Sentinel-2 and Sentinel-3 Mission Overview and Status

Craig Donlon – Sentinel-3 Mission Scientist and Ferran Gascon – Sentinel-2 QC manager and others @ESA

IOCCG-21, Santa Monica, USA 1-3rd March 2016 Overview

– What is Copernicus? – Sentinel-2 mission and status – Sentinel-3 mission and status Sentinel-3A OLCI first light Svalbard: 29 Feb 14:07:45-14:09:45 UTC Bands 4, 6 and 7 (of 21) at 490, 560, 620 nm Spain and Gibraltar 1 March 10:32:48-10:34:48 UTC California USA 29 Feb 17:44:54-17:46:54 UTC What is Copernicus?

Space Component

In-Situ Services A European Data Component response to global needs 6 What is Copernicus? – Space in Action for You!

• A source of information for policymakers, industry, scientists, business and the public • A European response to global issues: • manage the environment; • understand and to mitigate the effects of climate change; • ensure civil security • A user-driven programme of services for environment and security • An integrated Earth Observation system (combining space-based and in-situ data with Earth System Models)

Components & Competences

Coordinators: Partners:

Private Space Industries companies Component National Space Agencies

Overall Programme EMCWF EMSA Mercator snd Ocean Coordination FRONTEX Service Services operators Component EUSC EEA JRC

In-situ data are supporting the Space and Services Components Copernicus Funding

Funding for Development Funding for Operational Phase until 2013 (c.e.c.): Phase as from 2014 (c.e.c.):

~€ 3.7 B (from ESA and € 4.3 B for the EU) whole programme (from EU)

~€ 3 B ~€ 0.7 B for space for services

€ 2.3 B Space Component funding by ESA (ESA’s entrusted tasks) € 3.1 B € 0.8 B by EU (FP7, GIO and Grant) Planned Sentinel Schedule

Etna slopes sentinel-2 Sentinel-2 Mission Overview

• Spacecraft: 2 operating in twin configuration • Orbit: Sun-synchronous at 786 km (14+3/10 revs per day), with LTDN 10:30 AM • MultiSpectral Instrument (MSI): operating in pushbroom principle, filter based optical system • Spectral bands: 13 (VIS–NIR–SWIR spectral domains) • Spatial resolution: 10m / 20m / 60m • Swath: 290 km

Sentinel-2 A/B/C/D Sentinel-2 Second Generation A/B Sentinel-2 swath & coverage

SPOT5 60 x 60 km2 High revisit time (5 days at equator)  assured by twin 2 IRS P6 LISS III 141x141 km satellite observations Landsat ETM+ 180 x 172 km2 performed over a very large swath

Sentinel-2 290 x 290 km2 !!

Geographical Coverage: • All land masses 56° S bis 83° N incl. major islands All EU Islands < 20 km off the coast • All Mediterranean • Inland waters and all closed seas Satellite and Instrument

MultiSpectral instrument Satellite • Filter based push broom imager (280 kg, 1 m3) • Satellite mass: 1200 kg • Three mirrors silicon carbide telescope, with • Satellite power consumption: 1250 W dichroic beam splitter • Hydrazine propulsion system (120 kg - including • Focal plane arrays: Si CMOS VNIR detectors, provision for safe mode, debris avoidance and EOL HgCdTe SWIR detectors. orbit decrease for faster re-entry) • Onboard wavelet compression (divided by 3) • Accurate AOCS based on multi-head Star Tracker and fiber optic gyro • Integrated video & compression electronics (state of the art wavelet compression) • X band mission data distribution (520 Mbits/sec) • Radiometric resolution 12bits • Mission data onboard storage: 2.4 Tbits • Daily generated telemetry: 1.4 TB Sentinel-2 Mission Highlights

• 2 Satellites in twin formation, • Sun-synchronous orbit at 786 km (14+3/10 revs/day), with LTDN 10:30 AM • Revisit: 5 days at equator (with 2 satellites) under same viewing conditions;

• Multispectral Instrument: pushbroom with 13 bands in the VNIR and SWIR

• High spatial resolution: 10m, 20m and 60m;

• Wide field of view: 290 km • Duty cycle: average 17 min/orbit, maximum 32 min/orbit • Lifetime: 7.25 years, extendable to 12 years Sentinel-2A Launch last night!

June 23rd 2015 Sentinel-2 Products

Name High-level Description Production Preservation Volume Strategy Level-1B Top-of-atmophere Systematic Long-term ˜27 MB radiances in sensor (each 25x23km2) geometry

Level-1C Top-of-atmosphere Systematic Long-term ˜500 MB reflectances in cartographic (each 100x100km2) geometry

Level-2A Bottom-of-atmosphere On user N/A ˜600 MB reflectances in cartographic side* (each 100x100km2) geometry (prototype (using product) Sentinel-2 Toolbox**)

*: The possibility of a systematic global production of L2A is currently being explored. **: https://sentinel.esa.int/web/sentinel/toolboxes/sentinel-2

Level-1C / Algorithm

Level-0 Level-0 Level-1A Level-1B Level-1C Consolidated

TELEMETRY ANALYSIS DECOMPESSION RESAMPLING RADIOMETRIC - Geometry interpolation CORRECTIONS grid computation, - Resampling (B-splines). - Inv. on-board equalization, - Dark signal correction, - Blind pixels removal, - Cross-talk correction, PRELIMINARY QUICK- SWIR PIXELS - Relative response correction, LOOK AND CLOUD REARRANGEMENT - Defective/no-data correction, MASK GENERATION - Deconvolution/Denoising, CONVERSION TO - Binning of 60m bands. REFLECTANCES

PREVIEW IMAGE AND MASKS GENERATION GEOMETRIC VIEWING (defective pixels, cloud & MODEL REFINEMENT land/water)

- Refining of the viewing model using a global set of reference images, - Registration between VNIR and SWIR focal planes (optional). algorithms developed in cooperation with S2 assumption on coastal acquisitions 20 km offshore 280 km swath – lots of data.

https://sentinels.copernicus.eu/web /sentinel/missions/sentinel- 2/acquisition-plans Products Qualification On-going

data quality report on-line athttps://sentinels.copernicus.eu/documents/247904/685 211/Sentinel-2+Data+Quality+Report Level-1 Products Pre-Qualification

 Absolute geolocation performances (without geometric refinement) measured over 17 test sites.  Measurements in line with requirements.

Level-1 Products Pre-Qualification

 Multi-spectral registration performances measured show that the mean circular error over all band couples and detectors is lower than 0.23 pixel of the coarser band.

Level-1 Products Pre-Qualification

 Signal-to-Noise Ratio (SNR) calculated from images of the MSI sun diffuser.  Measured SNR values largely exceeding requirements. REQUIREMENTS S2-MPC RESULTS

SNR@Lref Lref SNR Margin Band/Unit - W/m2/Sr/μm - % B01 129 129.0 1016,50 688 B02 154 128.0 201,90 31 B03 168 128.0 228,60 36 B04 142 108.0 214,50 51 B05 117 74.5 238,50 104 B06 89 68.0 206,10 132 B07 105 67.0 208,80 99 B08 174 103.0 208,10 20 B8A 72 52.5 153,10 113 B09 114 9.0 164,70 44 San Francisco Bay Coastal Monitoring

Algae bloom, Baltic Sea

Venice boats

breakwater

surface effects

Courtesy K. Ruddick, D. V.d. Zande, RBINS

Copernicus Sentinel data 2015, RBINS processing Coral Reefs Monitoring London Wind Turbine Array (E. Channel)

(Quinten Vanhellemont & Kevin Ruddick, RBINS) 490nm native 10m resolution Hartog/Dorre Island, West Australia Sentinel–3 Mission Overview

• Operational mission in high-inclination, low Earth orbit • Full performance achieved with 2 satellites in orbit (S-3A,-3B)

Topography Mission Payload Optical Mission Payload providing providing  Sea surface topography data,  Sea and land color data, through a Topo P/L including a Ku-/C- through OLCI (Ocean and band Synthetic Aperture Radar Land Color Instrument) Altimeter (SRAL), a bi-frequency  Sea and land surface MicroWave Radiometer (MWR), and temperature, through the a Precise Orbit Determination SLSTR (Sea and Land (POD) including Surface Temperature . GNSS Receiver Radiometer) . DORIS . Laser Retro-Reflector

In addition, the payload design will allow  Data continuity of the Vegetation instrument (on SPOT4/5),  Enhanced fire monitoring capabilities Sentinel-3a launch from Plesetsk Cosmodrome 16th February 2016 Sentinel-3a launch from Plesetsk Cosmodrome 16th February 2016

(Credit: Antero Isola) Sentinel-3: Satellite Orbit details

S3B has a 180° phase separation on the same orbital plane

Instrument Swath Patterns Ground Track Patterns S3-A S3-B SRAL tracks at the equator: S3A = 104 km track separation S3A+B = 52 km separation 2 days

SRAL (>2 km) and MWR (20 km) nadir track

1400 km SLSTR (nadir) 1 Repeat Cycle 740 km SLSTR (oblique) 1270 km OLCI (27 days) SRAL orbit drivers: • Ground track repeatability, Orbit type Repeating frozen SSO • Dense spatial sampling Repeat cycle 27 days (14 + 7/27 orbits/day) Orbit control requirement: LTDN 10:00 • Ground track dead-band ±1km Average altitude 815 km Inclination 98.65° Sentinel-3 Optical Coverage (2 satellites with 180° phase separation)

OLCI mean 2.0 revisit time

with 2 Days satellites

0.5

1.0 SLSTR nadir- view mean revisit time Days with 2 satellites

0.2 S3 OLCI: Technical details

Basic configuration similar to MERIS: • 5 Camera Optical Sub Assemblies (COSA), • 5 Focal Plane Assemblies (FPA), • 5 Video Acquisition Modules (VAM), • 1 Scrambling Window Assembly (SWA), • 1 OLCI Electronic Unit (OEU) managing all the instrument functions, • 1 calibration assembly allowing radiometric and spectral calibration. Optical layout Ocean and Land Colour Imager

λ center Width Compared to MERIS: MERIS Bands Yellow substanace/detrital 412.5 10 100% overlap with SLSTR pigments • Chl. Abs. Max 442.5 10 • More spectral bands (from 15 to Chl & other pigments 490 10 21): 400-1020 nm Susp. Sediments, red tide 510 10 Chl. Abs. Min 560 10 • Broader swath: 1270 km Suspended sediment 620 10 Chl. Abs, Chl. fluorescence 665 10 • Full res. 300m acquired Chl. fluorescence peak 681.25 7.5 systematically for land & ocean Chl. fluorescence ref., Atm. Corr. 708.75 10 Vegetation, clouds 753.75 7.5 • Improved characterization, e.g. O2 R-branch abs. 761.25 2.5

straylight, camera boundary O2 P-branch abs. 778.75 15 characterization Atm corr 865 20

Vegetation, H2O vap. Ref. 885 10 • Timeliness: 3 hours NRT Level 2 H O vap., Land 900 10 product 2 New OLCI bands λ center Width Aerosol, in-water property 400 15 Reduced sun glint by camera tilt in Fluorescence retrieval 673.75 7.5 Atmospheric parameter 764.375 3.75 west (12.6°) Cloud top pressure 767.5 2.5 Atmos./aerosol correction 940 20 Atmos./aerosol correction 1020 40 Multi-sensor time series Band Set of OLCI in the Visible and the Near Infra-Red

SLSTR (After S. Sathyendranath) 555 659 865 • Higher spectral resolution than all previous sensors: Important for atmospheric correction, complex coastal waters, phytoplankton types • Consistency with MERIS: facilitates merging (no need to do band-shifting to establish inter-sensor biases) Sentinel-3 OLCI Basic Geometry

~09:00 LST 10:00 LST ~10:30 LST

1270 km OLCI Normalized Spectral Response Functions

CDR

SRD

Same behavior as for MERIS

FWHM for Oa21 will not be 40nm but 23nm due to optical transmission and detector response. Note, SNR performance remains compliant with large margins. (SNR EoL prediction: 197 vs Spec: 152) https://sentinels.copernicus.eu/web/sentinel/technical- guides/sentinel-3-olci/olci-instrument/spectral-response- (Pelloquin, C. and J. Nieke. 2012, OLCI and SLSTR simulated spectral response function-datafunction, S3 Calval WS, Oct-2012) OLCI instrument and camera test & characterisation sequence S3A-OLCI NEdL and SNR

1.E+03 S3A-OLCI* SNR shows 2000 excellent consistency

] 1 - 1.E+02 of intra-camera m

µ (different spectral . 1500 1 -

r 1.E+01 bands, spectra or full s SNR .

² ] -

- and reduced resolution Lref [

m R . 1.E+00

NeDL Typical 1000 N modes), and inter- S W [

SNR typical camera measurements.

L 1.E-01

D E 500 N

A good correlation , 1.E-02 L between the test results and the SNR 1.E-03 0 350 450 550 650 750 850 950 1050 model was established Wavelength [nm] *Measurements made prior to swap of cameras in June 2015. We expect similar performance for the Flight Model. OLCI and SLSTR diffuser material

OLCI: UV exposure and Hemi-reflectance stability (S3-RP-CSL-OLCI-09035) OLCI and SLSTR diffuser material:

• Same material for OLCI and SLSTR flat panel diffusers • Zenith Polymer®, an optical PTFE • Produced by SphereOptics/ElringKlinger AG, Germany • Production process according to “Space- Grade conditions” i.e., cleanness is closely monitored (e.g., PTFE raw material (powder) bake out at 24hrs, <1mbar, 95deg)

• Environmental conditions (thermal cycling, UV exposure, Radiation, ageing) were verified in a delta-qualification. • Strict contamination control during AIV. • All values are within requirements and Estimated total duration of sun illumination within budget. for OLCI diffusers over mission lifetime OLCI spectral calibration

Within OLCI a maximum of 3 peaks each with 15 spectral lines can be used.

100 Better Characterisation90 than for MERIS 80 Erbium-doped 70 Zenith 8% 60 Reflectivity (%) Reflectivity « pink diffuser » 50 40 350 450 550 650 750 850 950 1050 1150 Wavelength (nm) 3 absorption peaks are proposed for nominal spectral calibration

Peak @409 nm Peak @ 522 nm Peak @ 800 nm

100 100 100 90 90 90 80 80 80 70 70 70 60 60 60

Reflectivity (%) Reflectivity 50 Reflectivity (%) Reflectivity

Reflectivity (%) Reflectivity 50 50 40 40 40 30 30 400 404 408 412 416 420 515 520 525 530 535 790 795 800 805 810 Wavelength (nm) Wavelength (nm) Wavelength (nm) MERIS absorption bands

3 other peaks optional: 490 nm, 658 nm, 975 nm.

today

Level 1 pre-qualified core product release. Gradual ramp-up of operations and progressive release of level 2 pre-qualified core products Core products availability

Data provided to Reference sets commissioning of core products phase teams released to validation In addition, sample core products will be team/ expert released to all users as early as possible for users familiarisation ESA & EUMETSAT SHARE OPERATIONS

 EU Copernicus Regulation: full, open and free data policy, defining responsibilities for ESA and EUMETSAT and overall financial envelope  Dedicated EU-ESA and EU-EUMETSAT Copernicus agreements Operational Core Products - full technical documentation at https://sentinel.esa.int

– The operational core products description (including ATBD) is available at: http://sentinel.esa.int

Open and Free data access policy https://sentinels.copernicus.eu https://scihub.copernicus.eu/ Sentinel Data – Online Access

Online data access at: scihub.copernicus.eu

– ESA Data Hub Software (DHuS) provides an open source Web Interface

– Users can set scripts to automatically download data Scientific Toolboxes Sentinel 3 Toolbox

Sentinel-3 Toolbox:

• Visualisation & processing of Sentinel-3 OLCI and SLSTR data and other optical data • Uncertainty visualization and exploitation • Remote in-situ database access • Synergistic use of OLCI and SLSTR • Various OLCI and SLSTR data processors http://step.esa.int/

52

Application Modes I II III IV Interactive bulk / NRT EO data Cloud Exploitation Exploration processing processing centre Platform

SNAP SNAP gpt:_ gpt:_

SNAP Engine SNAP Engine:_

PC, notebook PC, notebook, tablet cluster, cloud server PC, notebook, tablet ESA Sentinel Satellite Smartphone App PRAGUE 09-13 MAY 2016

Main Objective: Presentation of Exploitation Results based on ESA Earth Observation Measurements

Important Dates: Themes: Deadline for abstract submission 16 October 2015 Atmosphere, Oceanography, Cryosphere, Notification of Acceptances End January 2016 Land, Hazards, Climate and Meteorology, Issue of Preliminary Programme February 2016 Solid Earth/Geodesy, Near-Earth Opening of Registration February 2016 Environment, Methodologies and Products, Release of the Final Programme at the symposium Open Science 2.0 Submission of Full Papers at the symposium http://lps16.esa.int Thank You – any Questions

Contact: [email protected] Sentinel-3 Toolbox

• Ingesting Sentinel-3 OLCI and SLSTR • Ingesting MERIS, AATSR, ATSR, MODIS, SeaWiFS, VIIRS, Proba-V, SPOT-VGT, Landsat TM, etc. • Basic functions: band maths, statistical analysis, spectrum analysis, geometric tools … • Uncertainty information exploitation – OLCI and SLSTR error variables – Various visualisation modes (blending and overlays) – Propagation build into band-maths (Standard Combined Uncertainty, GUM 95) • Remote in-situ database access tool (MERMAID for MERIS/OLCI) • OLCI / SLSTR / SYN Data Processors – OLCI/SLSTR collocation tool – Pixel classification (cloud screening) – Atmospheric correction and in-water processing (FLH-MCI, C2R, CoastColour, FuB,..) – SST & LST processing – Land thematic processing (fAPAR) • Commissioning phase support tools – OLCI/SLSTR collocation accuracy assessment (L1C product) – SLSTR PDU stitching tool ESA SEOM - Case2 Extreme Waters A study to improve the retrieval of water products in extreme scattering and absorbing coastal (and inland) waters, focus on S3 (primarily) and S2  Relevance/Applications: Water quality, Primary Production and C fixation, sediment transport, dredging + dumping, organic C from rivers, eutrophication, Rrver engineering  Problems to be addressed: ? • Atmospheric correction in turbid ? ? waters ? • CHL retrieval in high non-algae ? particle as well asd high CDOM ? loaded waters ? • Identification of water conditions, and selection of appropriate algorithm(s) for extreme ranges of water constituent concentrations • Large variability of SIOPs C2X Project: Team: Approach  Radiative transfer simulations, in-situ data from selected extreme water sites  Atmospheric correction • comparing & improving three different approaches (C2RCC - NN, polymer, NIR/SWIR) • exploiting new bands of S3 (O-400nm, O-1020nm, SLSTR-SWIR) • study applicability to S2 MSI  In-water algorithms • comparing & adapting candidate algorithms (spectrum inversion, semi-analytical, regression) • exploiting O-400nm band for absorbing waters and O-1020nm band for scattering  Prototyping and validation • match-ups and decorrelation analysis • list of demonstration sites, linked to local PIs • S3 OLCI & SLSTR in 2016  Link with international activities: IOCCG WGs, international science support team, ESA OC CCI, GEO Comm. Practice, … . Duration: April 2015 – March 2017

Marine Primary Production: Model Parameters from Space (MAPPS)

T.Jackson, H.Bouman, S.Sathyendranath, T.Platt, F.Melin

Objective: improve estimates of Marine Primary Production through the production of remote-sensing data based estimates of phytoplankton photosynthesis- irradiance parameters. contributors and details available at New database on photosynthesis irradiance www.global-maps.org/data parameters from across the globe, which will be published and made accessible through ESSD.

Inter-model comparison and model improvement and optimisation Example of model at SST of 22oC SST CHL PAR

Demonstration products should be available soon

Global satellite products of photosynthesis- irradiance parameters Ocean-Colour Component of ESA’s Climate Change Initiative

Objective: produce an uncertainty-characterised, inter- sensor bias-corrected, merged time series of ocean-colour May 2010 bias, log_10 Chl products for climate research, and engage with users V2 of the merged time series (SeaWiFS, MERIS and MODIS- A) released in March 2015. Specific aims of this version 2.0 release: • improves the in situ database used for uncertainty May 2010 RMSD, log_10 Chl characterisation • optimizes the uncertainty generation for the CCI data • improves consistency in many areas, including unifying the binning/mapping processing • improves bias correction, able to respond to temporal variation (primarily seasonal) • incorporates an improved cloud mask for MERIS Relative error in V2 Chl based on • benefits from a more automated quality assurance bias.

process Red vertical line: GCOS • extends the time series to the end of 2013 requirement for • refreshes the input datasets to the latest versions accuracy OC-CCI: Improved OC-CCI: Improved coverage in many uncertainty under-sampled regions characterisation in V2 compared with V1 OC-CCI:Future Plans

• Incorporate VIIRS into Bias in log_10 chlorophyll, per optical class the time series • Extend time series to 2015+ 800 Match-up in situ observations, per optical class • Prepare for Sentinel-3 400 • Improve Case-2 0 products RMSD in log_10 chlorophyll, per optical class

Acknowledgements ESA and OC-CCI thank the many members of the ocean-colour community who helped in many ways: validation data, participation in user consultation; feedback on products. A special thanks to NASA for their continued help and support. OC-CCI: Second version released in early 2015 Version 3 scheduled for release imminently Processing step Version 1 Version 2 Input datasets Latest versions available prior to Latest versions available prior to release in release in 2013 2015 Atmospheric MERIS: POLYMER V2 MERIS: Improved POLYMER V3 correction In situ database Initial version Extended version with substantial increase in match-ups with satellite data Binning Beam Binner for MERIS; L2Gen for Beam Binner for all sensors, improving SeaWiFS and MODIS- A consistency in this step Bias correction Static correction per pixel Improved bias correction; accounts for seasonal variation in bias Cloud masking Initial version of Idepix used for Improved cloud mask (Idepix 2.0) for MERIS; MERIS; : L2gen for SeaWiFS and no change for SeaWiFS and MODIS-A: L2gen MODIS-A Uncertainty Underlying optical classification Optical classification custom generated for generation based on in situ database on Rrs OC-CCI Version 2 satellite data Uncertainty Based on initial optical classification Benefited from OC-CCI-data-based optical characterisation and initial in situ database classification, and improved in situ database Quality Assurance Initial version, less automated More automated quality assurance process Time series Sep 1997 to Dec 2012 Sep 1997 to Dec 2013 OC-CCI: Work continues Match-up validation In situ database Uncertainty characterisation

Trends

Data assimilation into ecosystem models Pools of Carbon in the Ocean Project: Rationale

• At the interface between marine ecosystem and climate models and Earth Observation is the carbon-to-chlorophyll ratio, which is a highly variable quantity.

• In 2010, GEO produced a “Carbon Strategy Report”. In response, the CEOS produced a document on “Carbon from Space” (CEOS, 2014). Both GEO and CEOS Reports highlight the importance of carbon products. Main focus of POCO: Particulate Pools

Bigger More • Total Particulate Carbon Measure- pool data ments

• Particulate Organic Carbon more Smaller Less difficult • Phytoplankton Carbon pool data

Objectives • explore the potential of remote sensing for detecting particulate carbon pools in the ocean • compare with models • focus on climate studies Scientific Achievements:

• Algorithm selection: 10 Carbon algorithms reviewed and assessed • In situ database: ~36,000 POC data points assembled

• Model/carbon data interaction: Data OC-CCI Chl-a mg m-3 POC Stramski et al. 2008 assimilation and comparison of MIT Behrenfeld et al. 2005 Sathyendranath et al. 2009 model with EO • Uncertainty quantification: ~3,600 POC match-up points

• Some initial sresults on this slide Phytoplankton Carbon mg m-3 Phytoplankton Carbon Mg m-3 ) ) -3

Dutckiewitz et al . (2015)

Number of Match-ups Number of Match-ups Website: www.oceancarbon.net -3 POC, In Situ Stramski et al. 2008 m (mg al.2008 et Stramski POC POC (in situ) (mg m ) Twitter (@ESA_POCO) Colour and Light in the ocean from Earth Observation (CLEO) workshop 6-8 Sep 2016 ESA /ESRIN Frascati, Italy

The workshop is sponsored through selected SEOM projects, including: • Pools of Carbon in the Ocean • Photosynthetically Active Radiation and Primary Production • Extreme Case-2 Waters

Additional partner projects of ESA are: • Marine Photosynthesis Parameters from Space • Ocean Colour Climate Change Initiative • Synergistic Exploitation of Hyper-and Multispectral Sentinel-Measurements to Determine Phytoplankton Functional Types http://congrexprojects.com/2016-events/cleo

Please participate and help make workshop a success