Data assimilation in integrated hydrological modeling using ensemble Kalman filtering: evaluating the effect of ensemble size and localization on filter performance

• DOI: 10.5194/hess-19-2999-2015
• Publication Date: 2015-07-02T09:07:02Z
• Authors: J. Rasmussen; H. Madsen; K. H. Jensen; J. C. Refsgaard
• Keywords: Hydraulic head; Ensemble Learning

Abstract. Groundwater head and stream discharge is assimilated using the ensemble transform Kalman filter in an integrated hydrological model with aim of studying relationship between performance size. In attempt to reduce required number members, adaptive localization method used. The compared more common distance-based localization. terms hydraulic error members investigated for varying numbers spatial distributions groundwater observations or without assimilation parameter estimation. study shows that (1) are needed when fewer assimilated, (2) assimilating estimating parameters requires a much larger size than just observations. However, can be greatly reduced use localization, which by far outperforms conducted synthetic data only.