary formation mechanisms from

377 measurements of bulk abundances (particularly for CO). This rich atmospheric chemistry will be substantially

378 different from that on the Gas Giants, due to the weaker sunlight, the colder temperatures (changing reaction

379 rates and condensation processes, Moses et al., 2018), and the unusual ion-neutral chemistry resulting from

380 the complex magnetic field tilt and auroral processes (Dobrijevic et al. 2020). Condensation of these chemical

381 products (and water ice) can form thin haze layers observed in high-phase imaging (Rages et al., 1991), which

382 may add to the radiative warming of the stratosphere, be modulated by vertically-propagating waves, and

383 could sediment downwards to serve as condensation nuclei or aerosol coatings in the troposphere.

384 Furthermore, Uranus’ axial tilt presents an extreme test of coupled chemical and transport models, given

385 that each pole can spend decades in the shroud of winter darkness. The strength of vertical mixing may vary

17 Fletcher et al., Ice Giant Systems

386 with location, and disequilibrium tracers can be used to assess where mixing is strongest (Fletcher et al.,

387 2020; Cavalie et al., 2020). Such tracers include CO (Cavalie et al., 2017), para-H2 (Conrath et al., 1998) and

388 yet-to-be-detected phosphine (Moreno et al., 2009), arsine, germane, silane, or even tropospheric

389 hydrocarbons (ethane, acetylene). Fluorescence spectroscopy, infrared emissions and sub-mm sounding will

390 reveal the vertical, horizontal, and temporal variability of the chemical networks in the unique low-

391 temperature regimes of Uranus and Neptune.

392

393 What are the sources of energy responsible for heating the middle- and upper-atmospheres? Weak sunlight

394 alone cannot explain the high temperatures of the stratosphere (Li et al., 2018) and thermosphere (Herbert

395 et al., 1987), and this severe deficit is known as the energy crisis. Exploration of Uranus and Neptune may

396 provide a solution to this problem, potentially revealing how waves transport energy vertically from the

397 convective troposphere into the middle atmosphere, and how the currents induced by the asymmetric and

398 time-variable magnetic fields provide energy to the upper atmosphere via Joule heating. For example, the

+ 399 long-term cooling of Uranus’ thermosphere, observed via emission from H3 in the ionosphere (Melin et al.,

400 2019), appears to follow Uranus’ magnetic season, which may hint at the importance of particle precipitation

401 modulated by the magnetosphere in resolving the energy crisis. Solving this issue at Uranus or Neptune, via

402 wave observations and exploring magnetosphere-ionosphere-atmosphere coupling processes (e.g., via

403 aurora detected in the UV and infrared), will provide insights into the energetics of all planetary atmospheres.

404

405 How do planetary ionospheres enable the energy transfer that couples the atmosphere and

406 magnetosphere? In-situ radio occultations remain the only source for the vertical distribution of electron

407 density in the ionosphere (Majeed et al., 2004), a critical parameter for determining the strength of the

408 coupling between the atmosphere and the magnetosphere. The Voyager 2 occultations of both Uranus and

409 Neptune (Lindal et al, 1986, 1992) provided only two profiles for each planet, providing very poor constraints

410 on what drives the complex shape of the electron density profiles in the ionosphere, including the influx of

411 meteoritic material (Moses et al., 2017).

18 Fletcher et al., Ice Giant Systems

412

413 How do Ice Giant atmospheres change with time? In the decades since the Voyager encounters, Uranus has

414 displayed seasonal polar caps of reflective aerosols with changing winds (Sromovsky et al., 2015) and long-

415 term upper atmospheric changes (Melin et al., 2019); Neptune’s large dark anticyclones – and their

416 associated orographic clouds – have grown, drifted, and dissipated (Lebeau et al., 1998, Stratman et al., 2001,

417 Wong et al., 2018, Simon et al., 2019); and a warm south polar vortex developed and strengthened during

418 the Neptunian summer (Fletcher et al., 2014). What are the drivers for these atmospheric changes, and how

419 do they compare to the other planets? There have been suggestions that storm activity has occurred

420 episodically, potentially with a seasonal connection (de Pater et al., 2015; Sromovsky et al., 2015), but is this

421 simply driven by observational bias to their sunlit hemispheres? Mission scenarios for the early 2040s would

422 result in observations separated from those obtained by the Voyager 2 by 0.5 Uranian years and 0.25

423 Neptunian years. Orbital remote sensing over long time periods, sampling both summer and winter

424 hemispheres, could reveal the causes of atmospheric changes in a regime of extremely weak solar forcing, in

425 contrast to Jupiter and Saturn.

426

427 Summary: Investigations of dynamics, chemistry, cloud formation, atmospheric circulation, and energy

428 transport on Uranus and Neptune would sample a sizeable gap in our understanding of planetary

429 atmospheres, in an underexplored regime of weak seasonal sunlight, low temperatures, and extremes of

430 internal energy and vertical mixing.

431

432 2.3 Ice Giant Magnetospheres

433

434 What can we learn about astrophysical plasma processes by studying the highly-asymmetric, fast-rotating

435 Ice Giant magnetospheres? The off-centered, oblique and fast rotating planetary magnetic fields of Uranus

436 and Neptune give rise to magnetospheres that are governed by large scale asymmetries and rapidly evolving

437 configurations (e.g. Griton et al. 2018), with no other parallels in our Solar System. The solar wind that

19 Fletcher et al., Ice Giant Systems

438 impinges upon these two magnetospheres attains Mach numbers significantly larger than those found at

439 Earth, Jupiter, and Saturn, adding further to their uniqueness (Masters et al. 2018). Magnetospheric

440 observations should thus be a high priority in the exploration of the Ice Giants because they extend the

441 sampling of the vast parameter space that controls the structure and evolution of magnetospheres, thus

442 allowing us to achieve a more universal understanding of how these systems work. Insights would also be

443 provided to astrophysical plasma processes on similar systems that are remote to us both in space and time.

444 Such may be the magnetospheres of exoplanets or even that of the Earth at times of geomagnetic field

445 reversals, when the higher order moments of the terrestrial magnetic field become significant, as currently

+ 446 seen at the Ice Giants’ (Merrill and Mcfadden, 1999). Evidence for H3 ionospheric temperature modulations

447 at Uranus due to charged particle precipitation (Melin et al. 2011; 2013) is one of many indications reminding

448 us how strong a coupling between a planet and its magnetosphere can be, and why the study of the latter

449 would be essential also for achieving a system-level understanding of the Ice Giants.

450

451 A synergy between close proximity, remote sensing, and in-situ magnetospheric measurements at the Ice

452 Giants would redefine the state-of-the-art, currently determined by the single Voyager-2 flyby

453 measurements, and limited Earth-based auroral observations. Key questions that would guide such

454 observations are listed below:

455

456 Is there an equilibrium state of the Ice Giant magnetospheres? Voyager-2 spent only a few planetary

457 rotations within Uranus’ and Neptune’s magnetopauses, such that it was challenging to establish a nominal

458 configuration of their magnetospheres, their constituent particle populations, supporting current systems.

459 Furthermore, it is unclear whether the observations of this dynamic system represent any kind of steady

460 state, as ongoing magnetospheric reconfigurations were observed throughout each flyby, owing to the large

461 dipole tilts at the Ice Giants and their ~16 and ~17-hour rotation periods. The extent to which the two

462 magnetospheres are modified by internal plasma sources is also poorly constrained; Uranus’ magnetosphere

463 for instance was observed to be devoid of any appreciable cold plasma populations (McNutt et al., 1987).

20 Fletcher et al., Ice Giant Systems

464 The presence of strong electron radiation belts (Mauk et al. 2010), seems contradictory to the absence of a

465 dense, seed population at plasma energies, or could hint an efficient local acceleration process by intense

466 wave-fields (Scarf et al. 1987). A strong proton radiation belt driven by Galactic Cosmic Ray (GCR)-planet

467 collisions may reside closer to Uranus or Neptune than Voyager-2 reached (e.g. Stone et al. 1986), given that

468 Earth and Saturn, which are known to sustain such belts, are exposed to a considerably lower GCR influx than

469 the Ice Giants (Buchvarova and Belinov, 2009). Ion composition and UV aurora measurements hint that Triton

470 could be a major source of plasma in Neptune’s magnetosphere (Broadfoot et al., 1989, Richardson &

471 McNutt, 1990), although questions remain as to the effects of coupling between the magnetosphere and the

472 moon’s atmosphere and ionosphere in establishing the plasma source (Hoogeveen & Cloutier, 1996). The

473 magnetotails of both planets are expected to have very different structures to those seen at other

474 magnetized planets (Figure 4), with strong helical magnetic field components (Cowley, 2013; Griton and

475 Pantellini 2020), that may lead to a similarly helical topology of reconnection sites across the magnetospheric

476 current sheet (Griton et al. 2018). Whether the overall magnetospheric configuration is controlled more by

477 current sheet reconnection or a viscous interaction along the magnetopause (Masters et al. 2018) is also

478 unknown. Finally, measuring average escape rates of ionospheric plasma would offer further insights on

479 whether planetary magnetic fields protect planetary atmospheres from solar wind erosion (Wei et al. 2014,

480 Gunell et al., 2018). For such dynamic systems, long-term observations that average out rotational effects

481 and transients (e.g. Selesnick, 1988) are essential for achieving closure to all the aforementioned questions.

482

483 How do the Ice Giant magnetospheres evolve dynamically? The large tilts of the Ice Giant magnetic fields

484 relative to their planetary spin-axes hint that large-scale reconfigurations at diurnal time scales are

485 dominating short-term dynamics, a view supported by MHD simulations (Griton and Pantellini 2020; Griton

486 et al. 2018; Cao and Paty 2017; Mejnertsen et al. 2016). The rate of magnetic reconnection, for instance, is

487 predicted to vary strongly with rotation, and so is the rate of matter and energy transfer into the

488 magnetosphere and eventually upper atmosphere (Masters et al., 2014). Simulations do not capture how

489 such variations impact regions as the radiation belts, which would typically require a stable environment to

21 Fletcher et al., Ice Giant Systems

490 accumulate the observed, high fluxes of energetic particles (Mauk et al. 2010). A strong diurnal variability

491 may also affect the space weather conditions at the orbits of the Ice Giant moons, regulating the interactions

492 between the charged particle populations, their surfaces and exospheres (Plainaki et al., 2016; Arridge et al.,

493 2014) through processes like surface sputtering, ion pick up, and charge exchange, which may also feed the

494 magnetosphere with neutrals and low energy, heavy ion plasma (Lammer 1995). In the time-frame

495 considered here, the exploration of Neptune could provide an opportunity to study a “pole-on”

496 magnetosphere at certain rotational phases.

497

498 Additional sources of variability can be solar wind driven, such as Corotating Interaction Regions (CIRs) and

499 Interplanetary Coronal Mass Ejections (ICMEs) (Lamy et al. 2017). By the time they reach Uranus and

500 Neptune, ICMEs expand and coalesce and attain a quasi-steady radial width of ~2-3 AU (Wang and

501 Richardson, 2004) that could result in active magnetospheric episodes with week-long durations. With Triton

502 being a potential active source, Neptune’s magnetosphere may show variations at the moon’s orbital period

503 (Decker and Cheng 1994).

504

505 On longer time scales, seasonal changes of the planetary field’s orientation are most important, especially at

506 Uranus, because of its spin axis which almost lies on the ecliptic. There may be additional implications for the

507 long-term variability of the magnetosphere if magnetic field measurements reveal a secular drift of the

508 planetary field at any of the Ice Giants, in addition to providing constraints on the magnetic fields’ origin and

509 the planets’ interiors (Section 2.1).

510

511 How can we probe the Ice Giant magnetospheres through their aurorae? Aurorae form a unique diagnostic

512 of magnetospheric processes by probing the spatial distribution of active magnetospheric regions and their

513 dynamics at various timescales. Auroral emissions of Uranus and Neptune were detected by Voyager 2 at

514 ultraviolet and radio wavelengths. The Uranian UV aurora has been occasionally re-detected by the Hubble

515 Space Telescope since then. At NIR wavelengths auroral signatures remain elusive (e.g. Melin et al., 2019).

22 Fletcher et al., Ice Giant Systems

Figure 4 Typical magnetic field configuration in a terrestrial-like, solar wind driven magnetosphere (top) where the magnetic and spin axis are almost aligned, compared to a Uranus-like magnetospheric configuration (bottom), when the magnetic axis is 90° away from the spin axis, during equinox. The helicoidal magnetotail structure develops due to the planet’s fast rotation.

516

517 UV aurora are collisionally-excited H Ly-α and H 2 band emissions from the upper atmosphere, driven by

518 precipitating energetic particles. The radical differences of the Uranian UV aurora observed across different

519 seasons were assigned to seasonal variations of the magnetosphere/solar wind interaction (Broadfoot et al.,

520 1986; Lamy et al., 2012, 2017; Cowley, 2013; Barthélémy et al., 2014). The tentative detection of Neptunian

521 UV aurora did not reveal any clear planet-satellite interaction (e.g. with Triton) (Broadfoot et al., 1989).

522 Repeated UV spectro-imaging observations are essential to probe the diversity of Ice Giant aurora, assess

523 their underlying driver and constrain magnetic field models (e.g. Herbert, 2009).

524

525 Uranus and Neptune produce powerful auroral radio emissions above the ionosphere most likely driven by

526 the Cyclotron Maser Instability as at Earth, Jupiter, and Saturn. The Uranian and Neptunian Kilometric

527 Radiation are very similar (Desch et al., 1991, Zarka et al., 1995). They include (i) bursty emissions (lasting for

528 <10min) reminiscent of those from other planetary magnetospheres. A yet-to-be-identified time-stationary

529 source of free energy, able to operate in strongly variable magnetospheres, is thought to drive (ii) smoother

530 emissions (lasting for hours) that are unique in our solar system (Farrell, 1992). Long-term remote and in-situ

531 radio measurements are crucial to understand the generation of all types of Ice Giant radio emissions, to

23 Fletcher et al., Ice Giant Systems

532 complete the baseline for the search of exoplanetary radio emissions. Long-term monitoring of auroral

533 emissions will also be essential to precisely determine the rotation periods of these worlds.

534

535 Summary: The Ice Giant magnetospheres comprise unique plasma physics laboratories, the study of which

536 would allow us to observe and put to test a variety of astrophysical plasma processes that cannot be resolved

537 under the conditions that prevail at the terrestrial planets and at the Gas Giants. The strong planet-

538 magnetosphere links that exist further attest to the exploration of the magnetospheres as a key ingredient

539 of the Ice Giant systems.

540

541 2.4 Ice Giant Satellites – Natural & Captured

542

543 What can a comparison of Uranus’ natural satellites, the captured “Ocean World” Triton, and the myriad ice-

544 rich bodies in the Solar System, reveal about the drivers of active geology and potential habitability on icy

545 satellites? The satellite systems of Uranus and Neptune offer very different insights into moon formation

546 and evolution; they are microcosms of the larger formation and evolution of planetary systems. The

547 Neptunian system is dominated by the ‘cuckoo-like’ arrival of Triton (i.e., severely disrupting any primordial

548 Neptunian satellite system), which would be the largest known dwarf planet in the Kuiper belt if it were still

549 only orbiting the Sun. Neptune’s remaining satellites may not be primordial, given the degree of system

550 disruption generated by Triton’s arrival. In contrast, Uranus’s satellite system appears to be relatively

551 undisturbed since its formation — a highly surprising situation given that Uranus has the most severe axial

552 tilt of any planet, implying a dramatic collisional event in its past. Thus, these two satellite systems offer

553 laboratories for understanding the key planetary processes of formation, capture and collision.

554

555 What can the geological diversity of the large icy satellites of Uranus reveal about the formation and

556 continued evolution of primordial satellite systems? The five largest moons of Uranus (Miranda, Ariel,

557 Umbriel, Titania, Oberon) are comparable in sizes and orbital configurations to the medium-sized moons of

24 Fletcher et al., Ice Giant Systems

558 Saturn. However, they have higher mean densities, about 1.5 g/cm3 on average, and have different insolation

559 patterns: their poles are directed towards the Sun for decades at a time, due to the large axial tilt of Uranus.

560 The surfaces of Uranus’s five mid-sized moons exhibit extreme geologic diversity, demonstrating a complex

561 and varied history (Figure 1). On Ariel and Miranda, signs of endogenic resurfacing associated with tectonic

562 stress, and possible cryovolcanic processes, are apparent: these moons appear to have the youngest

563 surfaces. Geological interpretation has suffered greatly from the incomplete Voyager 2 image coverage of

564 only the southern hemisphere, and extremely limited coverage by Uranus-shine in part of the north (Stryk

565 and Stooke, 2008). Apart from a very limited set of images with good resolution at Miranda, revealing

566 fascinatingly complex tectonic history and possible re-formation of the moon (Figure 5), most images were

567 acquired at low to medium resolution, only allowing characterisation of the main geological units (e.g., Croft

568 and Soderblom, 1991) and strongly limiting any surface dating from the crater-size frequency distribution

569 (e.g. Plescia, 1987a, Plescia, 1987b). High-resolution images of these moons, combined with spectral data,

570 will reveal essential information on the tectonic and cryovolcanic processes and the relative ages of the

571 different geological units, via crater statistics and sputtering processes. Comparison with Saturn’s inner

572 moons system will allow us to identify key drivers in the formation and evolution of compact multiplanetary

573 systems.

574

575 What was the influence of tidal interaction and internal melting on shaping the Uranian worlds, and could

576 internal water oceans still exist? As in the Jovian and Saturnian systems, tidal interaction is likely to have

577 played a key role in the evolution of the Uranian satellite system. Intense tidal heating during sporadic

578 passages through resonances is expected to have induced internal melting in some of the icy moons

579 (Tittemore and Wisdom 1990). Such tidally-induced melting events, comparable to those expected on

580 Enceladus (e.g. Běhounková et al. 2012), may have triggered the geological activity that led to the late

581 resurfacing of Ariel and possibly transient hydrothermal activity. The two largest (>1500 km diameter)

582 moons, Titania and Oberon, may still harbour liquid water oceans between their outer ice shells and inner

583 rocky cores – remnants of past melting events. Comparative study of the static and time-variable gravity field

25 Fletcher et al., Ice Giant Systems

584 of the Uranian and Saturnian moons, once well-characterized, will constrain the likelihood and duration of

585 internal melting events, essential to characterize their astrobiological potential. Complete spacecraft

586 mapping of their surfaces could reveal recent endogenic activity.

587

588 Accurate radio tracking and astrometric measurements can also be used to quantify the influence of tidal

589 interactions in the system at present, providing fundamental constraints on the dissipation factor of Uranus

590 itself (Lainey, 2008). Gravity and magnetic measurements, combined with global shape data, will greatly

591 improve the models of the satellites' interiors, providing fundamental constraints on their bulk composition

592 (density) and evolution (mean moment of inertia). Understanding their ice-to-rock ratio and internal

593 structure will enable us to understand if Uranus' natural satellite system are the original population of bodies

594 that formed around the planet, or if they were disrupted.

595

596 What is the chemical composition of the surfaces of the Uranian moons? The albedos of the major Uranian

597 moons, considerably lower than those of Saturn’s moons (except Phoebe and Iapetus’s dark hemisphere),

598 reveal that their surfaces are characterized by a mixture of H2O ice and a visually dark and spectrally bland

599 material that is possibly carbonaceous in origin (Brown and Cruikshank, 1983). Pure CO2 ice is concentrated

600 on the trailing hemispheres of Ariel, Umbriel and Titania (Grundy et al., 2006), and it decreases in abundance

601 with increasing semimajor axis (Grundy et al., 2006; Cartwright et al., 2015), as opposed to what is observed

602 in the Saturnian system. At Uranus’ distance from the Sun, CO2 ice should be removed on timescales shorter

603 than the age of the Solar System, so the detected CO2 ice may be actively produced (Cartwright et al., 2015).

604

605 The pattern of spectrally red material on the major moons will reveal their interaction with dust from the

606 decaying orbits of the irregular satellites. Spectrally red material has been detected primarily on the leading

607 hemispheres of Titania and Oberon. H2O ice bands are stronger on the leading hemispheres of the classical

608 satellites, and the leading/trailing asymmetry in H2O ice band strengths decreases with distance from Uranus.

609 Spectral mapping of the distribution of red material and trends in H2O ice band strengths across the satellites

26 Fletcher et al., Ice Giant Systems

Figure 5 Best-resolution (roughly ~1 km/px) imagery of terrains seen in the moons of Uranus (top row) and Neptune (lower row) by Voyager 2. At Uranus, Umbriel and Titania are highly cratered with some major faults; Miranda displays spectacular and massive tectonic features; Ariel’s filled-fissured surface is suggestive of late cryovolcanic activity. At Neptune, Proteus has a surface suggestive of Saturn’s dust-rich moon Helene. The dwarf-planet-sized Triton has both the sublimation-related “cantaloupe terrain” (left) and active nitrogen gas geysers in the south polar terrains (right), with deposited dust visible as dark streaks. Credit: NASA/JPL-Caltech/Ted Stryk/collage M. Bannister. 610 and rings can map out infalling dust from Uranus's inward-migrating irregular satellites (Cartwright et al.,

611 2018), similar to what is observed in the Saturnian system on Phoebe/Iapetus (e.g., Tosi et al., 2010), and

612 could reveal how coupling with the Uranus plasma and dust environment influence their surface evolution.

613

614 Does Triton currently harbour a subsurface ocean and is there evidence for recent, or ongoing, active

615 exchange with its surface? Neptune’s large moon Triton, one of a rare class of Solar System bodies with a

616 substantial atmosphere and active geology, offers a unique opportunity to study a body comparable to the

617 dwarf planets of the rest of the trans-Neptunian region, but much closer. Triton shares many similarities in

618 surface and atmosphere with Pluto (Grundy et al. 2016), and both may harbour current oceans. Triton’s

619 retrograde orbit indicates it was captured, causing substantial early heating. Triton, in comparison with

620 Enceladus and Europa, will let us understand the role of tidally-induced activity on the habitability of ice-

621 covered oceans. The major discovery of plumes emanating from the southern polar cap of Triton (Soderblom

622 et al. 1990; see Figure 5) by Voyager 2, the most distant activity in the Solar System, is yet to be fully

623 understood (Smith et al. 1989).

27 Fletcher et al., Ice Giant Systems

624

625 Like Europa, Triton has a relatively young surface age of ~100 Ma (Stern and McKinnon 2000), inferred from

626 its few visible impact craters. Triton also displays a variety of distinctive curvilinear ridges and associated

627 troughs, comparable to those on Europa (Prockter et al. 2005) and especially apparent in the “cantaloupe

628 terrain”. This suggests that tidal stresses and dissipation have played an essential role in Triton’s geological

629 activity, and may be ongoing (Nimmo and Spencer 2015). Intense tidal heating following its capture could

630 have easily melted its icy interior (McKinnon et al. 1995). Its young surface suggests that Triton experienced

631 an ocean crystallization stage relatively recently (Hussmann et al. 2006), associated with enhanced surface

632 heat flux (Martin-Herrero et al. 2018). Combined magnetic, gravimetric and shape measurements from

633 multiple flybys or in orbit will allow us to detect if Triton possesses an ocean and to constrain the present-

634 day thickness of the ice shell. Correlating the derived shell structure and geological units will show if there is

635 exchange with the ocean. It is entirely unclear if the source(s) for Triton’s plumes reaches the ocean, as at

636 Enceladus.

637

638 Are seasonal changes in Triton’s tenuous atmosphere linked to specific sources and sinks on the surface,

639 including its remarkable plume activity? Triton’s surface has a range of volatile ices seen in Earth-based

640 near-infrared spectra, including N2, H2O, CO2, and CH4 (Quirico et al., 1999; Cruikshank et al., 2000; Tegler et

641 al, 2012). A 2.239 µm absorption feature suggests that CO and N2 molecules are intimately mixed in the ice

642 rather than existing as separate regions of pure CO and pure N2 deposits (Tegler et al., 2019). Mapping the

643 spatial variation of this absorption feature will constrain how the surface-atmosphere interaction affects the

644 surface composition, and more generally its climate and geologic evolution. Triton’s surface may also contain

645 complex organics from atmospheric photochemistry, like those of Pluto or Saturn’s moon Titan (e.g.

646 Krasnopolsky and Cruikshank 1995), as suggested by its yellowish areas (Thompson and Sagan 1990).

647 Identifying the organic compounds, and mapping out their correlation with recently active terrains and

648 geysers, will strongly raise the astrobiological potential of this exotic icy world.

649

28 Fletcher et al., Ice Giant Systems

650 Triton’s tenuous atmosphere is mainly molecular nitrogen, with a trace of methane and CO near the surface

651 (Broadfoot et al. 1989, Lellouch et al. 2010). Despite Triton's distance from the Sun and its cold temperatures,

652 the weak sunlight is enough to drive strong seasonal changes on its surface and atmosphere. Because CO and

653N 2 are the most volatile species on Triton, they are expected to dominate seasonal volatile transport across

654 its surface. Observation of increased CH4 partial pressure between 1989 and 2010 (Lellouch et al. 2010)

655 confirmed that Triton’s atmosphere is seasonably variable. The plumes of nitrogen gas and dust could be a

656 seasonal solar-driven process like the CO2 `spiders’ of the south polar regions of Mars, although an endogenic

657 origin is possible.

658

659 Are the smaller satellites of Neptune primordial? Voyager 2’s flyby led to the discovery of six small moons

660 inside Triton’s orbit: Naiad, Thalassa, Despina, Galatea, Larissa and Proteus. A seventh inner moon,

661 Hippocamp, has been recently discovered by HST observations orbiting between the two largest, Larissa and

662 Proteus. Almost nothing is known about these faint moons, which may post-date Triton’s capture rather than

663 being primordial. Only a new space mission could unveil basic features such as shape and surface

664 composition, shedding light on their origin.

665

666 How does an Ice Giant satellite system interact with the planets’ magnetospheres? Most of the major

667 moons of Uranus and Neptune orbit within the planets’ extensive magnetospheres (Figure 1). The tilt and

668 offset of both planets’ magnetic dipoles compared to their rotation axes mean that, unlike at Saturn, the

669 major moons at both Ice Giants experience continually-changing external magnetic fields. The potential

670 subsurface oceans of Titania, Oberon and Triton would be detectable by a spacecraft that can monitor for an

671 induced magnetic field. The moons in both systems orbit in relatively benign radiation environments, but

672 radiation belt particles could still drive sputtering processes at the inner moons’ surfaces and Triton’s

673 atmosphere. Triton could be a significant potential source of a neutral gas torus and magnetospheric plasma

674 at Neptune. Triton's ionosphere's transonic, sub-Alfvenic regime (Neubauer 1990; Strobel et al. 1990) may

675 generate an auroral spot in Neptune's upper atmosphere. No such interaction is anticipated at Uranus. Red

29 Fletcher et al., Ice Giant Systems

676 aurorae may also be present on Triton from N2 emission, providing valuable insights into Triton’s interaction

677 with its space environment.

678

679 Summary: Exploring Uranus’ and Neptune’s natural satellites, as well as Neptune’s captured moon Triton,

680 will reveal how ocean worlds may form and remain active, defining the extent of the habitable zone in the

681 Solar System. Both satellite systems are equally compelling for future orbital exploration.

682

683 2.5 Ice Giant Ring Systems

684

685 What processes shape the rings of Ice Giants, and why do they differ from the extensive rings of Saturn?

686 Uranus and Neptune both possess a complex system of rings and satellites (Figure 6, see also de Pater et al.,

687 2018, Nicholson et al., 2018). The rings exhibit narrow and dense ringlets, as well as fainter but broader dust

688 components. The moons can confine the rings gravitationally, and may also serve as sources and sinks for

689 ring material. Observations indicate rapid variability in the Uranian and Neptunian rings within decades. A

690 mission to the Ice Giants, with a dedicated suite of instruments, can answer fundamental questions on the

691 formation and evolution of the ring systems and the planets themselves:

692

693 What is the origin of the solar system ring systems, and why are they so different? The origin of the giant

694 planets’ rings is one of the unsolved mysteries in planetary science. Whereas all four giant planets do have

695 rings, their diversity of structure and composition argues for different formation scenarios (Charnoz et al.,

696 2018). It was hypothesized that the very massive Saturnian rings formed more than 3 Gyrs ago through tidal

697 destruction of a moon, or of a body on a path traversing the system. However, recent Cassini measurements

698 (Zhang et al., 2017, Kempf et al., 2018, Waite et al., 2018, Iess et al., 2019) argue for a younger ring age. In

699 contrast, Uranus’ and Neptune’s rings are far less massive and they have a different structure. Compared to

700 Saturn, their rings’ albedo is much lower, favouring a parent body that was a mixture of ice and dark material

701 (silicates and possibly organics). For instance, it was suggested (Colwell and Esposito, 1992, 1993) that these

30 Fletcher et al., Ice Giant Systems

702 two ring systems could result from the periodic destruction of moonlets though meteoroid bombardment,

703 in which case most of the ringlets would be only transient structures, currently in the process of re-accretion

704 to satellites. Among the Uranian dust rings the µ ring has a distinct blue spectral slope (de Pater et al. 2006).

705 In that regard it is similar to Saturn’s E ring, for which the blue slope results from the narrow size distribution

706 of its grains, formed in the cryo-volcanic activity of the moon Enceladus. Although the moon Mab is

707 embedded in the µ ring (Showalter et al, 2006) it appears much too small (12km radius) to be volcanically

708 active and create in this way the dust that forms the ring. Other dust rings of Uranus exhibit a red spectral

709 slope (de Pater et al. 2006), suggestive of dusty material released in micrometeoroid impacts on atmosphere-

710 less moons and the origin for different appearance of the µ ring is unknown. Recent ground-based

711 observations of thermal emission from the Uranian ring system (Molter et al., 2019) are consistent with the

712 idea that the rings are made up of larger particles, without micron-sized dust, and that their temperatures

713 result from low thermal inertia and/or slow rotation of the particles. Clearly, more data is needed to

714 ultimately settle the question of the origin and nature of Uranus’ and Neptune’s rings.

715

716 How do the ring-moon systems evolve? A variety of processes govern the evolution of planetary rings, many

717 of which are also fundamental to other cosmic disks. Important for the rings are viscous transport, ring-

718 satellite interactions, the self-gravity, as well as meteoroid bombardment and cometary impacts. These

719 processes may induce rapid evolution on timescales much shorter than the rings’ age. For instance, the

720 Neptune ring arcs were initially interpreted to be confined by resonances with the moon Galatea. However,

721 it was found that the arcs actually move away from the corresponding corotation sites (Dumas et al., 1999,

722 Namouni & Porco 2002) and that they evolve rapidly (de Pater et al., 2005). Also, for the Uranian rings

723 significant changes since the Voyager epoch are observed (de Pater et al., 2006, 2007). In the Uranus and

724 Neptune systems the role the moons play in sculpting the rings is even stronger than for Saturn’s rings.

725 Moreover, depending on their composition, the Uranus and Neptune ring systems may even extend beyond

726 the planet’s Roche Limit (Tiscareno et al., 2013). This implies that their path of evolution is different from the

727 Saturnian rings, inducing changes on more rapid timescales. Some of the edges of the Uranian rings are

31 Fletcher et al., Ice Giant Systems

Figure 6 Panel (a): Schematic diagram of the ring moon systems of Uranus, Neptune, and Saturn for comparison. The Roche radius (for icy ring particles) is marked as a dashed line. Panel (b) shows a combination of two Voyager images of the Uranus rings, taken at low (upper part) and high (lower part) phase angle (figure from Nicholson et al., 2018). At high phase angle the dusty components of the ring system stand out. Panel (c) shows a Voyager image of the rings of Neptune (Credit: NASA/JPL).

728 clearly confined by known satellites and others might be confined by yet undetected moons. Alternatively,

729 there might be different processes of confinement at work, in a similar manner as for narrow rings of Saturn,

730 for which shepherding moons are absent. Some of the dense rings of Uranus show sub-structure the origin

731 of which is unclear. Spacecraft investigation will answer the question if the structure is induced by resonant

732 interaction with moons, or if it represents intrinsic modes arising from instability and self-gravity of the rings.

733

734 What is the ring composition? The rings (as the moons) very likely consist of material that was present at

735 the location in the solar system where the planets themselves have formed. Therefore, the investigation of

736 the composition of the rings, while being interesting in its own right, may also shed light on models of planet

737 formation and migration in the solar system.

738

32 Fletcher et al., Ice Giant Systems

739 Imaging at high and intermediate phase angles will constrain the shapes and properties of known dust rings

740 and has the potential to discover new rings. Multicolour imaging at a range of phase angles will determine

741 the size distribution of the grains that form these rings. Imaging at low phase angles will probe the dense

742 rings and allow for a comprehensive search and discovery of yet unseen satellites that serve as sources for

743 ring material and that interact dynamically with the rings. Stellar occultations performed with a high-speed

744 photometer, or radio occultations, will determine the precise optical depths of the denser rings and resolve

745 their fine sub-structure (French et al., 1991). An IR spectrometer will determine the composition of the dense

746 rings. In a complementary manner a dust detector will directly determine the composition of the grains

747 forming the low optical depth dust rings (Postberg et al., 2009) and of particles lifted from the dense rings

748 (Hsu et al., 2018). A dust detector will also measure the flux and composition of interplanetary particles in

749 the outer solar system, which is a poorly known quantity of high importance for the origin and evolution of

750 the rings.

751

752 Summary: Ice Giant rings appear to be fundamentally different to those of Saturn, such that their origin,

753 evolution, composition and gravitational relationships with the icy satellites should provide key new insights

754 into the forces shaping ring systems surrounding planetary bodies.

755 3. Ice Giant Science in Context

756

757 The scientific themes highlighted in Section 2 span multiple disciplines within planetary science, and address

758 questions that touch on issues across astronomy. In this section we review how an Ice Giant mission must

759 be considered in the context of other fields and technologies that will be developing in the coming decades.

760

761 3.1 Astronomical Observatories

762

763 An Ice Giant System mission would be operating in the context of world-leading new facilities on or near

764 Earth. The 2020s will see the launch of the James Webb Space Telescope, able to provide spectral maps of 33 Fletcher et al., Ice Giant Systems

765 both Ice Giants from 1-30 µm but at a moderate spatial resolution and with limited temporal coverage. Earth-

766 based observatories in the 8-10 m class provide better spatial resolution at the expense of telluric

767 obscuration. A successor to the Hubble Space Telescope, which has been the key provider of visible and UV

768 observations of the Ice Giants, has yet to become a reality, but could be operating in the 2030s. And although

769 the next generation of Earth-based observatories in the 30+m class (the ELT, TMT, GMT) will provide exquisite

770 spatial resolution, this will remain limited to atmospheric and ionospheric investigations of the Earth-facing

771 hemisphere (with some disc-averaged spectroscopy of the satellites), leaving a multitude of fundamental

772 questions unanswered.

773

774 Additionally, distant remote sensing can only study phenomena that alter the emergent spectrum –

775 meteorology, seasonal variations, and ionospheric emissions (auroral and non-auroral). This limits our

776 understanding of the underlying mechanisms and means that ground- and space-based telescopes only serve

777 a narrow subset of the Ice Giant community, and cannot address the wide-ranging goals described in Section

778 2. Thus, there is no substitute for orbital exploration of one or both of these Ice Giant systems (alongside in

779 situ sampling of their atmospheres), but we envisage that these space missions will work in synergy with the

780 ground-based astronomy community, following successful examples of Galileo, Cassini, Juno and, ultimately,

781 JUICE and Europa Clipper. Support from Earth-based observatories, either on the ground or in space, will be

782 used to establish a temporal baseline for atmospheric changes (e.g., tracking storms), provide global context

783 for close-in observations from the orbiters, and plug any gaps in spectral coverage or spectral resolution in

784 the orbiter payload.

785

786 3.2 Heliophysics Connection

787

788 Missions to explore the Ice Giant Systems also resonate with the heliophysics community, as detailed

789 exploration of an oblique rotator can inform a universal model of magnetospheres. The panel on solar-wind

34 Fletcher et al., Ice Giant Systems

790 magnetosphere interactions of the 2013 heliophysics decadal survey5 identified how the magnetospheres of

791 Uranus and Neptune are fundamentally different from others in our Solar System, and sought to ensure that

792 magnetic field instruments would be guaranteed a place on outer planet missions, with a strong

793 recommendation being that NASA’s heliophysics and planetary divisions partner on a Uranus orbital mission.

794 They describe how Uranus offers an example of solar wind/magnetosphere interactions under strongly

795 changing orientations over diurnal timescales. Depending on the season, the effects of solar wind-

796 magnetosphere interaction vary dramatically over the course of each day.

797

798 There is also a need to understand how the solar wind evolves beyond 10 AU (Saturn orbit), as the states of

799 solar structures travelling within the solar wind (solar wind pressure pulses) are largely unknown due to the

800 lack of observations at such large heliocentric distances (Witasse et al., 2017). The long cruise duration of a

801 mission to Uranus or Neptune provides an excellent opportunity for both heliophysics and Ice Giant

802 communities if a space weather-monitoring package that includes a magnetometer, a solar wind analyser,

803 and a radiation monitor operates continuously during the cruise phase, to understand how conditions vary

804 out to 20-30 AU over a prolonged lifetime. Moreover, the complexity of the interactions of Uranus’s

805 magnetosphere with the solar wind provides an ideal testbed for the theoretical understanding of planetary

806 interactions with the solar wind, significantly expanding the parameter range over which scientists can study

807 magnetospheric structure and dynamics. The potential discoveries from its dynamo generation and its

808 variability stand to open new chapters in comparative planetary magnetospheres and interiors.

809

810 3.3 Exoplanet & Brown Dwarf Connection

811

812 The Ice Giant System mission will occur during an explosion in our understanding of planets beyond our Solar

813 System, through ESA’s Cosmic Vision missions Plato, Euclid, and ARIEL; through missions with international

5 https://www.nap.edu/catalog/13060/solar-and-space-physics-a-science-for-a-technological-society

35 Fletcher et al., Ice Giant Systems

Figure 7 Known transiting exoplanets in 2016, from the Kepler mission, showing sub-Neptunes as the most common planetary radius in the current census. Credit: NASA/Ames

814 partners like JWST and TESS; and through next-generation observatories like WFIRST, LUVOIR, Origins, and

815 HabEx. The physical and chemical processes at work within our own Solar System serve as the key foundation

816 for our understanding of those distant and unresolved worlds. Our Solar System provides the only laboratory

817 in which we can perform in-situ experiments to understand exoplanet formation, composition, structure,

818 dynamos, systems and magnetospheres. After the several highly-successful missions to the Gas Giants (and

819 the upcoming ESA/JUICE mission), dedicated exploration of the Ice Giants would place those discoveries into

820 a broader, solar-system context.

821

822 Planet Statistics: Uranus and Neptune represent our closest and best examples of a class of planets

823 intermediate in mass and size between the larger, hydrogen-helium-enriched gas giants, and the rocky

824 terrestrial worlds. Indeed, Neptune- and sub-Neptune-size worlds have emerged as commonplace in our

825 ever-expanding census of exoplanets (Figure 7). Neptune-size planets are among the most common classes

826 of exoplanet in our galaxy (Fulton et al., 2018). Fressin et al. (2013) suggest that this category of planets can

827 be found around 3-31% of sun-like stars, and Petigura et al (2017) suggests that sub-Neptunes remains the

828 most common category within Kepler’s survey. Based on statistics from the Kepler mission it is predicted that

829 TESS will detect over 1500 sub-Neptunes (2-4 RE) over the mission (Barclay et al. 2018). Microlensing surveys

830 (e.g., WFIRST, Penny et al., 2019) will also be more sensitive to lower-mass planets on wide orbits, and could

831 reveal new insights into extrasolar Ice Giants ahead of a mission to Uranus or Neptune. Given these planetary

832 occurrence rates, the exploration of bulk composition and interiors of our Ice Giants would provide strong

36 Fletcher et al., Ice Giant Systems

833 constraints on the most common outcome for planetary formation. However, we emphasise that being

834 Neptune-sized does not necessarily imply being Neptune-like, as a plethora of additional parameters come

835 into play to shape the environmental conditions on these worlds.

836

837 Atmospheric Insights: Although we are currently unable to resolve spatial contrasts on exoplanets and brown

838 dwarfs, comparisons of dayside (eclipse) and nightside (transit) spectra suggest the presence of powerful

839 winds on some targets that are responsible for redistribution of energy. Brightness variations as Brown

840 Dwarfs rotate suggest patchy cloud structures, but their rapid evolution was only understood when

841 compared with long-cadence Neptune observations (Apai et al., 2013, Stauffer et al. 2016, Simon et al. 2016).

842 Changes in exoplanet transit spectra with temperature indicate complex cloud condensation processes

843 (Morley et al., 2012; Wakeford et al., 2017). Atmospheric dynamics, chemistry, and cloud formation all vary

844 as a function of planetary age, distance from the host star, and the bulk enrichments of chemical species.

845 The smaller radii of Uranus and Neptune, compared to the larger gas giants, implies atmospheric processes

846 (zonal banding, storms, vortices) at work in a different region of dynamical parameter space, and one which

847 is unavailable elsewhere in our Solar System. Detailed exploration of our Ice Giants, in comparison to the

848 existing studies of the Gas Giants, would allow us to unravel these competing and complex phenomena

849 shaping the atmospheres of exoplanets and brown dwarfs. Importantly, measurements of Ice Giant

850 composition and dynamics can be directly compared to exoplanet and brown dwarf observations, placing

851 their formation location, age, and evolution into context, and vice versa. However, this cannot be done

852 without a detailed comparative dataset from our Solar System Ice Giants.

853

854 Magnetospheric Insights: Rymer et al. (2018) explore the importance of Ice Giant interactions with the wider

855 magnetic environment as a testbed for understanding exoplanets. Eccentric and complex orbital

856 characteristics appear commonplace beyond our Solar System, and Uranus is one of the only places where

857 radio emission, magnetospheric transport and diffusion resulting from a complex magnetospheric

858 orientation can be explored. The stability and strength of the Uranian radiation belts could also guide the

37 Fletcher et al., Ice Giant Systems

859 search for exoplanets with radiation belts. Finally, understanding the dynamos of Uranus and Neptune would

860 drastically improve our predictions of magnetic field strengths and exoplanet dynamo morphologies. Each

861 of these insights will be vital as exoplanetary science moves into an era of characterisation of atmospheric

862 composition, dynamics, clouds, and auroral/radio emissions.

863 4. Ice Giant Missions

864 4.1 Architectures: The Case for Orbiters

865

866 The scientific themes of Ice Giant System missions (summarised in Figure 9) are broad and challenging to

867 capture within a single mission architecture, but recent efforts by both ESA (e.g., the 2018-19 studies with

868 the Concurrent Design Facility for an M*-class mission6) and NASA (e.g., the 2017 Science Study Team report,

869 Hofstadter et al., 2019)7 have explored strategies to achieve many of the goals in Section 2. The joint NASA-

870 ESA science study team provided a detailed investigation of various combinations of flyby missions, orbiters,

871 multiple sub-satellites from a core spacecraft, satellite landers, and atmospheric probes. Strategies to

872 explore both Ice Giants with dual spacecraft have also been proposed (Turrini et al., 2014; Simon et al., 2018).

873 It was widely recognised that a flyby mission like Voyager, without any additional components like an entry

874 probe, was deemed to provide the lowest science return for the Ice Giants themselves, despite their lower

875 cost point. Without in situ measurements, and by providing only brief snapshots of the evolving atmospheres

876 and magnetospheres, and limited coverage of the satellites and rings, a flyby could not deliver on the highest-

877 priority science goals for an Ice Giant mission. Targeting Triton as a flyby, or the inclusion of an entry probe,

878 would increase the scientific reach, but would still prove inadequate for whole-system science. The study

879 found that an Ice Giant orbiter for either the Uranus or Neptune systems, alongside an in situ atmospheric

880 probe, would provide an unprecedented leap in our understanding of these enigmatic worlds. An orbiter

6 http://sci.esa.int/future-missions-department/61307-cdf-study-report-ice-giants/ 7 https://www.lpi.usra.edu/icegiants/

38 Fletcher et al., Ice Giant Systems

881 would maximise the time spent in the system to conduct science of interest to the entire planetary

882 community.

883

884 In our 2019 submission to ESA’s call for ideas to shape the planning of space science missions in the coming

885 decades (known as Voyage 2050), we therefore proposed that an orbital mission to an Ice Giant should be

886 considered as a cornerstone of ESA’s Voyage 2050 programme, if not already initiated with our international

887 partners in the coming decade. An ESA orbital mission, powered by radioisotope thermoelectric generators,

888 should be studied as an L-class mission to capitalise on the wealth of European experience of the Cassini and

889 JUICE missions. Alternatively, an M-class Ice Giant budget would allow a crucial contribution to an orbital

890 mission led by our international partners. The mass and mission requirements associated with additional

891 components, such as satellite landers, in situ probes, or secondary small satellites, must be tensioned against

892 the capabilities of the core payload, and the capability of the launch vehicle and propulsion. In all of these

893 cases, a formal study of the requirements and capabilities is necessary to mature the concept.

894

895 Payload Considerations: The 2017 NASA study found that payload masses of 90-150 kg could deliver

896 significant scientific return for a flagship-class mission, whilst the 2018 ESA CDF study identified 100 kg as a

897 realistic payload mass for a European orbiter. Different studies have resulted in different prioritisations for

898 instrumentation, but produced suites of orbiter experiments in common categories. Multi-spectral remote

899 sensing is required, using both imaging and spectroscopy, spanning the UV, visible, near-IR (e.g.,

900 atmosphere/surface reflectivity, dynamics; auroral observations), mid-IR, sub-mm, to centimetre

901 wavelengths (e.g., thermal emission and energy balance, atmospheric circulation). Such remote sensing is

902 also a requirement for characterising any atmospheric probe entry sites, or satellite lander sites. Direct

903 sensing of the magnetospheric and plasma environment would be accomplished via magnetometers, dust

904 detectors, plasma instruments, radio wave detectors, and potentially mass spectrometers. Radio science

905 would provide opportunities for interior sounding and neutral/ionospheric occultation studies. The provision

906 of such instruments would capitalise on European heritage on Cassini, JUICE, Rosetta, Venus/Mars Express,

39 Fletcher et al., Ice Giant Systems

907 and BepiColombo, but at the same time recognising the need to develop smaller, lighter, and less

908 power/data-intensive instruments, raising and maturing their technological readiness.

909

910 Orbit Considerations: Orbital missions to both Uranus and Neptune depend crucially on the chemical fuel

911 required for orbit insertion, which determines the deliverable mass. The potential use of aerocapture, using

912 atmospheric drag to slow down the spacecraft, permits larger payloads and faster trip times at the expense

913 of increased risk, which needs significant further study. Mission requirements and orbital geometries will

914 determine the inclination of orbital insertion – high geographical latitudes would benefit some atmospheric,

915 rings, and magnetospheric science, but satellite gravity assists and subsequent trajectory corrections would

916 be needed for exploration of the satellites, rings and atmosphere from a low-inclination orbit. High

917 inclinations are easier to achieve at Uranus, although Triton can be used to drive a satellite tour at Neptune.

918 The delivery and telemetry for an atmospheric entry probe must also be considered in an orbital tour design

919 (e.g., Simon et al., 2020). We also propose that distinct phases of an orbital tour be considered, balancing

920 moderate orbital distances (for remote sensing, outer magnetosphere science, and a satellite tour) with

921 close-in final orbits (for gravity science and inner magnetosphere), following the example of Cassini and Juno.

922 Multiple close flybys of major satellites are desirable to map their interiors, surfaces, and atmospheres via a

923 variety of techniques. Finally, multi-year orbital tours (at least ~3 years) would maximise our time in the

924 system, permitting the study of atmospheric and magnetospheric changes over longer time periods. The

925 2018 ESA CDF study confirmed the feasibility of orbital tours satisfying these scientific requirements.

926

927 Ring Hazards: The 2017 NASA-ESA report highlighted potentially unknown ring-plane hazards as a topic for

928 future study. Orbit insertion should be as close to the planet as possible to reduce the required fuel, but the

929 properties of Ice Giant rings remain poorly constrained. Potential options to mitigate this risk include: having

930 the insertion be further out (requiring more fuel); fly through the ring plane at an altitude where atmospheric

931 drag is high enough to reduce the number of particles, but not enough to adversely affect the spacecraft; use

932 a pathfinder spacecraft to measure the particle density ahead of time; or use Earth-based observations to

40 Fletcher et al., Ice Giant Systems

933 constrain the upper atmosphere/ring hazard. Detailed calculations on the location of this safe zone are

934 required.

935

936 4.2 Timeliness and Launch Opportunities

937

938 Trajectories to reach the Ice Giants depend on a number of factors: the use of chemical and/or solar-electric

939 propulsion (SEP) technologies; the lift capacity of the launch vehicle; the use of aerocapture/aerobraking;

940 and the need for gravity assists. The availability of Jupiter, as the largest planet, is key to optimal launch

941 trajectories, and the early 2030s offer the best opportunities. The synodic periods of Uranus and Neptune

942 with respect to Jupiter are ~13.8 and ~12.8 years, respectively, meaning that optimal Jupiter gravity-assist

943 (GA) windows occur every 13-14 years. The NASA-ESA joint study team identified chemical-propulsion

944 opportunities with a Jupiter GA in 2029-30 for Neptune, and a wider window of 2030-34 for Uranus. Such

945 windows would repeat in the 2040s, and a wider trade space (including the potential to use Saturn GA or

946 direct trajectories using the next generation of launch vehicles) should be explored. We stress that a mission

947 to an Ice Giant is feasible using conventional chemical propulsion.

948

949 Furthermore, a mission launched to Uranus or Neptune could offer significant opportunities for flybys of

950 Solar System objects en route, especially Centaurs, which are small bodies that orbit in the giant planet

951 region. This population has yet to be explored by spacecraft, but represents an important evolutionary step

952 between Kuiper Belt objects and comets. Around 300 are currently known, with 10% of these observed to

953 show cometary activity. In addition, we can expect to discover at least an order of magnitude more Centaurs

954 by the 2030s, following discoveries by the Large Synoptic Survey Telescope, increasing the probability that a

955 suitable flyby target can be found near to the trajectory to Uranus/Neptune. Some of the largest of the

956 Centaurs (~100 km scale objects) have their own ring systems, the origin of which has yet to be explained,

957 while smaller ones (1-10 km scale) could add an important ‘pre-activity’ view to better interpret data from

41 Fletcher et al., Ice Giant Systems

958 Rosetta’s exploration of a comet. The payload options described in Section 4 would be well suited to

959 characterise a Centaur during a flyby.

960

961 The launch time necessarily influences the arrival time, as depicted in Figure 8. Uranus will reach northern

962 summer solstice in 2030, and northern autumnal equinox in 2050. Voyager 2 observed near northern winter

963 solstice, meaning that the north poles of the planet and satellites were shrouded in darkness. These

964 completely unexplored northern terrains will begin to disappear into darkness again in 2050, where they will

965 remain hidden for the following 42 years (half a Uranian year).

966

967 Neptune passed northern winter solstice in 2005, and will reach northern spring equinox in 2046. After this

968 time, the southern hemispheres of the planet and satellites (most notably the plumes of Triton at high

969 southern latitudes) will sink into winter darkness, meaning that the Triton plumes – if they are indeed

970 restricted to the south – would no longer be in sunlight after ~2046, and would remain hidden for the next

971 ~82 years (half a Neptunian year).

42 Fletcher et al., Ice Giant Systems

Figure 8 Potential timeline of missions to Uranus (left) and Neptune (right), compared to the seasons on each Ice Giant. The white arrows show the approximate timescales for launch opportunities in the early 2030s, with arrival in the 2040s. 972

973 The 2028-34 launch opportunities were assessed by the joint ESA-NASA study team. Saturn GA was

974 considered but did not appear optimal for this launch window. Interplanetary flight times are 6 to 12 years

975 to Uranus, 8 to 13 years to Neptune, depending on launch year, mission architecture, and launch vehicle.

976 The greater challenge of reaching the Neptune system was reflected in their choice of detailed architecture

977 studies: five missions to Uranus (orbiters with/without probes; with/without SEP; and with different payload

978 masses), and a single orbiter and probe for Neptune. Both Uranus and Neptune were deemed equally

979 valuable as targets – Uranus standing out for its uniqueness; Neptune for the prospects of exploring Triton.

980 The choice between these two enticing destinations will ultimately be driven by launch opportunities and

981 deliverable mass to the systems.

982

43 Fletcher et al., Ice Giant Systems

983 We must capitalise on the current momentum within Europe, alongside our international partners, to make

984 use of the launch opportunities in the ~2030s. Such a mission would arrive at Uranus while we can still see

985 the totally-unexplored northern terrains, or at Neptune while we can still see the active geology of Triton.

986 Operations in the 2040s would also allow ESA to maintain Outer Solar System expertise from current and

987 future missions like JUICE. An Ice Giant explorer would therefore be active as a cornerstone of ESA’s Voyage

988 2050 strategic planning for space missions.

989

990 4.3 Mature and Developing Technologies

991

992 An ambitious mission to an Ice Giant System would largely build on existing mature technologies (e.g., see

993 the discussion of payload development in Section 4.1), but several challenges have been identified that, if

994 overcome, would optimise and enhance our first dedicated orbital mission to these worlds. Note that we

995 omit the need for ablative materials on atmospheric entry probes, which will be required for in situ science

996 (Mousis et al., 2018, Simon et al., 2020). Key areas where technology maturations are required include:

997

998 Space nuclear power: With the prospect of flying solar-powered spacecraft to 20 AU being non-viable, an

999 Ice Giant System mission must rely on radioisotope power sources, both for electricity and for spacecraft

1000 heating. In the US, existing MMRTGs (multi-mission radioisotope thermal generators), based on the decay of

1001 238Pu, will be re-designed to create eMMRTGs (“enhanced”) to increase the available specific power at the

1002 end of life, 4-5 of which were considered for the mission architectures studied in the 2017 NASA-ESA report.

1003 Previous M-class Uranus mission proposals have relied on US provision of these power sources for an ESA-

1004 led mission. However, ESA continues to pursue the development of independent power sources based on

1005 241Am (Ambrosi et al., 2019). Whilst the power density is lower than that of 238Pu, the half-life is much longer,

1006 and much of the material is available from the reprocessing of spent fuel from European nuclear reactors,

1007 extracted chemically from plutonium to a ceramic oxide form. Prototypes for both radioisotope heater units

1008 (warming the spacecraft) and thermoelectric generators (providing spacecraft power) have now been

44 Fletcher et al., Ice Giant Systems

1009 demonstrated, and development is continuing for operational use late in the next decade (Ambrosi et al.,

1010 2019). An Ice Giant System mission could benefit tremendously from this independent European power

1011 source.

1012

1013 Hardware longevity: Given the 6- to 13-year interplanetary transfer, coupled with the desire for a long

1014 orbital tour, Ice Giant orbiters must be designed to last for a long duration under a variable thermal load

1015 imposed by gravity assists from the inner to the outer solar system. This poses constraints on the reliability

1016 of parts and power sources, as well as the need to develop optimised operational plans for the long cruise

1017 phases, such as the use of hibernation modes following the example of New Horizons.

1018

1019 Telemetry/Communications: All missions to the giant planets are somewhat constrained by competition

1020 between advanced instrumentation and the reduced data rates at large distances from Earth, but the case

1021 at Uranus and Neptune is most severe. The use of Ka-band in the downlink, currently supported by both the

1022 US Deep Space Network and two out of three ESA Deep Space Antennas, allows for mitigating this issue by

1023 increasing the achievable daily data volumes, but a careful optimization of the science tour remains critical.

1024 The available power, length of the downlink window, and the need to balance science data, engineering and

1025 housekeeping telemetry, all contribute to determining the overall data rate. For example, the NASA/JPL

1026 mission study8 in 2017 suggested that use of a 35-W traveling wave tube amplifier (the current power limit

1027 for space-qualified TWTAs) and a 34-m ground-station could provide 15 kbps at Uranus, whereas this could

1028 increase to 30 kbps with a 70-W TWTA. The 2018 ESA CDF study9 suggested that the data rate of a future M-

1029 class orbital mission to Uranus could use a 100-W TWTA to deliver 94 kbps Ka-band downlink from Uranus,

1030 or 42 kbps from Neptune. Assuming 3.2 hours/day for communications, this equates to daily data volumes

1031 for science of 1.09 and 0.48 Gb, respectively. Using current 35-W TWTAs, the ESA CDF study also suggested

1032 that lower data rates of 31 and 14 kbps could be achievable at Uranus and Neptune respectively, requiring

8 https://www.lpi.usra.edu/icegiants/mission_study/Full-Report.pdf 9 https://sci.esa.int/web/future-missions-department/-/61307-cdf-study-report-ice-giants

45 Fletcher et al., Ice Giant Systems

1033 the use of longer downlink windows. Even under the most optimistic assumptions discussed above, it is clear

1034 that some data optimization strategy would still be required. We would welcome detailed studies of new

1035 communications technologies, such as optical communications, as a general enabling technology for solar

1036 system exploration. However, we recognise that achieving the required directionality of a downlink laser

1037 from beyond 5AU will be challenging.

1038

1039 Launch Vehicles: The market for launch vehicles is changing dramatically both in Europe and in the US. Ice

1040 Giant mission concepts have traditionally considered Jupiter gravity assists to provide realistic flight times

1041 and sufficient payload delivered to each system. However, we also advocate investigation of the direct-

1042 transfer trajectories enabled by the heavy-lift capacity of the next generation of launch vehicles. This may

1043 (i) obviate the need for Jupiter/Saturn GA, so allow more flexibility in launch dates; and (ii) open up the

1044 possibilities for launching multiple spacecraft that share the same faring.

1045

1046 5. Summary and Perspectives

1047

1048 This article reviews and updates the scientific rationale for a mission to an Ice Giant system, advocating that

1049 an orbital mission (alongside an atmospheric entry probe, Mousis et al., 2018) be considered as a cornerstone

1050 of ESA’s Voyage 2050 programme, working in collaboration with international partners to launch the first

1051 dedicated mission to either Uranus or Neptune. Using technologies both mature and in development, the

1052 Ice Giants community hopes to capitalise on launch opportunities in the 2030s to reach the Ice Giants. As

1053 shown in Figure 9, an Ice Giant System mission would engage a wide community, drawing expertise from a

1054 vast range of disciplines within planetary science, from surface geology to planetary interiors; from

1055 meteorology to ionospheric physics; from plasma scientists to heliophysicists. But this challenge is also

1056 interdisciplinary in nature, engaging those studying potentially similar Neptune-size objects beyond our Solar

1057 System, by revealing the properties of this underexplored class of planetary objects. As Neptune’s orbit

1058 shapes the dynamical properties of objects in the distant solar system, an Ice Giant System mission also draws 46 Fletcher et al., Ice Giant Systems

1059 in the small-bodies community investigating objects throughout the Outer Solar System, from Centaurs, to

1060 TNOs, to Kuiper Belt Objects, and contrasting these with the natural satellites of Uranus.

1061

1062 To launch a mission to Uranus and/or Neptune in the early 2030s, the international Ice Giant community

1063 needs to significantly raise the maturity of the mission concepts and instrument technological readiness. At

1064 the time of writing (late 2019), we await the outcomes of both ESA’s Voyage 2050 process, and the next US

1065 planetary decadal survey. Nevertheless, this should not delay the start of a formal science definition and

1066 study process (pre-phase A conceptual studies), so that we are ready to capitalise on any opportunities that

1067 these strategic planning surveys provide. Fully costed and technologically robust mission concepts need to

1068 be developed, studied, and ready for implementation by 2023-25 to have the potential to meet the upcoming

1069 window for Jupiter gravity assists between 2029-2034. We hope that the scientific themes and rationale

1070 identified in this review will help to guide that conceptual study process, to develop a paradigm-shifting

1071 mission that will help redefine planetary science for a generation of scientists and engineers.

1072

1073 Most importantly, the Ice Giant System mission will continue the breath-taking legacy of discovery of the

1074 Voyager, Galileo, Cassini, and Juno missions to the giant planets (in addition to the forthcoming JUICE and

1075 Europa Clipper missions). A dedicated orbiter of an Ice Giant is the next logical step in our exploration of the

1076 Solar System, completing humankind’s first reconnaissance of the eight planets. It will be those discoveries

1077 that no one expected, those mysteries that we did not anticipate, and those views that no human has

1078 previously witnessed, which will enthuse the general public, and inspire the next generation of explorers to

1079 look to the worlds of our Solar System. We urge both ESA and NASA to take up this challenge.

1080

1081

47 Fletcher et al., Ice Giant Systems

Figure 9 Left: Our last views of Uranus and Neptune from a robotic spacecraft, taken by Voyager 2 three decades ago (Credit: NASA/JPL/E. Lakdawalla). Will we see these views again before a half-century has elapsed? Right: Seven science themes for Voyage 2050 that could be addressed by an Ice Giant System mission.

1082

1083 Acknowledgements

1084

1085 Fletcher was supported by a Royal Society Research Fellowship and European Research Council Consolidator

1086 Grant (under the European Union's Horizon 2020 research and innovation programme, grant agreement No

1087 723890) at the University of Leicester. Mousis acknowledges support from CNES. Hueso was supported by

1088 the Spanish MINECO project AYA2015-65041-P (MINECO/FEDER, UE) and by Grupos Gobierno Vasco IT1366-

1089 19 from Gobierno Vasco. UK authors acknowledge support from the Science and Technology Facilities

1090 Council (STFC). The authors are grateful to Beatriz Sanchez-Cano, Colin Snodgrass, Hannah Wakeford,

1091 Marius Millot, and Radek Poleski for their comments on early drafts of this article, and to two anonymous

1092 reviewers for their helpful suggestions. We are grateful to the members of ESA’s Voyage 2050 Senior

1093 Committee for allowing us to present this review at the Voyage 2050 workshop in Madrid in October 2019

1094 (see https://www.cosmos.esa.int/web/voyage-2050/workshop-programme).

48 Fletcher et al., Ice Giant Systems

1095

1096 References

1097 Ambrosi, R. M., Williams, H., Watkinson, E. J., et al. (2019), European Radioisotope Thermoelectric

1098 Generators (RTGs) and Radioisotope Heater Units (RHUs) for Space Science and Exploration, Space Science

1099 Reviews, 215, 55, 10.1007/s11214-019-0623-9.

1100

1101 Apai, D., Radigan, J., Buenzli, E., et al. (2013), HST Spectral Mapping of L/T Transition Brown Dwarfs Reveals

1102 Cloud Thickness Variations, The Astrophysical Journal, 768, 121, 10.1088/0004-637X/768/2/121.

1103

1104 Arridge, C. S., Achilleos, N., Agarwal, J., et al. (2014), The science case for an orbital mission to Uranus:

1105 Exploring the origins and evolution of ice giant planets, Planetary and Space Science, 104, 122,

1106 10.1016/j.pss.2014.08.009.

1107

1108 Arridge, C. S., Agnor, C. B., André, N., et al. (2012), Uranus Pathfinder: exploring the origins and evolution of

1109 Ice Giant planets, Experimental Astronomy, 33, 753, 10.1007/s10686-011-9251-4.

1110

1111 Atreya, S. K., Hofstadter, M. H., In, J. H., et al. (2020), Deep Atmosphere Composition, Structure, Origin, and

1112 Exploration, with Particular Focus on Critical in situ Science at the Icy Giants, Space Science Reviews, 216, 18,

1113 10.1007/s11214-020-0640-8.

1114

1115 Bailey, E., & Stevenson, D. J. (2015), Modeling Ice Giant Interiors Using Constraints on the H2-H2O Critical

1116 Curve, AGU Fall Meeting Abstracts, 2015, P31G-03.

1117

1118 Bali, E., Audétat, A., & Keppler, H. (2013), Water and hydrogen are immiscible in Earth's mantle, Nature, 495,

1119 220, 10.1038/nature11908.

1120 49 Fletcher et al., Ice Giant Systems

1121 Barclay, T., Pepper, J., & Quintana, E. V. (2018), A Revised Exoplanet Yield from the Transiting Exoplanet

1122 Survey Satellite (TESS), The Astrophysical Journal Supplement Series, 239, 2, 10.3847/1538-4365/aae3e9.

1123

1124 Barthélemy, M., Lamy, L., Menager, H., et al. (2014), Dayglow and auroral emissions of Uranus in H2 FUV

1125 bands, Icarus, 239, 160, 10.1016/j.icarus.2014.05.035.

1126

1127 Batalha, N. M., Rowe, J. F., Bryson, S. T., et al. (2013), Planetary Candidates Observed by Kepler. III. Analysis

1128 of the First 16 Months of Data, The Astrophysical Journal Supplement Series, 204, 24, 10.1088/0067-

1129 0049/204/2/24.

1130

1131 Běhounková, M., Tobie, G., Choblet, G., et al. (2012), Tidally-induced melting events as the origin of south-

1132 pole activity on Enceladus, Icarus, 219, 655, 10.1016/j.icarus.2012.03.024.

1133

1134 Boué, G., & Laskar, J. (2010), A Collisionless Scenario for Uranus Tilting, The Astrophysical Journal, 712, L44,

1135 10.1088/2041-8205/712/1/L44.

1136

1137 Broadfoot, A. L., Herbert, F., Holberg, J. B., et al. (1986), Ultraviolet Spectrometer Observations of Uranus,

1138 Science, 233, 74, 10.1126/science.233.4759.74.

1139

1140 Broadfoot, A. L., Atreya, S. K., Bertaux, J. L., et al. (1989), Ultraviolet Spectrometer Observations of Neptune

1141 and Triton, Science, 246, 1459, 10.1126/science.246.4936.1459.

1142

1143 Brown, R. H., & Cruikshank, D. P. (1983), The Uranian satellites: Surface compositions and opposition

1144 brightness surges, Icarus, 55, 83, 10.1016/0019-1035(83)90052-0.

1145

50 Fletcher et al., Ice Giant Systems

1146 Buchvarova, M., & Velinov, P. (2009), Cosmic ray spectra in planetary atmospheres, Universal Heliophysical

1147 Processes, 257, 471, 10.1017/S1743921309029718.

1148

1149

1150 Cao, X., & Paty, C. (2017), Diurnal and seasonal variability of Uranus's magnetosphere, Journal of Geophysical

1151 Research (Space Physics), 122, 6318, 10.1002/2017JA024063.

1152

1153 Cartwright, R. J., Emery, J. P., Rivkin, A. S., et al. (2015), Distribution of CO2 ice on the large moons of Uranus

1154 and evidence for compositional stratification of their near-surfaces, Icarus, 257, 428,

1155 10.1016/j.icarus.2015.05.020.

1156

1157 Cartwright, R. J., Emery, J. P., Pinilla-Alonso, N., et al. (2018), Red material on the large moons of Uranus:

1158 Dust from the irregular satellites?, Icarus, 314, 210, 10.1016/j.icarus.2018.06.004.

1159

1160 Cavalié, T., Venot, O., Selsis, F., et al. (2017), Thermochemistry and vertical mixing in the tropospheres of

1161 Uranus and Neptune: How convection inhibition can affect the derivation of deep oxygen abundances, Icarus,

1162 291, 1, 10.1016/j.icarus.2017.03.015.

1163

1164 Cavalié, T., Venot, O., Miguel, Y., et al. (2020), The deep composition of Uranus and Neptune from in situ

1165 exploration and thermochemical modeling, Space Science Reviews, accepted, 10.1007/s11214-020-00677-8.

1166

1167 Charnoz, S., Canup, R. M., Crida, A., et al. (2018), The Origin of Planetary Ring Systems, Planetary Ring

1168 Systems. Properties, Structure, and Evolution, 517, 10.1017/9781316286791.018.

1169

1170 Colwell, J. E., & Esposito, L. W. (1993), Origins of the rings of Uranus and Neptune. 2. Initial conditions and

1171 ring moon populations, Journal of Geophysical Research, 98, 7387, 10.1029/93JE00329.

51 Fletcher et al., Ice Giant Systems

1172

1173 Colwell, J. E., & Esposito, L. W. (1992), Origins of the rings of Uranus and Neptune 1. Statistics of satellite

1174 disruptions, Journal of Geophysical Research, 97, 10227, 10.1029/92JE00788.

1175

1176 Conrath, B. J., Gierasch, P. J., & Ustinov, E. A. (1998), Thermal Structure and Para Hydrogen Fraction on the

1177 Outer Planets from Voyager IRIS Measurements, Icarus, 135, 501, 10.1006/icar.1998.6000.

1178

1179 Cowley, S. W. H. (2013), Response of Uranus' auroras to solar wind compressions at equinox, Journal of

1180 Geophysical Research (Space Physics), 118, 2897, 10.1002/jgra.50323.

1181

1182 Croft, S. K., & Soderblom, L. A. (1991), Geology of the Uranian satellites., Uranus, 561.

1183

1184 Cruikshank, D. P., Schmitt, B., Roush, T. L., et al. (2000), Water Ice on Triton, Icarus, 147, 309,

1185 10.1006/icar.2000.6451.

1186

1187 de Pater, I., Romani, P. N., & Atreya, S. K. (1991), Possible microwave absorption by H 2S gas in Uranus' and

1188 Neptune's atmospheres, Icarus, 91, 220, 10.1016/0019-1035(91)90020-T.

1189

1190 de Pater, I., Hammel, H. B., Showalter, M. R., et al. (2007), The Dark Side of the Rings of Uranus, Science, 317,

1191 1888, 10.1126/science.1148103.

1192

1193 de Pater, I., Sromovsky, L. A., Fry, P. M., et al. (2015), Record-breaking storm activity on Uranus in 2014,

1194 Icarus, 252, 121, 10.1016/j.icarus.2014.12.037.

1195

1196 de Pater, I., Gibbard, S. G., & Hammel, H. B. (2006), Evolution of the dusty rings of Uranus, Icarus, 180, 186,

1197 10.1016/j.icarus.2005.08.011.

52 Fletcher et al., Ice Giant Systems

1198

1199 de Pater, I., Gibbard, S. G., Chiang, E., et al. (2005), The dynamic neptunian ring arcs: evidence for a gradual

1200 disappearance of Liberté and resonant jump of courage, Icarus, 174, 263, 10.1016/j.icarus.2004.10.020.

1201

1202 de Pater, I., Fletcher, L. N., Luszcz-Cook, S., et al. (2014), Neptune’s global circulation deduced from multi-

1203 wavelength observations, Icarus, 237, 211, 10.1016/j.icarus.2014.02.030.

1204

1205 de Pater, I., Renner, S., Showalter, M. R., et al. (2018), The Rings of Neptune, Planetary Ring Systems.

1206 Properties, Structure, and Evolution, 112, 10.1017/9781316286791.005.

1207

1208 Decker, R. B., & Cheng, A. F. (1994), A model of Triton's role in Neptune's magnetosphere, Journal of

1209 Geophysical Research, 99, 19027, 10.1029/94JE01867.

1210

1211 Desch, M. D., Kaiser, M. L., Zarka, P., et al. (1991), Uranus as a radio source., Uranus, 894.

1212

1213 Dobrijevic, M., Loison, J. C., Hue, V., et al. (2020), 1D photochemical model of the ionosphere and the

1214 stratosphere of Neptune, Icarus, 335, 113375, 10.1016/j.icarus.2019.07.009.

1215

1216 Dodson-Robinson, S. E., & Bodenheimer, P. (2010), The formation of Uranus and Neptune in solid-rich feeding

1217 zones: Connecting chemistry and dynamics, Icarus, 207, 491, 10.1016/j.icarus.2009.11.021.

1218

1219 Dumas, C., Terrile, R. J., Smith, B. A., et al. (1999), Stability of Neptune's ring arcs in question, Nature, 400,

1220 733, 10.1038/23414.

1221

1222 Farrell, W. M. (1992), Nonthermal radio emissions from Uranus, Planetary Radio Emissions III, 241.

1223

53 Fletcher et al., Ice Giant Systems

1224 Feuchtgruber, H., Lellouch, E., de Graauw, T., et al. (1997), External supply of oxygen to the atmospheres of

1225 the giant planets, Nature, 389, 159, 10.1038/38236.

1226

1227 Fletcher, L. N., de Pater, I., Orton, G. S., et al. (2020), Ice Giant Circulation Patterns: Implications for

1228 Atmospheric Probes, Space Science Reviews, 216, 21, 10.1007/s11214-020-00646-1.

1229

1230 Fletcher, L. N., de Pater, I., Orton, G. S., et al. (2014), Neptune at summer solstice: Zonal mean temperatures

1231 from ground-based observations, 2003-2007, Icarus, 231, 146, 10.1016/j.icarus.2013.11.035.

1232

1233 French, R. G., Nicholson, P. D., Porco, C. C., et al. (1991), Dynamics and structure of the Uranian rings., Uranus,

1234 327.

1235

1236 Fressin, F., Torres, G., Charbonneau, D., et al. (2013), The False Positive Rate of Kepler and the Occurrence of

1237 Planets, The Astrophysical Journal, 766, 81, 10.1088/0004-637X/766/2/81.

1238

1239 Fulton, B. J., & Petigura, E. A. (2018), The California-Kepler Survey. VII. Precise Planet Radii Leveraging Gaia

1240 DR2 Reveal the Stellar Mass Dependence of the Planet Radius Gap, The Astronomical Journal, 156, 264,

1241 10.3847/1538-3881/aae828.

1242

1243 Gierasch, P. J., & Conrath, B. J. (1987), Vertical temperature gradients on Uranus: Implications for layered

1244 convection, Journal of Geophysical Research, 92, 15019, 10.1029/JA092iA13p15019.

1245

1246 Griton, L., Pantellini, F., & Meliani, Z. (2018), Three-Dimensional Magnetohydrodynamic Simulations of the

1247 Solar Wind Interaction With a Hyperfast-Rotating Uranus, Journal of Geophysical Research (Space Physics),

1248 123, 5394, 10.1029/2018JA025331.

1249

54 Fletcher et al., Ice Giant Systems

1250 Griton, L., & Pantellini, F. (2020), Magnetohydrodynamic simulations of a Uranus-at-equinox type rotating

1251 magnetosphere, Astronomy and Astrophysics, 633, A87, 10.1051/0004-6361/201936604.

1252

1253 Grundy, W. M., Binzel, R. P., Buratti, B. J., et al. (2016), Surface compositions across Pluto and Charon,

1254 Science, 351, aad9189, 10.1126/science.aad9189.

1255

1256 Grundy, W. M., Young, L. A., Spencer, J. R., et al. (2006), Distributions of H 2O and CO 2 ices on Ariel, Umbriel,

1257 Titania, and Oberon from IRTF/SpeX observations, Icarus, 184, 543, 10.1016/j.icarus.2006.04.016.

1258

1259 Guillot, T. (1995), Condensation of Methane, Ammonia, and Water and the Inhibition of Convection in Giant

1260 Planets, Science, 269, 1697, 10.1126/science.7569896.

1261

1262 Guillot, T. (2019), Uranus and Neptune are key to understand planets with hydrogen atmospheres, arXiv e-

1263 prints, arXiv:1908.02092.

1264

1265 Gunell, H., Maggiolo, R., Nilsson, H., et al. (2018), Why an intrinsic magnetic field does not protect a planet

1266 against atmospheric escape, Astronomy and Astrophysics, 614, L3, 10.1051/0004-6361/201832934.

1267

1268 Gurnett, D. A., Kurth, W. S., Cairns, I. H., et al. (1990), Whistlers in Neptune's magnetosphere: Evidence of

1269 atmospheric lightning, Journal of Geophysical Research, 95, 20967, 10.1029/JA095iA12p20967.

1270

1271 Helled, R., Bodenheimer, P., Podolak, M., et al. (2014), Giant Planet Formation, Evolution, and Internal

1272 Structure, Protostars and Planets VI, 643, 10.2458/azu_uapress_9780816531240-ch028.

1273

1274 Helled, R., & Bodenheimer, P. (2014), The Formation of Uranus and Neptune: Challenges and Implications for

1275 Intermediate-mass Exoplanets, The Astrophysical Journal, 789, 69, 10.1088/0004-637X/789/1/69.

55 Fletcher et al., Ice Giant Systems

1276

1277 Helled, R., Anderson, J. D., Podolak, M., et al. (2011), Interior Models of Uranus and Neptune, The

1278 Astrophysical Journal, 726, 15, 10.1088/0004-637X/726/1/15.

1279

1280 Helled, R., Anderson, J. D., & Schubert, G. (2010), Uranus and Neptune: Shape and rotation, Icarus, 210, 446,

1281 10.1016/j.icarus.2010.06.037.

1282

1283 Herbert, F. (2009), Aurora and magnetic field of Uranus, Journal of Geophysical Research (Space Physics),

1284 114, A11206, 10.1029/2009JA014394.

1285

1286 Herbert, F., Sandel, B. R., Yelle, R. V., et al. (1987), The upper atmosphere of Uranus: EUV occultations

1287 observed by Voyager 2, Journal of Geophysical Research, 92, 15093, 10.1029/JA092iA13p15093.

1288

1289 Hofstadter, M., Simon, A., Atreya, S., et al. (2019), Uranus and Neptune missions: A study in advance of the

1290 next Planetary Science Decadal Survey, Planetary and Space Science, 177, 104680,

1291 10.1016/j.pss.2019.06.004.

1292

1293 Hofstadter, M. D., & Butler, B. J. (2003), Seasonal change in the deep atmosphere of Uranus, Icarus, 165, 168,

1294 10.1016/S0019-1035(03)00174-X.

1295

1296 Hoogeveen, G. W., & Cloutier, P. A. (1996), The Triton-Neptune plasma interaction, Journal of Geophysical

1297 Research, 101, 19, 10.1029/95JA02761.

1298

1299 Hsu, H.-W., Schmidt, J., Kempf, S., et al. (2018), In situ collection of dust grains falling from Saturn's rings into

1300 its atmosphere, Science, 362, aat3185, 10.1126/science.aat3185.

1301

56 Fletcher et al., Ice Giant Systems

1302 Hueso, R., & Sánchez-Lavega, A. (2019), Atmospheric Dynamics and Vertical Structure of Uranus and

1303 Neptune's Weather Layers, Space Science Reviews, 215, 52, 10.1007/s11214-019-0618-6.

1304

1305 Hueso, R., Guillot, T., Sánchez-Lavega, A. (2020). Atmospheric dynamics and convective regimes in the non-

1306 homogeneous weather layers of the ice giants, Philosophical Transactions A. Submitted.

1307

1308 Hussmann, H., Sohl, F., & Spohn, T. (2006), Subsurface oceans and deep interiors of medium-sized outer

1309 planet satellites and large trans-neptunian objects, Icarus, 185, 258, 10.1016/j.icarus.2006.06.005.

1310

1311 Iess, L., Militzer, B., Kaspi, Y., et al. (2019), Measurement and implications of Saturn's gravity field and ring

1312 mass, Science, 364, aat2965, 10.1126/science.aat2965.

1313

1314 Irwin, P. G. J., Toledo, D., Garland, R., et al. (2018), Detection of hydrogen sulfide above the clouds in Uranus's

1315 atmosphere, Nature Astronomy, 2, 420, 10.1038/s41550-018-0432-1.

1316

1317 Jacobson, R. A. (2014), The Orbits of the Uranian Satellites and Rings, the Gravity Field of the Uranian System,

1318 and the Orientation of the Pole of Uranus, The Astronomical Journal, 148, 76, 10.1088/0004-6256/148/5/76.

1319

1320 Jacobson, R. A. (2009), The Orbits of the Neptunian Satellites and the Orientation of the Pole of Neptune,

1321 The Astronomical Journal, 137, 4322, 10.1088/0004-6256/137/5/4322.

1322

1323 Karkoschka, E., & Tomasko, M. G. (2011), The haze and methane distributions on Neptune from HST-STIS

1324 spectroscopy, Icarus, 211, 780, 10.1016/j.icarus.2010.08.013.

1325

1326 Kaspi, Y., Showman, A. P., Hubbard, W. B., et al. (2013), Atmospheric confinement of jet streams on Uranus

1327 and Neptune, Nature, 497, 344, 10.1038/nature12131.

57 Fletcher et al., Ice Giant Systems

1328

1329 Kegerreis, J. A., Eke, V. R., Gonnet, P., et al. (2019), Planetary giant impacts: convergence of high-resolution

1330 simulations using efficient spherical initial conditions and SWIFT, Monthly Notices of the Royal Astronomical

1331 Society, 487, 5029, 10.1093/mnras/stz1606.

1332

1333 Kempf, S., Altobelli, N., Srama, R., et al. (2018), The Age of Saturn's Rings Constrained by the Meteoroid Flux

1334 Into the System, EGU General Assembly Conference Abstracts, 10791.

1335

1336 Krasnopolsky, V. A., & Cruikshank, D. P. (1995), Photochemistry of Triton's atmosphere and ionosphere.,

1337 Journal of Geophysical Research, 100, 21,271,286.

1338

1339 Lainey, V. (2008), A new dynamical model for the Uranian satellites, Planetary and Space Science, 56, 1766,

1340 10.1016/j.pss.2008.02.015.

1341

1342 Lambrechts, M., Johansen, A., & Morbidelli, A. (2014), Separating gas-giant and ice-giant planets by halting

1343 pebble accretion, Astronomy and Astrophysics, 572, A35, 10.1051/0004-6361/201423814.

1344

1345 Lammer, H. (1995), Mass loss of N 2 molecules from Triton by magnetospheric plasma interaction, Planetary

1346 and Space Science, 43, 845, 10.1016/0032-0633(94)00214-C.

1347

1348 Lamy, L., Prangé, R., Hansen, K. C., et al. (2017), The aurorae of Uranus past equinox, Journal of Geophysical

1349 Research (Space Physics), 122, 3997, 10.1002/2017JA023918.

1350

1351 Lamy, L., Prangé, R., Hansen, K. C., et al. (2012), Earth-based detection of Uranus' aurorae, Geophysical

1352 Research Letters, 39, L07105, 10.1029/2012GL051312.

1353

58 Fletcher et al., Ice Giant Systems

1354 LeBeau, R. P., & Dowling, T. E. (1998), EPIC Simulations of Time-Dependent, Three-Dimensional Vortices with

1355 Application to Neptune's Great Dark SPOT, Icarus, 132, 239, 10.1006/icar.1998.5918.

1356

1357 Leconte, J., Selsis, F., Hersant, F., et al. (2017), Condensation-inhibited convection in hydrogen-rich

1358 atmospheres . Stability against double-diffusive processes and thermal profiles for Jupiter, Saturn, Uranus,

1359 and Neptune, Astronomy and Astrophysics, 598, A98, 10.1051/0004-6361/201629140.

1360

1361 Lellouch, E., de Bergh, C., Sicardy, B., et al. (2010), Detection of CO in Triton's atmosphere and the nature of

1362 surface-atmosphere interactions, Astronomy and Astrophysics, 512, L8, 10.1051/0004-6361/201014339.

1363

1364 Li, C., Le, T., Zhang, X., et al. (2018), A high-performance atmospheric radiation package: With applications to

1365 the radiative energy budgets of giant planets, Journal of Quantitative Spectroscopy and Radiative Transfer,

1366 217, 353, 10.1016/j.jqsrt.2018.06.002.

1367

1368 Lindal, G. F., Lyons, J. R., Sweetnam, D. N., et al. (1987), The atmosphere of Uranus: Results of radio

1369 occultation measurements with Voyager 2, Journal of Geophysical Research, 92, 14987,

1370 10.1029/JA092iA13p14987.

1371

1372 Lindal, G. F. (1992), The Atmosphere of Neptune: an Analysis of Radio Occultation Data Acquired with

1373 Voyager 2, The Astronomical Journal, 103, 967, 10.1086/116119.

1374

1375 Majeed, T., Waite, J. H., Bougher, S. W., et al. (2004), The ionospheres-thermospheres of the giant planets,

1376 Advances in Space Research, 33, 197, 10.1016/j.asr.2003.05.009.

1377

1378 Martin-Herrero, A., Romeo, I., & Ruiz, J. (2018), Heat flow in Triton: Implications for heat sources powering

1379 recent geologic activity, Planetary and Space Science, 160, 19, 10.1016/j.pss.2018.03.010.

59 Fletcher et al., Ice Giant Systems

1380

1381 Masters, A. (2014), Magnetic reconnection at Uranus' magnetopause, Journal of Geophysical Research

1382 (Space Physics), 119, 5520, 10.1002/2014JA020077.

1383

1384 Masters, A. (2018), A More Viscous-Like Solar Wind Interaction With All the Giant Planets, Geophysical

1385 Research Letters, 45, 7320, 10.1029/2018GL078416.

1386

1387 Masters, A., Achilleos, N., Agnor, C. B., et al. (2014), Neptune and Triton: Essential pieces of the Solar System

1388 puzzle, Planetary and Space Science, 104, 108, 10.1016/j.pss.2014.05.008.

1389

1390 Mauk, B. H., & Fox, N. J. (2010), Electron radiation belts of the solar system, Journal of Geophysical Research

1391 (Space Physics), 115, A12220, 10.1029/2010JA015660.

1392

1393 McKinnon, W. B., & Leith, A. C. (1995), Gas drag and the orbital evolution of a captured Triton., Icarus, 118,

1394 392, 10.1006/icar.1995.1199.

1395

1396 McKinnon, W. B., Stern, S. A., Weaver, H. A., et al. (2017), Origin of the Pluto-Charon system: Constraints

1397 from the New Horizons flyby, Icarus, 287, 2, 10.1016/j.icarus.2016.11.019.

1398

1399 McNutt, R. L., Selesnick, R. S., & Richardson, J. D. (1987), Low-energy plasma observations in the

1400 magnetosphere of Uranus, Journal of Geophysical Research, 92, 4399, 10.1029/JA092iA05p04399.

1401

1402 Mejnertsen, L., Eastwood, J. P., Chittenden, J. P., et al. (2016), Global MHD simulations of Neptune's

1403 magnetosphere, Journal of Geophysical Research (Space Physics), 121, 7497, 10.1002/2015JA022272.

1404

60 Fletcher et al., Ice Giant Systems

1405 Melin, H., Stallard, T., Miller, S., et al. (2011), Seasonal Variability in the Ionosphere of Uranus, The

1406 Astrophysical Journal, 729, 134, 10.1088/0004-637X/729/2/134.

1407

1408 Melin, H., Fletcher, L. N., Stallard, T. S., et al. (2019), The H3+ ionosphere of Uranus: decades-long cooling

1409 and local-time morphology, Philosophical Transactions of the Royal Society of London Series A, 377,

1410 20180408, 10.1098/rsta.2018.0408.

1411

1412 Melin, H., Stallard, T. S., Miller, S., et al. (2013), Post-equinoctial observations of the ionosphere of Uranus,

1413 Icarus, 223, 741, 10.1016/j.icarus.2013.01.012.

1414

1415 Merrill, R. T., & McFadden, P. L. (1999), Geomagnetic polarity transitions, Reviews of Geophysics, 37, 201,

1416 10.1029/1998RG900004.

1417

1418 Millot, M., Coppari, F., Rygg, J. R., et al. (2019), Nanosecond X-ray diffraction of shock-compressed superionic

1419 water ice, Nature, 569, 251, 10.1038/s41586-019-1114-6.

1420

1421 Molter, E. M., de Pater, I., Roman, M. T., et al. (2019), Thermal Emission from the Uranian Ring System, The

1422 Astronomical Journal, 158, 47, 10.3847/1538-3881/ab258c.

1423

1424 Moreno, R., Marten, A., & Lellouch, E. (2009), Search for PH3 in the Atmospheres of Uranus and Neptune at

1425 Millimeter Wavelength, AAS/Division for Planetary Sciences Meeting Abstracts #41, 28.02.

1426

1427 Morley, C. V., Fortney, J. J., Marley, M. S., et al. (2012), Neglected Clouds in T and Y Dwarf Atmospheres, The

1428 Astrophysical Journal, 756, 172, 10.1088/0004-637X/756/2/172.

1429

61 Fletcher et al., Ice Giant Systems

1430 Moses, J. I., & Poppe, A. R. (2017), Dust ablation on the giant planets: Consequences for stratospheric

1431 photochemistry, Icarus, 297, 33, 10.1016/j.icarus.2017.06.002.

1432

1433 Moses, J. I., Fletcher, L. N., Greathouse, T. K., et al. (2018), Seasonal stratospheric photochemistry on Uranus

1434 and Neptune, Icarus, 307, 124, 10.1016/j.icarus.2018.02.004.

1435

1436 Mousis, O., Atkinson, D. H., Cavalié, T., et al. (2018), Scientific rationale for Uranus and Neptune in situ

1437 explorations, Planetary and Space Science, 155, 12, 10.1016/j.pss.2017.10.005.

1438

1439 Namouni, F., & Porco, C. (2002), The confinement of Neptune's ring arcs by the moon Galatea, Nature, 417,

1440 45, 10.1038/417045a.

1441

1442 Nettelmann, N., Helled, R., Fortney, J. J., et al. (2013), New indication for a dichotomy in the interior structure

1443 of Uranus and Neptune from the application of modified shape and rotation data, Planetary and Space

1444 Science, 77, 143, 10.1016/j.pss.2012.06.019.

1445

1446 Neubauer, F. M. (1990), Satellite plasma interactions, Advances in Space Research, 10, 25, 10.1016/0273-

1447 1177(90)90083-C.

1448

1449 Nicholson, P. D., De Pater, I., French, R. G., et al. (2018), The Rings of Uranus, Planetary Ring Systems.

1450 Properties, Structure, and Evolution, 93, 10.1017/9781316286791.004.

1451

1452 Nimmo, F., & Spencer, J. R. (2015), Powering Triton's recent geological activity by obliquity tides: Implications

1453 for Pluto geology, Icarus, 246, 2, 10.1016/j.icarus.2014.01.044.

1454

62 Fletcher et al., Ice Giant Systems

1455 Orton, G. S., Fletcher, L. N., Moses, J. I., et al. (2014), Mid-infrared spectroscopy of Uranus from the Spitzer

1456 Infrared Spectrometer: 1. Determination of the mean temperature structure of the upper troposphere and

1457 stratosphere, Icarus, 243, 494, 10.1016/j.icarus.2014.07.010.

1458

1459 Pearl, J. C., Conrath, B. J., Hanel, R. A., et al. (1990), The albedo, effective temperature, and energy balance

1460 of Uranus, as determined from Voyager IRIS data, Icarus, 84, 12, 10.1016/0019-1035(90)90155-3.

1461

1462 Pearl, J. C., & Conrath, B. J. (1991), The albedo, effective temperature, and energy balance of Neptune, as

1463 determined from Voyager data, Journal of Geophysical Research, 96, 18921, 10.1029/91JA01087.

1464

1465 Penny, M. T., Gaudi, B. S., Kerins, E., et al. (2019), Predictions of the WFIRST Microlensing Survey. I. Bound

1466 Planet Detection Rates, The Astrophysical Journal Supplement Series, 241, 3, 10.3847/1538-4365/aafb69.

1467

1468 Petigura, E. A., Howard, A. W., Marcy, G. W., et al. (2017), The California-Kepler Survey. I. High-resolution

1469 Spectroscopy of 1305 Stars Hosting Kepler Transiting Planets, The Astronomical Journal, 154, 107,

1470 10.3847/1538-3881/aa80de.

1471

1472 Plainaki, C., Lilensten, J., Radioti, A., et al. (2016), Planetary space weather: scientific aspects and future

1473 perspectives, Journal of Space Weather and Space Climate, 6, A31, 10.1051/swsc/2016024.

1474

1475 Plescia, J. B. (1987a), Cratering history of the Uranian satellites: Umbriel, Titania, and Oberon, Journal of

1476 Geophysical Research, 92, 14918, 10.1029/JA092iA13p14918.

1477

1478 Plescia, J. B. (1987b), Geological terrains and crater frequencies on Ariel, Nature, 327, 201,

1479 10.1038/327201a0.

1480

63 Fletcher et al., Ice Giant Systems

1481 Podolak, M., & Helled, R. (2012), What Do We Really Know about Uranus and Neptune?, The Astrophysical

1482 Journal, 759, L32, 10.1088/2041-8205/759/2/L32.

1483

1484 Podolak, M., Weizman, A., & Marley, M. (1995), Comparative models of Uranus and Neptune, Planetary and

1485 Space Science, 43, 1517, 10.1016/0032-0633(95)00061-5.

1486

1487 Pollack, J. B., Hubickyj, O., Bodenheimer, P., et al. (1996), Formation of the Giant Planets by Concurrent

1488 Accretion of Solids and Gas, Icarus, 124, 62, 10.1006/icar.1996.0190.

1489

1490 Postberg, F., Kempf, S., Schmidt, J., et al. (2009), Sodium salts in E-ring ice grains from an ocean below the

1491 surface of Enceladus, Nature, 459, 1098, 10.1038/nature08046.

1492

1493 Prockter, L. M., Nimmo, F., & Pappalardo, R. T. (2005), A shear heating origin for ridges on Triton, Geophysical

1494 Research Letters, 32, L14202, 10.1029/2005GL022832.

1495

1496 Quirico, E., Douté, S., Schmitt, B., et al. (1999), Composition, Physical State, and Distribution of Ices at the

1497 Surface of Triton, Icarus, 139, 159, 10.1006/icar.1999.6111.

1498

1499 Rages, K., Pollack, J. B., Tomasko, M. G., et al. (1991), Properties of scatterers in the troposphere and lower

1500 stratosphere of Uranus based on Voyager imaging data, Icarus, 89, 359, 10.1016/0019-1035(91)90183-T.

1501

1502 Redmer, R., Mattsson, T. R., Nettelmann, N., et al. (2011), The phase diagram of water and the magnetic

1503 fields of Uranus and Neptune, Icarus, 211, 798, 10.1016/j.icarus.2010.08.008.

1504

1505 Reinhardt, C., Chau, A., Stadel, J., et al. (2020), Bifurcation in the history of Uranus and Neptune: the role of

1506 giant impacts, Monthly Notices of the Royal Astronomical Society, 492, 5336, 10.1093/mnras/stz3271.

64 Fletcher et al., Ice Giant Systems

1507

1508 Richardson, J. D., & McNutt, R. L. (1990), Low-energy plasma in Neptune's magnetosphere, Geophysical

1509 Research Letters, 17, 1689, 10.1029/GL017i010p01689.

1510

1511 Roman, M. T., Fletcher, L. N., Orton, G. S., et al. (2020), Uranus in Northern Midspring: Persistent Atmospheric

1512 Temperatures and Circulations Inferred from Thermal Imaging, The Astronomical Journal, 159, 45,

1513 10.3847/1538-3881/ab5dc7.

1514

1515 Rymer, A., Mandt, K., Hurley, D., et al. (2019), Solar System Ice Giants: Exoplanets in our Backyard., Bulletin

1516 of the American Astronomical Society, 51, 176.

1517

1518 Safronov, V. S. (1966), Sizes of the largest bodies falling onto the planets during their formation, Soviet

1519 Astronomy, 9, 987.

1520

1521 Sánchez-Lavega, A., Sromovsky, L. A., et al. (2018), Gas Giants. In: Galperin, B., Read, P.L., (eds) Zonal Jets

1522 Phenomenology, Genesis and Physics. Cambridge (doi: 10.1017/9781107358225).

1523

1524 Scarf, F. L., Gurnett, D. A., Kurth, W. S., et al. (1987), Voyager 2 plasma wave observations at Uranus, Advances

1525 in Space Research, 7, 253, 10.1016/0273-1177(87)90226-2.

1526

1527 Selesnick, R. S. (1988), Magnetospheric convection in the nondipolar magnetic field of Uranus, Journal of

1528 Geophysical Research, 93, 9607, 10.1029/JA093iA09p09607.

1529

1530 Showalter, M. R., & Lissauer, J. J. (2006), The Second Ring-Moon System of Uranus: Discovery and Dynamics,

1531 Science, 311, 973, 10.1126/science.1122882.

1532

65 Fletcher et al., Ice Giant Systems

1533 Simon, A. A., Fletcher, L. N., Arridge, C., et al. (2020), A Review of the in Situ Probe Designs from Recent Ice

1534 Giant Mission Concept Studies, Space Science Reviews, 216, 17, 10.1007/s11214-020-0639-1.

1535

1536 Simon, A. A., Wong, M. H., & Hsu, A. I. (2019), Formation of a New Great Dark Spot on Neptune in 2018,

1537 Geophysical Research Letters, 46, 3108, 10.1029/2019GL081961.

1538

1539 Simon, A. A., Stern, S. A., & Hofstadter, M. (2018), Outer Solar System Exploration: A Compelling and Unified

1540 Dual Mission Decadal Strategy for Exploring Uranus, Neptune, Triton, Dwarf Planets, and Small KBOs and

1541 Centaurs, arXiv e-prints, arXiv:1807.08769.

1542

1543 Simon, A. A., Rowe, J. F., Gaulme, P., et al. (2016), Neptune's Dynamic Atmosphere from Kepler K2

1544 Observations: Implications for Brown Dwarf Light Curve Analyses, The Astrophysical Journal, 817, 162,

1545 10.3847/0004-637X/817/2/162.

1546

1547 Slattery, W. L., Benz, W., & Cameron, A. G. W. (1992), Giant impacts on a primitive Uranus, Icarus, 99, 167,

1548 10.1016/0019-1035(92)90180-F.

1549

1550 Smith, B. A., Soderblom, L. A., Banfield, D., et al. (1989), Voyager 2 at Neptune: Imaging Science Results,

1551 Science, 246, 1422, 10.1126/science.246.4936.1422.

1552

1553 Smith, M. D., & Gierasch, P. J. (1995), Convection in the outer planet atmospheres including ortho-para

1554 hydrogen conversion., Icarus, 116, 159, 10.1006/icar.1995.1118.

1555

1556 Soderblom, L. A., Kieffer, S. W., Becker, T. L., et al. (1990), Triton's Geyser-Like Plumes: Discovery and Basic

1557 Characterization, Science, 250, 410, 10.1126/science.250.4979.410.

1558

66 Fletcher et al., Ice Giant Systems

1559 Soderlund, K. M., Heimpel, M. H., King, E. M., et al. (2013), Turbulent models of ice giant internal dynamics:

1560 Dynamos, heat transfer, and zonal flows, Icarus, 224, 97, 10.1016/j.icarus.2013.02.014.

1561

1562 Sromovsky, L. A., de Pater, I., Fry, P. M., et al. (2015), High S/N Keck and Gemini AO imaging of Uranus during

1563 2012-2014: New cloud patterns, increasing activity, and improved wind measurements, Icarus, 258, 192,

1564 10.1016/j.icarus.2015.05.029.

1565

1566 Sromovsky, L. A., Karkoschka, E., Fry, P. M., et al. (2014), Methane depletion in both polar regions of Uranus

1567 inferred from HST/STIS and Keck/NIRC2 observations, Icarus, 238, 137, 10.1016/j.icarus.2014.05.016.

1568

1569 Stanley, S., & Bloxham, J. (2004), Convective-region geometry as the cause of Uranus' and Neptune's unusual

1570 magnetic fields, Nature, 428, 151, 10.1038/nature02376.

1571

1572 Stanley, S., & Bloxham, J. (2006), Numerical dynamo models of Uranus' and Neptune's magnetic fields, Icarus,

1573 184, 556, 10.1016/j.icarus.2006.05.005.

1574

1575 Stauffer, J., Marley, M. S., Gizis, J. E., et al. (2016), Spitzer Space Telescope Mid-IR Light Curves of Neptune,

1576 The Astronomical Journal, 152, 142, 10.3847/0004-6256/152/5/142.

1577

1578 Stern, S. A., & McKinnon, W. B. (2000), Triton's Surface Age and Impactor Population Revisited in Light of

1579 Kuiper Belt Fluxes: Evidence for Small Kuiper Belt Objects and Recent Geological Activity, The Astronomical

1580 Journal, 119, 945, 10.1086/301207.

1581

1582 Stevenson, D. J. (1986), The Uranus-Neptune Dichotomy: the Role of Giant Impacts, Lunar and Planetary

1583 Science Conference, 1011.

1584

67 Fletcher et al., Ice Giant Systems

1585 Stoker, C. R., & Toon, O. B. (1989), Moist convection on Neptune, Geophysical Research Letters, 16, 929,

1586 10.1029/GL016i008p00929.

1587

1588 Stone, E. C., Cooper, J. F., Cummings, A. C., et al. (1986), Energetic Charged Particles in the Uranian

1589 Magnetosphere, Science, 233, 93, 10.1126/science.233.4759.93.

1590

1591 Stratman, P. W., Showman, A. P., Dowling, T. E., et al. (2001), EPIC Simulations of Bright Companions to

1592 Neptune's Great Dark Spots, Icarus, 151, 275, 10.1006/icar.2001.6603.

1593

1594 Strobel, D. F., Cheng, A. F., Summers, M. E., et al. (1990), Magnetospheric interaction with Triton's

1595 ionosphere, Geophysical Research Letters, 17, 1661, 10.1029/GL017i010p01661.

1596

1597 Stryk, T., & Stooke, P. J. (2008), Voyager 2 Images of Uranian Satellites: Reprocessing and New

1598 Interpretations, Lunar and Planetary Science Conference, 1362.

1599

1600 Sun, Z.-P., Schubert, G., & Stoker, C. R. (1991), Thermal and humidity winds in outer planet atmospheres,

1601 Icarus, 91, 154, 10.1016/0019-1035(91)90134-F.

1602

1603 Tegler, S. C., Grundy, W. M., Olkin, C. B., et al. (2012), Ice Mineralogy across and into the Surfaces of Pluto,

1604 Triton, and Eris, The Astrophysical Journal, 751, 76, 10.1088/0004-637X/751/1/76.

1605

1606 Tegler, S. C., Stufflebeam, T. D., Grundy, W. M., et al. (2019), A New Two-molecule Combination Band as a

1607 Diagnostic of Carbon Monoxide Diluted in Nitrogen Ice on Triton, The Astronomical Journal, 158, 17,

1608 10.3847/1538-3881/ab199f.

1609

68 Fletcher et al., Ice Giant Systems

1610 Thompson, W. R., & Sagan, C. (1990), Color and chemistry on Triton, Science, 250, 415,

1611 10.1126/science.11538073.

1612

1613 Tiscareno, M. S., Hedman, M. M., Burns, J. A., et al. (2013), Compositions and Origins of Outer Planet Systems:

1614 Insights from the Roche Critical Density, The Astrophysical Journal, 765, L28, 10.1088/2041-8205/765/2/L28.

1615

1616 Tittemore, W. C., & Wisdom, J. (1990), Tidal evolution of the Uranian satellites III. Evolution through the

1617 Miranda-Umbriel 3:1, Miranda-Ariel 5:3, and Ariel-Umbriel 2:1 mean-motion commensurabilities, Icarus, 85,

1618 394, 10.1016/0019-1035(90)90125-S.

1619

1620 Tosi, F., Turrini, D., Coradini, A., et al. (2010), Probing the origin of the dark material on Iapetus, Monthly

1621 Notices of the Royal Astronomical Society, 403, 1113, 10.1111/j.1365-2966.2010.16044.x.

1622

1623 Turrini, D., Politi, R., Peron, R., et al. (2014), The comparative exploration of the ice giant planets with twin

1624 spacecraft: Unveiling the history of our Solar System, Planetary and Space Science, 104, 93,

1625 10.1016/j.pss.2014.09.005.

1626

1627 Venturini, J., & Helled, R. (2017), The Formation of Mini-Neptunes, The Astrophysical Journal, 848, 95,

1628 10.3847/1538-4357/aa8cd0.

1629

1630 Waite, J. H., Perryman, R. S., Perry, M. E., et al. (2018), Chemical interactions between Saturn's atmosphere

1631 and its rings, Science, 362, aat2382, 10.1126/science.aat2382.

1632

1633 Wakeford, H. R., Visscher, C., Lewis, N. K., et al. (2017), High-temperature condensate clouds in super-hot

1634 Jupiter atmospheres, Monthly Notices of the Royal Astronomical Society, 464, 4247,

1635 10.1093/mnras/stw2639.

69 Fletcher et al., Ice Giant Systems

1636

1637 Wang, C., & Richardson, J. D. (2004), Interplanetary coronal mass ejections observed by Voyager 2 between

1638 1 and 30 AU, Journal of Geophysical Research (Space Physics), 109, A06104, 10.1029/2004JA010379.

1639

1640 Wei, Y., Pu, Z., Zong, Q., et al. (2014), Oxygen escape from the Earth during geomagnetic reversals:

1641 Implications to mass extinction, Earth and Planetary Science Letters, 394, 94, 10.1016/j.epsl.2014.03.018.

1642

1643 Witasse, O., Sánchez-Cano, B., Mays, M. L., et al. (2017), Interplanetary coronal mass ejection observed at

1644 STEREO-A, Mars, comet 67P/Churyumov-Gerasimenko, Saturn, and New Horizons en route to Pluto:

1645 Comparison of its Forbush decreases at 1.4, 3.1, and 9.9 AU, Journal of Geophysical Research (Space Physics),

1646 122, 7865, 10.1002/2017JA023884.

1647

1648 Wong, M. H., Tollefson, J., Hsu, A. I., et al. (2018), A New Dark Vortex on Neptune, The Astronomical Journal,

1649 155, 117, 10.3847/1538-3881/aaa6d6.

1650

1651 Zarka, P., & Pedersen, B. M. (1986), Radio detection of uranian lightning by Voyager 2, Nature, 323, 605,

1652 10.1038/323605a0.

1653

1654 Zarka, P., Pedersen, B. M., Lecacheux, A., et al. (1995), Radio emissions from Neptune., Neptune and Triton,

1655 341.

1656

1657 Zhang, Z., Hayes, A. G., Janssen, M. A., et al. (2017), Exposure age of Saturn's A and B rings, and the Cassini

1658 Division as suggested by their non-icy material content, Icarus, 294, 14, 10.1016/j.icarus.2017.04.008.

1659

1660

70

Declaration of interests

☒ The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

☐The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: