The COASTALT Project: Towards an Operational Use of Satellite Altimetry in the Coastal Zone

Size: px

Start display at page:

Download "The COASTALT Project: Towards an Operational Use of Satellite Altimetry in the Coastal Zone"

Clemence Walsh
5 years ago
Views:

The COASTALT Project: Towards an Operational Use of Satellite Altimetry in the Coastal Zone Stefano Vignudelli (CNR, Italy) Paolo Cipollini*, Scott Gleason, Christine Gommenginger, Helen Snaith

1 The COASTALT Project: Towards an Operational Use of Satellite Altimetry in the Coastal Zone Stefano Vignudelli (CNR, Italy) Paolo Cipollini*, Scott Gleason, Christine Gommenginger, Helen Snaith (NOCS, UK) Henrique Coelho (Hidromod, Portugal) Joana Fernandes (U Porto, Portugal) Jesus Gómez-Enri (U Cádiz, Spain) Cristina Martin-Puig (Starlab, Spain) Phil Woodworth (POL, UK) Salvatore Dinardo (SERCo/ESRIN, Italy) Jérôme Benveniste (ESA/ESRIN, Italy) + the whole COASTALT crew! * Project coordinator cipo@noc.soton.ac.uk

to get our hands on Because of land effects + calm water effects on waveforms, inaccurate

2 Rationale for coastal altimetry Satellite altimetry designed for open ocean BUT coastal region has enormous strategic importance 17 years of data over the coastal ocean are still to get our hands on Because of land effects + calm water effects on waveforms, inaccurate corrections, etc normally flagged as ʻbadʼ in the official products - but they can be recovered! 2

Many possible uses sea level, currents, waves not only long term studies and climatologies, but also specific hazardous events (surges) Assimilation into coastal models Coastal tides

3 Many possible uses sea level, currents, waves not only long term studies and climatologies, but also specific hazardous events (surges) Assimilation into coastal models Coastal tides Coastal Altimetry is a legitimate component of coastal observing systems See Cipollini et al., Community White Paper, OceanObsʼ09 Example of Coastal Tidal Model, MOHID run by Hidromod, Lisbon 3

4 How to recover the information A. Specialized retracking Use better waveform models, accounting for change of shape in coastal environment (next slide) Use specialized (2-D or sequential) retracking techniques B. Improved Corrections Most crucial is the correction of path delay due to water vapour ( wet tropospheric correction) Some applications require correction of tidal and high-frequency signals, which are also difficult to model in the coastal zone 4

5 Example of coastal waveforms Pass direction ENVISAT pass 294 (descending) Cycle 39 Bonifacio Strait Cycle 40 Capraia Island Capraia Bonifacio Strait Cycle 41 5

6 Example of coastal waveforms Capraia Island ENVISAT pass 294 (descending) 6

7 Example of coastal waveforms Capraia Island 7

8 Example of wet tropospheric problems South Africa Effect of land S Africa J. Maina Mbui for ALTICORE-Africa 8

The Coastal Altimetry Community In the last few years, several groups have started tackling these problems Coastal Altimetry Workshops (Silver Spring 2008, Pisa 2008, Frascati 2009, Porto 2010?

9 The Coastal Altimetry Community In the last few years, several groups have started tackling these problems Coastal Altimetry Workshops (Silver Spring 2008, Pisa 2008, Frascati 2009, Porto 2010?) Community White Paper at OceanObsʼ09 Space Agencies do support these efforts: NASA/CNES via some OST-ST projects CNES funded PISTACH Project (2007/09) ESA funds COASTALT (Phase1: 2008/09) In this talk we will see COASTALT objectives, main results and recommendations 9

10 COASTALT - objectives COASTALT aims to lead to the definition, specification and prototyping of a new pulse-limited radar altimetry coastal zone product. In COASTALT this is done over a number of study regions: NW Mediterranean (incl Corsica Channel) West Britain Portugal Coast This new product is the seed for future routine reprocessing of coastal altimetry data by ESA including the reprocessing of whole ESA Radar Altimetry archive (ERS-1, ERS-2, ENVISAT) exploitation of CryoSat and Sentinel-3 over the coastal zone PISTACH focuses on NASA/CNES Jason-1, Jason-2 instead 10

11 Results User requirements & products Joint PISTACH/COASTALT Survey in 2008 Modelling community is an important user community Wind and Waves matter to a good share of the users in addition to SSH Some scope for near-real-time or even realtime delivery of coastal altimetry Need clear quality flags together with all the separate corrections Need good documentation doing NetCDF product specification and user handbook 11

Results Corrections: the way forward Example: Difference between a local tidal model and a global one (GOT00) over the White Sea (courtesy of S. Lebedev / A.

12 Results Corrections: the way forward Example: Difference between a local tidal model and a global one (GOT00) over the White Sea (courtesy of S. Lebedev / A. Sirota for ALTICORE) COASTALT reviewed the whole spectrum of corrections Wet Tropospheric correction: DLM (Dynamically Linked Model) approach: Use models to extend radiometer observations Model/remove land effect (developed by PISTACH) GPS-based wet tropo Ionospheric corrections Extend dual-freq open ocean corrections using GIM model (based on GPS) IB and HF dealiasing Investigate and use local models Also need better data screening and editing

beginning and end of the land contaminated segment, by using a linear adjustment (using time as interpolation coordinate) 2.

13 Results Wet Tropo DLM technique Two types of algorithm 1. Island type or ʻdouble-endedʻ valid radiometer points on each side of the segment Model field is adjusted to the radiometer field, at the beginning and end of the land contaminated segment, by using a linear adjustment (using time as interpolation coordinate) 2. Continental coastline (ʻsingle-endedʼ) only valid radiometer points on one side of the segment Model field is adjusted to the radiometer field, at the beginning or at the end of the land contaminated segment, by using a bias correction 13

14 Results DLM Example of Results Model DLM 14

15 Results Wet tropo - GPD technique GNSS (GPS) based First, estimate STD - Slant Total Delay from GPS observations Then, map into ZTD - Zenith Total Delay using MIT GAMIT software and Vienna mapping functions Estimate ʻdryʼ component (ZHD) from meteo data derive ʻwetʼ component (ZWD) (For Validation) Needs a good network of coastal stations, and possibly coincident meteo observations ZTD accuracy currently ~3 mm. Wet accuracy depends on how good the ʻdryʼ is but we can get to less that 1 cm. (for validation only, otherwise the ZTD is used) 15

16 Results GPD example of results ZWD (Wet tropo) for Envisat Pass 160 (Cycles 59 and 62) 16

17 Results Comparison with Tide Gauges Median Absolute Deviations TG altimeter (corrected with the various techniques) M.J. Fernandes et al., subm. to IEEE GRSL Conclusion (over a number of TGs): GPD does as well as DLM (and better than ECMWF model-based correction) - and there is room for improvement (more stations, new mapping functions, etc) 17

18 Results Waveforms in Coastal zone Study of waveforms in coastal zone and island passes We observed effects of land and effects of calm waters in the coastal strip Land normally gives ʻdarkʼ features (less signal) Calm water cause quasi-specular reflections bright features These features migrate in the waveform/gatenumber space following hyperbolae Features are reproduced by a simple model of the land/ ocean/calm waters response The idea is that this should allow removal of the land/calm waters contamination prior to retracking Innovative retrackers (Bayes, 2-D): retrack each waveform not in isolation but using info from adjacent ones 18

cycle 49, bright target due to wave sheltering in NW bay

19 Results Example Pianosa Island 3 km Gate no. observed Flight direction Envisat Ascending track 128 In cycle 49, bright target due to wave sheltering in NW bay (Golfo della Botte) J. Gómez-Enri et al., IEEE GRSL, in press simulated

20 Results The COASTALT Processor Core deliverable of the Project Flexible, configurable and modular Baseline processor coded in C (I/O) and FORTRAN (retracking) Read L2 SGDR files Least-square fitting (weighted or unweighted) Brown, Specular and Mixed waveform models Generate NetCDF output files Coastal Geophysical Data Records (CGDRs) User-defined Geophysical Corrections module Allows addition of any user-generated correction CGDRs being tested and validated 20

21 Results Processor block diagram 21

22 COASTALT: the future Need for continuous maintenance/update/development of processor and products Improve CGDRs, add more fields Add new retrackers (including 2-D, Bayes) Continue research on corrections and roll new ones out GPD Wet Tropo is main potential improvement Also need more testing of local tidal/hf models Validation on the three pilot areas Preliminary validation carried out but need more mature products Ideally extend to other places (India, S Africa, etc) Outreach / contribute to capacity building in Coastal Altimetry Drafted dedicated tutorials for the Basic Radar Altimetry Toolbox (BRAT) will be implementef in near future 22

Conclusions Coastal Altimetry important new field of research COASTALT Project has delivered: Product recommendation, product definition, user handbook Recommendation on corrections DLM and GPD Wet

23 Conclusions Coastal Altimetry important new field of research COASTALT Project has delivered: Product recommendation, product definition, user handbook Recommendation on corrections DLM and GPD Wet Tropo Techniques latter is particularly promising Analysis of coastal waveforms Preliminary design of innovative retracking techniques COASTALT processor up and running Coordination work: CA workshops, COASTALT-SWT 23

24 . and good news for the future ESA is extending the support for a Phase 2 of COASTALT for 2010 This will allow to capitalize on the R&D investment made in Phase 1, and in particular: Improve products & corrections, add more fields Develop and plug in innovative retrackers Perform full validation Release the coastal product to Users COASTALT will remain at the centre of the Coastal Altimetry community The project will co-organize 4 th Coastal Altimetry Workshop (Porto, Oct 2010? date/place TBC) 24

Pulse-Pair (Doppler) Processing of Envisat Individual Echoes

Pulse-Pair (Doppler) Processing of Envisat Individual Echoes R. Abileah 1, S. Vignudelli 2 1 jomegak, San Carlos CA, USA 2 CNR Istituto di Biofisica, Pisa, Italy Outline Envisat individual echoes (IE)