Probable maximum precipitation for a coastal site of Tarapur, Maharashtra, India, using intensity–duration–frequency curves

Abstract

Probable maximum precipitation (PMP) for Tarapur, a coastal industrial area, has been assessed by developing intensity–duration–frequency (IDF) curves using Atomic Energy Regulatory Board recommended probabilistic methods. The analysis identified Fréchet distribution as the best fit according to goodness of fit tests; however, its heavy-tailed nature led to implausible PMP estimates when compared with similar studies. Consequently, the Gumbel distribution, with parameters estimated using the least squares method, was selected as a more appropriate model for extreme rainfall events. The study also identified the most optimal plotting position formulations (PPFs) for the considered meterorological parameter. The Hazen PPF was identified to be better, for the considered rainfall data and was used to determine the PMP estimates for different storm durations. The distribution parameters were applied to yield the PMP estimates for 100, 1000, and 10000-year return period across all durations. The PMP presented as IDF charts prepared for this site, will be useful for modeling runoff characteristics and determining maximum water levels.

Keywords

• extreme value analysis
• intensity–duration–frequency curves
• least square
• order statistics approach
• probable maximum precipitation