Using Data Interpolating Empirical Orthogonal Functions (DINEOF) Algorithm for filling missing data of AVHHR satellite images

Abstract

Using Data INterpolating Empirical Orthogonal Functions (DINEOF) Algorithm for filling the missing data of AVHHR satellite images. Andreas Nikolaidis(*,a,b), Stavros Stylianou(b,Georgios Georgiou(b), Diofantos Hajimitsis(a) and Evangelos Akylas(a) a. Department of Civil Engineering and Geomatics, Cyprus University of Technology, 30 Archbishop Kyprianou Str., 3036, Limassol, Cyprus b. University of Cyprus, PO BOX 20537, 1678, Nicosia, Cyprus, Oceanography Centre KEY WORDS: Remote Sensing, Cyprus, Mediterranean, DINEOF, ArcGIS, data reconstruction. During the last decade, Rixen (2005) and Alvera-Azkarate (2010) presented the DINEOF (Data Interpolating Empirical Orthogonal Functions) method, a EOF-based technique to reconstruct missing data in satellite images. The application of DINEOF method, proved to provide relative success in various experimental trials (Wang and Liu, 2013; Nikolaidis et al., 2013;2014), and tends to be an effective and computationally affordable solution, on the problem of data reconstruction, for missing data from geophysical fields, such as chlorophyll-a, sea surface temperatures or salinities and geophysical fields derived from satellite data. Implementation of this method in a GIS system will provide with a more complete, integrated approach, permitting the expansion of the applicability over various aspects. This may be especially useful in studies where various data of different kind, have to be examined. For this purpose, in this study we have implemented and present a new GIS toolbox that aims to automate the usage of the algorithm, incorporating the DINEOF codes provided by GHER (GeoHydrodynamics and Environment Research Group of University of Liege) into the ArcGIS®. ArcGIS® is a well known standard on Geographical Information Systems, used over the years for various remote sensing procedures, in sea and land environment alike. A case-study of filling the chlorophyll-a missing data in the Mediterranean Sea area, for a 18-day period is analyzed, as an example for the effectiveness and simplicity of the usage of this toolbox. The specific study focuses to chlorophyll-a MODIS satellite data collected by CNR-ISAC (Italian National Research Council, Institute of Atmospheric Sciences and Climate), from the respective products of MyOcean2® organization, that provides free online access to Level 3, with 1 km resolution. In particular, all the daily products with an initial level of only 27% data coverage were successfully reconstructed.