The return period is a statistical parameter used in “frequency” analysis and is a measure of the probable time interval between the occurrence of a given event and that of at least equal or greater event. Return period analysis is used mainly for any construction of hydraulic small structure e.g., culvert, bridges, highways, railways etc. based on the annual maximum rainfall. The return period analysis is firsthand information for construction of any hydraulic structures in a particular catchment area. The present study is carried out in the Upper Godavari River basin area which is having three river sub-basins namely Sub-Basin-1 (SB1), Sub-Basin-2 (SB2) and Sub-Basin-3 (SB3) based on the daily annual maximum rainfall data of more than 115 years of nearby the three stations. These three stations at Yeola, Sinner and Igatpuri are located nearby SB1, SB2 and SB3 respectively. The study is carried out for computation Return Period Values (RPV) for T-year t-hour return period rainfall values (where T=100-year, 200-year and 500-year, t=1-day and 2-day). The RPV is computed by using Gumbel extreme value distribution.
The available annual maximum station rainfall data (1901 to 2020) is arranged in a running window of 50 years with increment of 10 years i.e., 1901-1950, 1911-1960, 1921-1970, 1931-1980, 1941-1990, 1951-2000, 1961-2010 and 1971-2020 of Yeola, Sinner, and Igatpuri for SB1, SB2 and SB3 respectively to study the trend in extreme rainfall events. These results are also compared with the RPVs calculated by using the total available rainfall data from 1901-2020, a span of 118 years.
The Return Period Analysis reveals that the annual extreme rainfall is initially increasing and last two spans showed the decreasing trend over catchments SB1 and SB2 (inner plateau region) whereas annual extreme rainfall is increasing over the catchment area SB3 (Hilly region) for all the all the 50 years spans.
In other words it could say that 100, 200 & 500yrs RPVs for 1-day, 2-day over SB1, SB2, & SB3 calculated by different running window of 50 years with increment of 10 years are within the limit of ± 15% same that of the total available rainfall data except one or two windows.
Return Period Values, Gumble distribution, Annual Extreme Rainfall
