Flood prediction: analyzing land use scenarios and strategies in Sumber Brantas and Kali Konto watersheds in East Java, Indonesia

Previous studies have emphasized the significant influence of land use and land cover (LULC) on flood hazard severity. However, the analysis has been restricted to a single dataset and scenario. This study is carried out to analyze the land use options for flood prediction by examining three distinc...

Full description

Bibliographic Details
Main Authors: Putra, Aditya Nugraha, Alfaani, Salsabila Fitri, Saputra, Danny Dwi, Andhika, Yosi, Wisnubroto, Erwin Ismu, Admajaya, Fandy Tri, Maritimo, Febrian, Paimin, Saskia Karyna, Kusumawati, Irma Ardi, Prasetya, Novandi Rizky, Sugiarto, Michelle Talisia, Nita, Istika, Sudarto, Sudarto, Sujarwo, Sujarwo, Rayes, Mochtar Lutfi, Suprayogo, Didik, Ismail, Mohd. Hasmadi, van Noordwijk, Meine
Format: Article
Language:English
Published: Springer Science and Business Media B.V. 2025
Online Access:http://psasir.upm.edu.my/id/eprint/120778/
http://psasir.upm.edu.my/id/eprint/120778/1/120778.pdf
Description
Summary:Previous studies have emphasized the significant influence of land use and land cover (LULC) on flood hazard severity. However, the analysis has been restricted to a single dataset and scenario. This study is carried out to analyze the land use options for flood prediction by examining three distinct scenarios namely business as usual (BAU), regional spatial planning (RSP), and land capability (LC). The BAU (2025) scenario was forecasted by using a multitemporal LULC baseline (2017, 2019, 2021 and 2022) and modelled with the ANN Cellular Automata-Markov Chain. The RSP and LC scenarios were developed based on the official regional spatial planning of Malang Regency and Batu City, while LC was developed through the land capability classification limiting factor method. These scenarios were applied to predict flood levels using the InVEST model, incorporating factors such as rainfall depth, hydrologic soil group, curve number, and a biophysical table for infiltration analysis, by using SCS Curve Number analysis in InVEST. The result shows a decline in forest cover (from 31 to 23%) and agroforestry (from 3 to 2%) to correspond with a 16% increase in flood hazard levels. This correlation was identified using pearson model and validated (Kappa accuracy) through ground-check surveys, achieving an overall classification accuracy of 75%. If there are no interventions, high and very high flood hazard levels could escalate to 12% and 4% in 2025. In contrast, the RSP and LC scenarios show promise in reducing flood hazards by 16% and 10%, respectively. Remarkably, the LC scenario has shown to be the most effective strategy for the land use approach, showcasing a potential to prevent flood hazards because it maintains the existence of forests according to their land capabilities.