Pengembangan Model Rasionalisasi Stasiun Hujan Menggunakan Metode Jaringan Syaraf Tiruan dan Kerapatan WMO (World Meteorological Organization) Pada DAS Ciliwung

Abstract

The accuracy of hydrological data is crucial for water resource planning. The Ciliwung watershed in Indonesia faces a challenge as the rainfall monitoring stations are too close to each other, resulting in the need for a review of the quality and quantity of data. One method to address this issue is the use of the Artificial Neural Network Method to rationalize the rain post network based on the needs of the watershed. The Ciliwung watershed with an area of 370,8 km² and a main river length of 124,1 km is influenced by 9 rain stations that are scattered, for the upstream and midstream there are 5 rain stations that will be used as case studies. The results of the assessment were analyzed using the WMO standard, for the upstream and midstream parts of the downstream including tropical mountainous areas with an area of influence of 100-250 km2 where 12 combinations of rain stations were found that met the standard. The modeling results using JST also demonstrate relatively small error values of less than 10%. In modeling with 5 years of data is 3.298% in the combination of loss of Sta. Gunung Mas Plantation. For 2 years of data is 5,868% in the combination of the loss of Sta. Gadog and Cibinong. The need for relocation at one of the stations, Sta. Gunung Mas Plantation, is highlighted based on the analysis of 5 years and 2 years of data. One of the factors because in the upper part of the Ciliwung watershed the distance between stations is too close, causing less ideal distribution of rain stations.