|Validation of MERIS water products for Belgian coastal waters: 2002-2003|
Park, Y.; De Cauwer, V.; Nechad, B.; Ruddick, K. (2004). Validation of MERIS water products for Belgian coastal waters: 2002-2003, in: (2004). Proceedings of the MERIS and AATSR calibration and geographical validation workshop held in Frascati, 20-24 October 2003. ESA Special Publications, : pp. 1-16
In: (2004). Proceedings of the MERIS and AATSR calibration and geographical validation workshop held in Frascati, 20-24 October 2003. ESA Special Publications. European Space Agency: Frascati, more
In: ESA Special Publications. ESA Publications Office, more
Coastal waters; Remote sensing; ANE, Belgium, Belgian Coast [gazetteer]; Marine
|Project|| Top | Authors |
- BELCOLOUR - Optical remote sensing of coastal waters, more
This paper describes the validation of MERIS products for Belgian coastal waters for the period 2002-2003. The MERIS Level 2 water products are compared with sea-borne measurements if match-up conditions are satisfied. A qualitative assessment of MERIS imagery was made to provide recommendations to improve image quality as in our previous report. This shows that undetected cloud contaminated pixels can severely deteriorate image quality. Patches of non-positive water-leaving reflectance at blue bands are also found for both case 1 and case 2S waters, and water constituent retrieval may be unrealistic for those patches. Furthermore, a quantitative comparison for the match-up data is made focusing on a comparison of water-leaving reflectance. Compared with last year's MERIS processing ('200210' version), the latest MERIS processing ('200308' version) gives better results for the spectral shape of water leaving reflectance in the blue and green bands, but differences in the absolute magnitude of reflectance are found. No apparent bias of MERIS reflectance with respect to the seaborne reflectance is found from the scatter plot of all the spectral reflectance data of 2002 and 2003. However, the relative difference is larger for blue bands and NlR bands. Differences in reflectance at visible bands could not be related to differences in either the imposed NlR band ratio of water reflectance or aerosol optical thickness with respect to our match-up data. This should be further investigated to improve the atmospheric correction procedure.