Yueying Wang / Abstract : Satellite images have become a major source of data for atmospheric, ocean and land studies. Those applications and data processing algorithms typically require complete remote sensing data, while satellite images often present missing values due to cloud cover and sensor specific problems, for instance the scan-line corrector (SLC) failure of Landsat 7 Enhanced Thematic Mapper Plus (ETM+) sensor in 2003. A variety of gap-filling techniques have been proposed to tackle the problem. In this talk, I will first review several geostatistical and spatiotemporal based gap-filling methods, and then propose an alternative functional model for data imputation, which is less computationally intensive. In the approach, FPCA and tensor-product of bivariate spline and B-spline are used for fitting spatiotemporal data. Finally, some experimental analyses results will be presented to illustrate and evaluate our algorithm.