Download PDFOpen PDF in browserProbability Distribution and Frequency Analysis of Consecutive Days Maximum Rainfall at Sambra (Belagavi), Karnataka, IndiaEasyChair Preprint 530412 pages•Date: April 8, 2021AbstractIn this case study, Daily Rainfall Data (1984-2019) of SambraRaingauge station in North Karnataka is used. An attempt was made to fit various probability distribution functions to the datasets of 1 day and 2 to 5 consecutive days annual maximum rainfall. The goodness of fit of probability distribution functions were tested by comparing the Chi-square (χ2) values. No single probability distribution was adequate to describe the entire datasets. Various trendlines were also fitted to the rainfall datasets mentioned above; the best fit was decided based on the value of coefficient of determination R2, no single trendline equation was able to describe the entire datasets. The magnitudes of 1 day as well as 2 to 5 consecutive days annual maximum rainfall corresponding to 2 to 100 years return period were estimated best fit distribution function, it was found that even though Normal distribution function had low Chi-square value comparatively, it cannot be used overall for estimation of rainfall values of different return periods for all the datasets. Rainfall was also estimated by best fit trendline equation i.e.polynomial 3rd order, for all the datasets corresponding to 2 to 100 years return period. It was observed the rainfall values predicted for 100 years return period for 1 to 5 consecutive days maximum rainfall were extremely high and unrealistic with respect to climate conditions of Sambra region. Chi-square test (χ2) was conducted between observed rainfall and predicted rainfall by different trendline equations to ascertain the bestfit as determined by R2, it was not able to establish the same results as determined by coefficient of determination. Keyphrases: Chi-square test, Rainfall, Trendline equation, frequency analysis, probability distribution
|