Estimation of spatial correlation structure of gauged rainfall in network under semi-arid situation of Krishna basin at watershed scale

Abstract

Kriging serves as a geospatial interpolation method for delineating optimal radial zones around existing rainfall stations within the flat land watershed of Machkula (4D5B3) situated in the Krishna basin. This analysis primarily focuses on smaller (15 and 30 minutes) and medium (45 minutes to 1 hour 45 minutes) rainfall events, examining their spatial distribution. For longer duration events (>2 hours), the analysis emphasizes depth across 22 gauging stations. The study utilizes data from 22 gauged rainfall network stations spanning the period of 2014-2020 to establish relationships between recurring events of varying duration (15 minutes to 2 hours). These relationships are evaluated in terms of the weighted influence of both the number and depth of events, considering the spatial morphology and constraints of the watershed. The spatial distribution of rainfall within the Machkula watershed is found to exhibit significant uniformity in both the number and depth of events within a radial zone extending up to 4 kilometers around each station. This uniformity is supported by strong goodness-of-fit statistics (R2) exceeding 0.8.