Climate Change Impact Assessment Using K-Nearest Neighbor Weather Generator: Case Study of the Nakdong River Basin in Korea

Abstract

This study presents a methodology for assessing impacts of climate change in regional-scale hydrology using the K-nearest neighbor weather generator (WG) model in combination with the outputs of global circulation models (GCMs) and hydrologic rainfall-runoff model. The Nakdong River Basin in Korea is used as a case study. The study applies a systematic approach to select the variables for the WG model from five meteorological variables available in the basin. In addition, the GCMs’ projections based on a B1 emission scenario are incorporated into the proposed WG model to reflect the impacts of climate change, consequently generating the meteorological series for 60 years. Meteorological data for historic and two climate scenarios (dry and wet) are generated. The generated time series of meteorological variables are used with the streamflow synthesis and reservoir regulation rainfall-runoff model to calculate the streamflow at 25 sub-basins. The results demonstrate that the WG models combined with the GCMs’ outputs are able to provide future weather conditions for the assessment of climate change impacts. One of the major findings of the study is the potential severity of drought impacts that may increase the historical drought duration in the basin up to three times.