WRF-PDM: Prognostic approach for discharge prediction in ungauged catchment

P. K. Srivastava, T. Islam, George Petropoulos, M. Gupta

Research output: Contribution to journalArticlepeer-review

37 Downloads (Pure)

Abstract

In this study, the mesoscale model WRF (Weather Research & Forecasting model) is used for dynamical downscaling of European Centre for Medium Range Weather Forecasts (ECMWF) ERA interim reanalysis global datasets for obtaining hydro-meteorological variables. The WRF estimated precipitation and Evapo-transpiration are then used as input data for the Probability Distributed Model (PDM) for discharge prediction. For performance evaluation of the integrated framework, objective function like Nash Sutcliffe Efficiency (NSE) is used. Analysis of NSE indicates values of 0.85 and 0.82 during the calibration and validation respectively for the combination observed rainfall and station based reference Evapotranspiration (ETo). On the other hand, a marginally lower performance is reported by the combination Observed Rainfall and WRF based ETo (NSEcal=0.82; NSEval=0.80), while a very poor performance is reported by the combination Rainfall and ETo when both derived from WRF (NSEcal=0.58; NSEval=0.06). The overall analysis suggests that the WRF-PDM can be used for discharge prediction in the absence of ground based measurements. This study provides valuable information to the hydro-meteorologist on downscaled weather variables from global datasets and its applicability to rainfall-runoff modeling for river discharge prediction
Original languageEnglish
Pages (from-to)129-132
Number of pages4
JournalEuropean Water
Volume57
Publication statusPublished - 19 Oct 2017

Keywords

  • WRF-PDM
  • GLUE
  • rainfall-runoff
  • discharge
  • sensitivity analysis and uncertainty estimation

Fingerprint

Dive into the research topics of 'WRF-PDM: Prognostic approach for discharge prediction in ungauged catchment'. Together they form a unique fingerprint.

Cite this