Reconstruction of Missing Satellite Total Suspended Matter Data over the Southern North Sea and English Channel using Empirical Orthogonal Function Decomposition of Satellite Imagery and Hydrodynamical Modelling

摘要

Optical remote sensing data archives generally have many gaps caused by clouds or other retrieval problems. However, for the light forcing of ecosystem models continuous fields are required. For parameters exhibiting strong spatial and temporal correlations for regions of similar dynamics or from day to day, the missing data can be estimated by use of statistical techniques. In this context, the Data Interpolation with Empirical Orthogonal Functions (DINEOF) method is used for reconstruction of complete space-time information for surface total suspended matter (TSM) and chlorophyll a from a 5-year archive of MODIS and MERIS products over the Southern North Sea and English Channel. The DINEOF univariate methodology has been previously demonstrated for Mediterranean sea surface temperature data (Alvera-Azcarate et al., 2005, Beckers et al., 2006). Alvera-Azcarate et al (2007) showed that SST reconstructions could be improved by using a multivariate approach in which SST, chlorophyll and wind fields were taken into account together for the analyses.Here, TSM images will be used in combination with information from the COHERENS hydrodynamical model to provide a complete and continuous estimate of surface TSM for the Southern North Sea throughout the period 2003-2005. In addition to the remotely sensed TSM, the DINEOF multivariate analysis will consider wind fields, depth integrated currents, surface elevations and bottom stresses. Reconstucted images are compared with the original incomplete images. Validation of the method is achieved by estimation of information removed from the training data by exclusion of entire images and by addition of artificial clouds. The data reconstruction technique has further applications in the processing and quality control of optical remote sensing data. Perspectives will be outlined for improving the quality control of retrieved parameters and for the improvement of retrievals by adding statistical information to the conventional spectral processing.References: Alvera-Azcarate, A., Barth, A., Rixen, M., and Beckers, J.-M.: Reconstruction of incomplete oceanographic data sets using Empirical Orthogonal Functions. Application to the Adriatic Sea, Ocean Modelling, 9, 325–346, 2005.Alvera-Azcarate, A., Barth, A., Beckers, J. M., and Weisberg, R. H.: Multivariate Reconstruction of Missing Data in Sea Surface Temperature, Chlorophyll and Wind Satellite Fields, Journal of Geophysical Research, 112, C03008, doi:10.1029/2006JC003660, 2007.Beckers J.-M., A. Barth & A. Alvera-Azcarate, DINEOF reconstruction of clouded images including error maps. Application to the Sea-Surface Temperature around Corsican Island, Ocean Sciences, 2: 183–199, 2006.