Prediction of the Hydrological Drought Index (HDI) based on Rain Simulation Data using an Artificial Neural Network (ANN) in the Randugunting Watershed, Rembang Regency

Abstract

Abstract Randugunting Watershed, with the main river Randugunting River, originates from the North Limestone Mountains of Rembang Regency. Because of its cyclical nature, the river flow discharge is continually subject to change. The discharge rises during rainy weather and falls during dry weather. Due to this issue, Rembang will likely experience drought throughout the dry season. The objectives of this study are to: (1) determine the drought index and its sharpness based on the Hydrological Drought Index Method (HDI) and (2) make drought predictions using the HDI based on discharge for the years 2023–2026 using simulated rain data with Artificial Neural Networks (ANN). Calculation of regional rainfall using the Thiessen Polygon Method. To find simulated rain and drought prediction based on the ANN with Backpropagation algorithm using Mathlab software. From the analysis indicate that the number of months that would experience wet conditions are decreasing from 7 months in 2023 and 2024 to 6 months in 2025 and only 3 months in 2026. In the anticipated drought from 2023 to 2026, there will be 14 dry months, 8 very dry months, and 3 extremely dry months. The examination of the hydrological drought’s duration revealed that the typical drought period fell between May and November.