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: