CryoSat Plus for Oceans: analysis of the state-of-the-art Marc Naeije 1, Christine Gommenginger 2, Thomas Moreau 3, Salvatore Dinardo 4, David Cotton 5, and Jérôme Benveniste 4 1 TUDelft, 2 NOC, 3 CLS, 4 ESA, 5 SatOC

CryoSat Plus for Oceans ESA’s CryoSat-2 mission is the first one to carry a radar altimeter that can operate in SAR mode Although the primary aim is land and marine ice monitoring, the SAR mode capability of the CryoSat-2 SIRAL altimeter offers potential benefits for ocean applications The “CryoSat Plus for Oceans” (CP4O) project exploits CryoSat-2 data over the ocean CP4O is supported by the ESA under the Support to Science Element (STSE) Programme, and by CNES, and brings together an expert European consortium, led by SatOC (UK)

CP4O project objectives Build a sound scientific basis for new applications of CryoSat-2 data over the open ocean, polar ocean, coastal seas and for sea-floor mapping Generate and evaluate new methods and products that will enable the full exploitation of the capabilities of the CryoSat-2 SIRAL altimeter, and extend their application beyond the initial mission objectives Ensure that the scientific return of the CryoSat-2 mission is maximised … and … be prepared for full exploitation of future SAR enabled altimeter missions like Sentinel-3 and Jason-CS

CP4O partners & themes Open Ocean Polar Ocean Coastal Zone Sea floor mapping Geophysical corrections

CP4O theme objectives Open Ocean Low Rate Mode: o Assess accuracy /continuity with respect to previous and concurrent missions SAR Mode: o RDSAR Mode Generate LRM like products from SAR mode Ensuring continuity from coastal zone to open ocean, & LRM to SAR mode o New SAR re-tracking schemes o Ability to map fine scale features Coastal Zone SAR Mode o Investigate ability to map fine scale coastal features o Approaches to minimise land contamination SARIN Mode o Potential to discriminate & mitigate contamination from off-nadir land targets

Theme objectives cont’d Polar Ocean LRM, SAR and RDSAR o Develop and evaluate processing schemes applicable to sea-ice affected regions: Mean Sea Surface, Mean Dynamic Topography, polar ocean circulation Improvements of polar tide models Sea-Floor Mapping SAR Mode o Investigate ability to resolve short-wavelength sea surface signals caused by sea-floor topography and to map uncharted sea- mounts/trenches o Critically depends on > 1 year continuous data over suitable region Geophysical corrections Ionosphere, wet troposphere, regional tide models

SAR vs pulse-limited t=0 t<0 t>0 RED LRM GREEN SAR 1D multi-looked 20Hz waveform 128 gates multi-looking

SAR advantages More independent looks lead to improved retrieval precision o Two-fold improvement according to numerical studies by Jensen & Raney (1998) Finer spatial resolution along track o ~300 meters along-track Higher SNR o ~10 db more Better performance close to land o especially for track ~90° to coastline Less sensitivity to sea state

9 THE FUTURE: Sentinel-3 Surface Topography Mission LRM: open ocean SAR: global coastal ocean + sea ice SAR open loop: ice sheet margins SAR open/closed loop: land & inland waters THE PRESENT: Cryosat-2 LRM: most of open ocean + most of land SAR: some oceanic & coastal regions + sea ice SARIN: few pilot regions (ocean & land) + land ice

SAR innovations Reduced SAR techniques (RDSAR): processing FBR data into LRM waveform data, and re-track producing pseudo LRM (PLRM) dealing with the LRM/SAR differences in PRF o SAMOSA approach processes subsets of waveforms, remaining issues subset optimization and time tag bias o CNES RDSAR from the Cryosat Processing Prototype (CPP) based on averaging all pulses from 4 SAR bursts (256 FBR echoes), CPP_LRM and CPP_RDSAR are routinely ingested in CNES/SALP products o NOAA/Altimetrics RDSAR as used in RADS based on stacking and averaging all individual SAR echoes CP4O compares and validates these techniques

SAR innovations cont’d ascending SAR descending 1-Hz values 20-Hz Std Dev Matching PLRM to LRM –Different timing biases –Updates to range –Improved backscatter –Improved SWH –New SSB model

SAR innovations cont’d Dedicated SAR altimetry (Delay-Doppler) processing, not only for CryoSat-2 but also to prepare for Sentinel-3 and Jason-CS missions (SAR waveform is very different from LRM waveform): o (semi-) analytical approach (Halimi). Initial 3 parameter approach not suited well for CryoSat because of mispointing, but a 5 parameters model (including mispointing) has been developed o Numerical re-tracking as done by CNES CPP, accounts for mispointing o Analytical SAMOSA-3 model, accounts for mispointing Different methods need further comparison and validation Sigma0 retrieval in SAR mode needs further development

SAR/SARIN re-tracking over Polar Ocean We compared 7 re-trackers for the Arctic area: o Traditional OCOG o Traditional Threshold o 5 Parameter Beta o Leading Edge (LE) + OCOG (R4) o Leading Edge (LE) + Threshold (R5) o Maxima as Threshold o ESA re-tracker (standard CryoSat-2 L2 products) Performance evaluated using gravity field method and SLA method leading to conclusion that applying re-tracking only on the leading edge of the returned waveforms significantly improves sea height

Polar re-tracking cont’d Hz standard deviations of SLA for R5 (LE+threshold), and R4 (LE+OCOG) retrackers, S is Summer and W is Winter

Corrections and models Presently used corrections and models on ESA product should be replaced by state-of-the art: o Wet troposphere : no onboard MWR available – use ERA model, MWR (altimeter and non-altimeter missions) and GNSS (coast/islands) - CP4O progress in developing optimal data combination technique (objective analysis) o Ionosphere : CryoSat single frequency altimeter –TEC from models and external sources like GNSS. JPL GIM current best global solution. Regionally this can be improved by regional ionosphere TEC from the Spectre project o Global Tides : Use high resolution empirical or hydrodynamical like DTU10, EOT10a, TPXO7.2, GOT4.8, or FES2012. Current RADS default GOT4.8, but not suited for shallow waters o Regional Tides : Use of regionally tuned higher resolution models like the COMAPI model for the NE Atlantic

Corrections cont’d

Models cont’d ModelsMain componentsLong-period tidesNon-linear components FES2004M2, S2, K1, O1, N2, K2, P1, Q1, S1, 2N2, SsaMf, Mm, MSqm, Mtm M4 TPXO7.2M2, S2, K1, O1, N2, K2, P1, Q1Mf, MmM4, MS4, MN4 GOT4.8M2, S2, K1, O1, N2, K2, P1, Q1, S1/M4 DTU10M2, S2, K1, O1, N2, K2, P1, Q1 + S1 from GOT4.7 /M4 from GOT4.7 EOT10aM2, S2, K1, O1, N2, K2, P1, Q1, S1, 2N2Mf, MmM4 COMAPI regional atlases M2, S2, K1, O1, N2, K2, P1, Q1, S1, 2N2 + Minor components: E2, J1, L2, La2, Mu2, Nu2, R2, Ro1, Sig1, T2 Mf, Mm, MSqm, Mtm M4, MS4, MN4, S4, M6, MK4, SN4, SK4, 2Q1, MP1, 2MK6, 2MN6, 2MS6, 2SM2, 2SM6, KJ2, MK3, MKS2, MO3, MSK6, MSN2, MSN6

State of the Art Known issues with ESA Cryosat-2 products (Baseline A and B) o Some will be addressed in new IPF for FD Marine Product o Need to resolve mispointing, time tag, tracking point issues o Effect of truncation of waveform trailing edge in Baseline B Does it change sensitivity of retrieved SSH to mispointing? impact on coastal applications (mitigate land signals) ? Other issues o Is there an effect of long waves, wave direction on SAR SSH and SWH ? o Spreading of the SAR leading edge (in baseline B) impacts C2 SAR retrieval accuracy o Sea State Bias model for SAR waveform re-tracking

State of the art cont’d Considering our analyses and the known problems in the CryoSat-2 data products published by ESA we have come to the conclusion that currently, for oceanographic applications, LRM L2 and SAR L2 are not useful and that SAR L1b is sub-optimal At the same time the CP4O team recognizes the exceptional performance of the SIRAL altimeter on CryoSat-2, and is confident that in due course the high- quality state-of-the-art products possible will be available from the ESA processing chain and CP4O hopes to contribute to this adequately Alternative data sets come from CNES CPP and RADS The state-of-the-art-activity of CP4O will produce a Preliminary Analysis Report and a Development and Validation Plan, which will be available in

Product development and validation The heart of CP4O is the development and validation of algorithms and processing schemes for new ocean products, based on CryoSat-2 data. This involves the creation of 7 new experimental altimeter data sets, and 4 data sets with new geophysical corrections These experimental data sets will cover limited regions and time periods (see: next sheet). Because the version of CryoSat data currently distributed by ESA has known anomalies for ocean applications initial work will use data from alternative processing chains, including the CNES CryoSat Processing Prototype (CPP) and the NOAA processed data available on RADS The outputs of the product development activity will be Algorithm Theoretical Basis Documents and Product Validation Reports for each processing scheme. These will be available in early autumn

Development cont’d Initial Development and Validation Large scale assessment 1LRM for Open OceanGlobal (RADS & CLS) 2RDSAR for Open OceanNE AtlanticGlobal 3SAR for Open OceanNE AtlanticGlobal 4SAR for Coastal OceanSouth Coast UKGlobal / regional Gulf of Cadiz, North- West Mediterranean & German Bight 5SARIn for Coastal OceanCuba, Chilean CoastN/A 6SAR for Polar OceanArctic (initially Baffin Bay) 7SAR for Sea Floor MappingPacific / North Pacific 8Improved wet trop correctionGlobal, full C2 mission 9Improved iono correctionMediterannean Sea, European continental shelf 10 Improved regional tidesNorth East Atlantic (coastal) 11Other improved correctionsGlobal (RADS)

Development: early results Jason-2 L2 LRM Cryosat-2 L1B SAR (Baseline A) Cryosat-2 L1B SAR (Baseline B) Cryosat-2 SAR (Hs < 2m) SSH std 20Hz: cm 1Hz: cm Jason-2 LRM (Hs < 2m) SSH std 20Hz: cm 1Hz: cm

French Polynesian Atolls J2 SWH C2 SAR SWH CryoSat-2 SAR (red) and Jason-2 (blue) data from Iles Palliser atolls in the Pacific, confirming the ability of SAR mode to retain track. French Polynesian Atolls Development: early results

CP4O: next steps Impact assessment o Based on the processing schemes and experimental data sets validated larger data sets will be generated and more widely evaluated, to analyse and quantify the impact and benefits of the new CryoSat-2 products. For this a “round robin” methodology will be applied similar to those applied in projects such as the ESA Sea Level Climate Change Impacts Project. The output will be an Impact Assessment Report, available early 2014 Future exploitation – scientific road map o Finally, the results of the Impact Assessment will be reviewed against the User Requirements Baseline generated early in the project and used to define an agreed Scientific Roadmap to ensure fullest possible exploitation of CryoSat-2 data over the oceans, and to support the transfer the results into scientific and operational activities

Final words SAR mode altimetry offers an exciting opportunity to oceanographers, opening up a number of potential new applications, as well as significant improvements to more well established products. Results will be shared with the oceanographic community as our project progresses. We plan to hold a number of open workshops at which we will present our latest findings and invite others working in this field to also present their work 1. We also intend to establish an expert group to provide an independent evaluation of our work. The ultimate aim is to maximise the exploitation of CryoSat-2 data in oceanographic applications, and to build a sound scientific foundation for future satellite missions carrying SAR mode altimeters, starting with Sentinel-3 in the near future. 1 First workshop opportunity is the peer-review of the mid term output of CP4O held at ESRIN (Frascati) – interested to join? Please, contact the presenter at lunch break