Estimated Parameters of Rain Flow Distribution Using L-Moment Method in South Sulawesi, Indonesia

Abstract

This research is an applied research which aims to determine the type of distribution of each district/city in South Sulawesi Province. The data used are secondary data belonging to the Water Resources Management Office (PSDA) from 58 rainfall stations with a span of 31 years . The L-Moment method is a method that defines the probability density function of a distribution with its first 4 characteristics. The four characteristics are L-location 𝜆𝜆 1 , L-variability 𝜆𝜆 2 , L-skewness 𝜏𝜏 3 dan L-curtosis 𝜏𝜏 4 . Several types of distribution in this study are the distribution of Generalized Extreme Value (GEV), Generalized Logistics (GLo), Generalized Pareto (GPa), Lognormal III (LN3), Pareto type III (Pe3) and Gumbel. Determination of the best distribution based on 3 test indicators, namely the minimum value of Root Mean Square Error (RMSE) and Mean Absolute Error (MAE) and the maximum value of the Correlation Coefficient (CC). The results obtained are Soppeng district have the GPa distribution; The results of this research can be utilized by the PSDA Service, Public Works Service, civil engineering sector, and researchers in the fields of hydrology, climatology and geography to plan irrigation development, prevention of future disasters or the need for science.

Stochastic Model of the Annual Maximum Rainfall Series Using Probability Distributions

“…Their findings revealed that among these distributions, the GEVD demonstrated superior performance. Another study conducted by Sanusi et al [11] discovered that the annual maximum rainfall data in various South Sulawesi Province districts, such as Pinrang and Takalar, was best represented by the GEVD. Susanti et al [12] concluded that the GEVD performed better than the GPA distribution for maximum rainfall data in Pekanbaru City, Riau.…”