A bootstrap regional model for assessing the long-term impacts of climate change on river discharge. (2019)

• Record Type: Journal Article
• Authors: Li, Chi-Yu; Lin, Shiu-Shin; Lin, Yi-Fong; Kan, Ping-Shiuan
• Is Part Of: International journal of hydrology science and technology. Volume 9:Number 1(2019)
• Page Start: 84
• Page End: 108
• Publication Date: 2019
• Subjects: climate change -- genetic algorithm -- neural network -- regional river model -- bootstrap; Hydrology -- Periodicals; Water resources development -- Periodicals; Hydrology; 553.705
• Languages: English
• ISSNs: 2042-7808

Abstract

Water resources in Taiwan come predominantly from rivers. Hence, it is important to understand the impact of future climate scenarios for policymaking. To investigate the impact of accelerating climate change on river flow in Taiwan, a regional flow impact model (RFIM) was developed. The RFIM is based on the radial basis function neural network. It adapts the genetic algorithm for parameter optimisation and the bootstrap method for quantifying uncertainties in the model and its results. The study area is the Taiwan Island, divided into four water resource management regions: North, Middle, South and East. After the RFIMs were developed for different regions, various future weather scenarios predicted from global circulation models were applied. The results suggest that the average discharge increases at a higher rate in the Middle and the East and the uncertainty of future discharge is higher in the Middle and the South of Taiwan Island.