1 st Envisat Validation Workshop MERIS Conclusions and recommendations

1. 1st Envisat Validation Workshop MERIS Conclusions and recommendations

2. Plan Introduction Calibration In-flight Vicarious Verification Validation Cloud and Water Vapour Land Products Water Products Conclusion

3. Introduction • Instrument works well. • In-flight radiometric and spectral calibration performed regularly. Spectral calibration campaigns took place giving an in-flight spectral characterization results with unprecedented accuracy. • Verification activity was more demanding than originally expected but produced good results. • Following the verification phase, the vicarious calibration and the validation activities started. • First results have been presented during the workshop.

4. Calibration Team Conclusions • Second spectral campaigns confirm early results. • - Overall accuracy 0.2 nm, resolution 0.05 nm • Radiometric calibration results showed a little degradation in some camera. (below 2.5%) • Vicarious calibration results (based on field measurements) showed good agreement (within the accuracy of the method used). • Inter comparison with AATSR showed good agreements. • Inter comparison results over desert sites showed an over estimation of Meris compared to last century’s sensors.

5. Calibration Team Recommendations • Spectral characterization • Repeat spectral characterization campaigns every six months • Improve the instrument model using calibrated data • Instrument Degradation: • Include a pixel-wise degradation model in the L1b processing • Run a diffuser ageing calibration in SciLo configuration (planned) • BRDF modelling • Improve the BRDF model based on in-flight data • Develop a diffuser degradation model • Study the Speckle effect • Run a vigneting check for diffuser illumination extremes (planned)

6. Calibration Team Recommendations • Radiative Transfer Codes • Inter-compare the Radiative Transfer Codes used for Vicarious Calibration • Tool Box Improvements • Include a L1b conversion to TOA reflectance & TOA normalized radiance • Include the RTC in the Toolbox for vicarious calibration exercises • METRIC Improvements • Include Snow, Buoys, instrumented site extraction • Include IOCCG Diagnostic data sites extraction • Calibration Plan • Systematic acquisition of data over designated targets and distribution via FTP • Hold Regular Cal/Val workshops

7. Calibration Team Recommendations • Exploitation of available spectral campaign data • SeaWiFs, Chris/Proba • Programming of new spectral campaigns for radiometric calibration • Spot/Vegetation, ATM, MODIS, GLI, AATSR • Calibration Plan • Systematic comparison of MERIS and AATSR based on macro-pixels • Systematic acquisition of data over designated targets and distribution via FTP • Hold Regular Cal/Val workshops

8. Verification Team Conclusions • All the products are verified and ready for the validation loop. • Minor problems have been identified for some products (Land Surface Pressure, Water Vapour above clouds, Case2 products). • Additional interaction between verification and validation team is needed. • DPM/IODD/TDS including smile correction and pressure upgrade will be available by Christmas as recommended during the calibration workshop.

9. Verification Team Recommendations • Level 0 data corresponding to match-up need to be retrieved and processed with appropriate calibration files. • Atmospheric correction over glint area needs to be revised. • Atmospheric correction above complex water needs improvement (different treatment of absorbing aerosol). • PCDs are ok, but the logic may need some simplification. • Additional flag for land product to be added.

10. Cloud and Water Vapour Team Conclusions • CTP works quite well when comparing with radiosoundings. • WV in a quite good agreement with Radiosoundings, radiometer, gps and Modis. Apparently Meris under-estimates somewhat. • WV over clouds and land needs more investigation. • Cloud Optical Thickness ‘looks’ ok, but it is too early to have a conclusion. Need 3 months of data for statistical analysis. • First estimation results: • products range method accuracy COT 0-200 MODIS 40% • CTP 250-1000hPa MODIS-radiosonde <50 hPa • WV land 0.2-5.0 g/cm² MODIS-radiosonde-gps <15% • WV clouds 0-2 g/cm² radiosonde <20% • WV water 0.2-4.0 g/cm² radiosonde tbc

11. Cloud and Water Vapour Team Recommendation • New neural network for CTP using the precise wavelengths. Planned. • 3 aircraft campaigns were made in November. Results will be analysed. • Aircraft campaign are planned in May. • Check carefully the meteo products. Averaging effect. • (A)ATSR data IR can be used for CTP validation. • Use the reflectance from Meris to generate an albedo map for improving the atmosphere products . • For WV above ocean take into account the aerosol effect.

12. Interlude : Mean Water Vapour Content from 2002-11-28 to 2002-12-03

13. Land Team Conclusion • The first results are very promising. • Independent investigations of MGVI show good agreement in comparison with SeaWIFS. • Cross-Validation of variables using other sensors (also w/o products) show consistency.

14. Land Team Recommendation • New Look Up Table has been generated. Based on ESL first validation, the results are improved. If investigation confirms the first results, The new land LUT should be loaded in the processor. • Include additional test on cloud screening based on results of vegetation validation. • Need more data (including FR data). • Improve PCDs and include new science flags in the product. • Compare MERIS MGVI with the corresponding products from MOS, VEGETATION, MODIS.

15. Land Team Recommendation • Investigate red edge classification capability in MERIS FR data. • Establish more combinations with MODIS over e.g. VALERI and MODLAND sites. Intercomparison. over global sites. • Need to have a spatial and temporal composite product to support/validate global vegetation maps. • Need for exact knowledge of the MERIS geometric performance.

16. Land Team Recommendation • Pressure over Land: • Recommendation: shift the O2 band 1.25 nm toward the red. • Use of pressure into cloud flagging. • Aerosol: • Need to improve RO_DDV versus ARVI. Extend the concept of DDV to less dark vegetation.

17. Ocean Colour Team Conclusions • A significant effort was put in the intercalibration of optical instruments and in the definition of common measurements laboratories protocols. • The number of match-ups for marine products is too limited to conclude on the validation exercise. • The first preliminary results show that water leaving reflectance seem quite good. The validation measurements data set is small to draw any further conclusion. Some spectra with strange shapes have been reported by some PIs, but in many cases, the corresponding data are flagged as invalid by the processor. • PCDs are very severe. • Algorithm is unable to discriminate absorbing aerosols from absorbing waters when the turbid flag is not raised.

18. Ocean Colour Team Recommendations • Process missing match ups • Improve the mission_planning overpasses to include camera 5 • Detection of Clouds • Improve cloud detection using O2 results • Develop cirrus clouds, • Improve the atmospheric correction over absorbing waters • Revise PCDs and Flags (new flags = white cap etc.) • Smoother transition of Case2_s AC correction • Improve the Sun-glint correction

19. General Recommendations • Full Resolution Background Regional Missions • All land surfaces and coastal zones

20. Conclusion • MERIS instrument is in a very good shape. • The validation has started, however,more match up data are needed to conclude on the achieved accuracy. • More time is needed to look at the data and make a complete synthesis. • The first results look very promising. • The products can be distributed following smile update early next year. • IPF updates foreseen by May 2003 to include the validation findings.