The use of spatial statistics in the geosciences has become increasingly popular over the last few years. The framework of spatial statistics is a versatile sub-field of statistics that exploits the spatial correlation structure of measurements and process values at different locations. It allows for diverse applications such as gap-filling global satellite data, merging different data products, changing the spatial support of data using downscaling and predicting the extent of spatial phenomena such as pollution fields.

This talk will provide a methodological overview of spatial statistics with an emphasis on aspects relevant to large satellite datasets. The second part of the talk will consist of examples, including gap-filled global CO2 concentration maps based on satellite data, merging ground-based and satellite measurements for aerosol data, inverse modeling and the prediction of the spatial extent of hypoxia in lakes and along the ocean shore line.