OPTIMIZATION OF M PARAMETER ON THE MONONOBE FORMULA WITH ERROR REDUCTION AND REGRESSION METHOD BASED ON SHORT DURATION RAINFALL DATA

• Main Author: Prahardi, Tulus
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/61975
• Institution: Institut Teknologi Bandung

In mining activities, especially in Indonesia, the amount of rainfall that occurs will be an obstacle. To overcome these obstacles, it is necessary to design a drainage system. One of the steps in designing a drainage system is the determination of the intensity of the planned rain. Due to limitations on measurement, available data is often limited to daily rainfall. Mononobe’s formula is a formula that is generally used to calculate rainfall intensity from daily rainfall. In this formula, there is m parameter which is a coefficient that is influenced by climate conditions depending on the research location. The purpose of this study is to determine the optimal value of the m parameter, optimal here means that the error generated by the Mononobe’s formula will be minimal at the calculated m value. The method used in this research is by reducing the root mean square error (RMSE) and mean absolute percentage error (MAPE). Regression analysis was performed using quantile regression to strengthen the results of the previous analysis. The data used in this study is shortduration rain data which is reviewed daily at the Air Laya Mine location owned by PT Bukit Asam Tbk in year 1995 – 2001. The data picked for research are 30 and 100 most extreme data. From average rainfall intensity for each duration and daily rainfall, some value picked corresponding to their specified return. In 100 most extreme data, matched distributions and parameter estimates gave inconsistent results, so no further analysis was carried out on this data. Inconsistent parameter estimates found in high return periods, so the analysis is limited to small return periods of 2, 5, and 7 years. Error-values with MAPE do not treat errors in shortduration rains equally, so the estimation results with RMSE are more preferred for analysis. The optimal value of m is found in the range of 0.62 – 0.68 which obtained from the error reduction method with RMSE and is supported by the regression results in small return period.