Comparison of Various Reanalyses Gridded Data with Observed Data from Meteorological Stations over India

Abstract

Meteorological data such as precipitation and temperature are important for hydrological modelling. In areas where there is sparse observational data, an alternate means for obtaining information for different impact modelling and monitoring activities is provided by reanalysis products. Evaluating their behaviour is crucial to know their uncertainties. Therefore, we evaluated two reanalyses gridded data products, viz., Coordinated Regional Climate Downscaling Experiment (CORDEX) and National Centers for Environment Predictors and GCM (General Circulation Model) predictor variables (NCEP); two station based gridded data products, viz., Asian Precipitation - Highly-Resolved Observational Data Integration Towards Evaluation (APHRODITE) and India Meteorological Department (IMD) gridded data; one satellite based gridded data product i.e., Tropical Rainfall Measuring Mission (TRMM); and one merged data product, i.e., Global Precipitation Climatology Project (GPCP). These products were compared with IMD observed station data for 1971 to 2010 to evaluate their behaviour in terms of fitness by using statistical parameters such as NSE, CRM and R2. APHRODITE and TRMM gridded data showed overall good results for precipitation followed by IMD, GPCP, CORDEX and NCEP. APHRODITE also showed good agreement for mean temperature. CORDEX and NCEP gave a promising result for minimum and maximum temperatures with NCEP better than CORDEX.