Streamflow data assimilation for soil moisture analysis

摘要

Streamflow depends on the soil moisture of a river catchment and canbe measured with relatively high accuracy. The soil moisture in theroot zone influences the latent heat flux and, hence, the quantityand spatial distribution of atmospheric water vapour andprecipitation. As numerical weather forecast and climate modelsrequire a proper soil moisture initialization for their land surfacemodels, we enhanced an Ensemble Kalman Filter to assimilatestreamflow time series into the multi-layer land surface modelTERRA-ML of the regional weather forecast model COSMO. The impact ofstreamflow assimilation was studied by an observing systemsimulation experiment in the Enz River catchment (located at thedownwind side of the northern Black Forest in Germany). The resultsdemonstrate a clear improvement of the soil moisture field in thecatchment. We illustrate the potential of streamflow dataassimilation for weather forecasting and discuss its spatial andtemporal requirements for a corresponding, automated river gaugingnetwork.