A global analysis of spatial correlation lengths of water storage anomalies

This study has been run in the context of the European Union research project G3P (Global Gravity-based Groundwater Product) on developing Groundwater storage (GW) as a new product for the EU Copernicus Services. GW variations can be derived on a global scale by subtracting from total water storage (TWS) variations based on the GRACE/GRACE-FO satellite missions variations in other water storage compartments such as soil moisture, snow, surface water bodies, and glaciers. Due to the nature of data acquisition by GRACE and GRACE-FO, the data need filtering in order to reduce North-South-oriented striping errors. However, this also leads to a spatially smoothed TWS signal. For a consistent subtraction of all individual storage compartments from GRACE-based TWS, the individual data sets for all other hydrological compartments need to be filtered in a similar way as GRACE-based TWS.

In order to test different filter methods, we used compartmental water storage data of the global hydrological model WGHM. The decorrelation filter known as DDK filter that is routinely used for GRACE and GRACE-FO data introduced striping artifacts in the smoothed model data. Thus, we can conclude that the DDK filter is not suitable for filtering water storage data sets that do not exhibit GRACE-like correlated error patterns. Alternatively, an isotropic Gaussian filter might be used. The best filter width of the Gaussian filter is determined by minimizing the differences between the empirical spatial correlation functions of each water storage and the spatial correlation function of GRACE-based TWS. We also analyzed time variations of correlation lengths such as seasonal effects. Finally, the selected filter widths are applied to each compartmental storage data set to remove them from TWS and to obtain the GW variations. 

 

Acknowledgement :

This study received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement nº 870353.