Automatic Prediction of High-Resolution Daily Rainfall Fields for Multiple Extents: The Potential of Operational Radar

• DOI: 10.1175/2007jhm792.1
• Publication Date: 2008-01-03T15:16:30Z
• Authors: J. M. Schuurmans; M. F. P. Bierkens; E. J. Pebesma; R. Uijlenhoet
• Keywords: Variogram; Weather radar

Abstract This study investigates the added value of operational radar with respect to rain gauges in obtaining high-resolution daily rainfall fields as required distributed hydrological modeling. To this end data from Netherlands national gauge network (330 nationwide) is combined an experimental (30 within 225 km2). Based on 74 selected events (March–October 2004) spatial variability investigated at three extents: small (225 km2), medium (10 000 and large (82 875 From analysis it shown that semivariograms show no clear dependence season. Predictions point are performed for all extents using different geostatistical methods: (i) ordinary kriging (OK; only), (ii) external drift (KED), (iii) collocated cokriging (OCCK), latter two both range-corrected composites—a standard product Royal Meteorological Institute (KNMI). The focus here automatic prediction. For extent, alone perform better than radar, while larger lower densities, performs overall (OK). Methods (KED OCCK) prove be more accurate either (OK) or particular, extents. positively related correlation between data. Using a pooled semivariogram almost good event-based semivariograms, which convenient if prediction automated. An interesting result terms estimating error (kriging variance) especially where number points estimate rather unstable.