Meteorological Sub-Divisional Scale Comparison Between Two Indian Rain Gauge-Based Rainfall Datasets for the Southwest Monsoon Season

Abstract

A monthly rainfall dataset for India at country, regional and meteorological sub-divisional scales was developed by the Indian Institute of Tropical Meteorology (IITM) based on a fixed network of 306 rain gauges. This dataset has been constructed when long period data was not available at many locations and there was not much computing power available. This data has been used worldwide for rainfall analysis over India. In this study, this rainfall dataset has been compared with a larger network of rain gauges maintained by the India Meteorological Department (IMD) for the southwest monsoon period of 1901–2010 at meteorological sub-divisional scale. Two different rain gauge networks can give rise to divergent estimates of rainfall, in general from differences in network density or location of individual rain gauges in each network, assuming measurement errors have small effect. Although mean monthly and seasonal monsoon rainfall and their interannual variability in both IITM and IMD datasets are similar, IITM dataset shows larger difference from IMD data for several meteorological sub-divisions. The long-term trends and frequency of occurrence of deficient and excess monsoon rainfall also show considerable differences between these two rainfall datasets. Data from a sparse network is not representative at meteorological sub-divisions associated with rather larger spatial variations in the southwest monsoon rainfall. For instance, IITM dataset has 11 rain gauges compared to 147 IMD rain gauges over a meteorological sub-division—South Interior Karnataka, and mean absolute difference in monthly monsoon rainfall estimates becomes about 25% when compared for rather shorter period using station data. It is also demonstrated that inclusion of additional rain gauges substantially improves the quality of IITM monthly rainfall estimates over this specific meteorological sub-division.