New algorithms for estimating chlorophyll-a in the Spanish waters of the Western Mediterranean Sea from multiplatform imagery

Abstract

ABSTRACTThe utility of global and regional algorithms for retrieving surface chlorophyll-a values from satellite images of MODIS (Moderate-Resolution Imaging Spectroradiometer) onboard Aqua and Terra Satellites, SeaWiFS (Sea-Viewing Wide Field-of-View Sensor), MERIS (MEdium Resolution Imaging Spectrometer) and VIIRS (Visible Infrared Imaging Radiometer Suite) is tested in a wide region of the Western Mediterranean Sea, the Spanish Mediterranean. The comparison between chlorophyll-a concentration obtained from samples collected and satellite concentrations calculated with global and regional algorithms for this area demonstrates that satellites overestimate the surface chlorophyll-a. In this work, a set of new algorithms, namely SMED (Spanish MEDiterranean algorithms), are proposed to improve the estimations of surface chlorophyll-a in our study area. The SMED set of algorithms employs the linear function to fit the standard Maximum Band Ratios (MBR) to the in situ surface chlorophyll-a concentration (in logarithmic scale). The implementation of the SMED algorithms is simple, and the accuracy is as good as more complex algorithms like OC5 (5 band algorithms tuned for European Atlantic and tested successfully in the Mediterranean Sea). The improvement of SMED algorithms is important respect to MedOC (built based in data of the Mediterranean Sea) for all sensors, ranging 17%-37% in terms of the coefficient of determination (R2). In the case of OC algorithms (standard global algorithms currently operative for most ocean colour sensors), SMED also improves the estimates in 11%-22%, except in 4% for MERIS. Additionally, when comparing satellite estimates with in situ chlorophyll-a in log-log scale, SMED achieved the best correspondence, even with respect to OC5. During the calibration procedure MERIS remote sensing reflectances (RRS) showed a higher sensitivity to the presence of mineral aerosols, and therefore it is strongly recommended to include the value of Aerosol Optical Thickness (AOT) in the algorithms for chlorophyll in the region under study. A regional algorithm for VIIRS ocean colour sensor is also proposed based on a matchup database built from spatially and temporary distributed SMED chlorophyll-a values of MODIS-Aqua and VIIRS RRS. Our analysis suggests that SeaWiFS, MODIS-Aqua, and VIIRS based SMED algorithms can support generating a multiplatform time series in the region.