Download the paper:

Note: Full text may not be available for papers that have not yet been published.


Entekhabi, D., R. H. Reichle, R. D. Koster, and W. T. Crow:
"Performance Metrics for Soil Moisture Retrievals and Application Requirements"
Journal of Hydrometeorology, 11, 832-840, doi:10.1175/2010JHM1223.1, 2010.

Abstract:
Quadratic performance metrics such as root-mean-square error (RMSE) and time series correlation are often used to assess the accuracy of geophysical retrievals (satellite measurements) with respect to true fields. These metrics are related; nevertheless each has advantages and disadvantages. In this study we explore the relation between the RMSE and correlation metrics in the presence of biases in the mean as well as in the amplitude of fluctuations (standard deviation) between estimated and true fields. Such biases are common, for example, in satellite retrievals of soil moisture and impose constraints on achievable and meaningful RMSE targets. Finally we introduce an approach for converting a requirement in an applications product into a corresponding requirement for soil moisture accuracy. The approach can help with the formulation of soil moisture measurement requirements. It can also help determine the utility of a given retrieval product for applications.


Home

NASA-GSFC / GMAO / Rolf Reichle