Uncertainty Impacts of Climate Change and Downscaling Methods on Future Runoff Projections in the Biliu River Basin

Zhu Zhu,Fu Fu,Liu Liu,Yue Yue,Zhang Zhang,Wu Wu,Qi Qi

doi:10.3390/w11102130

Abstract

This paper assesses the uncertainties in the projected future runoff resulting from climate change and downscaling methods in the Biliu River basin (Liaoning province, Northeast China). One widely used hydrological model SWAT, 11 Global Climate Models (GCMs), two statistical downscaling methods, four dynamical downscaling datasets, and two Representative Concentration Pathways (RCP4.5 and RCP8.5) are applied to construct 22 scenarios to project runoff. Hydrology variables in historical and future periods are compared to investigate their variations, and the uncertainties associated with climate change and downscaling methods are also analyzed. The results show that future temperatures will increase under all scenarios and will increase more under RCP8.5 than RCP4.5, while future precipitation will increase under 16 scenarios. Future runoff tends to decrease under 13 out of the 22 scenarios. We also found that the mean runoff changes ranging from −38.38% to 33.98%. Future monthly runoff increases in May, June, September, and October and decreases in all the other months. Different downscaling methods have little impact on the lower envelope of runoff, and they mainly impact the upper envelope of the runoff. The impact of climate change can be regarded as the main source of the runoff uncertainty during the flood period (from May to September), while the impact of downscaling methods can be regarded as the main source during the non-flood season (from October to April). This study separated the uncertainty impact of different factors, and the results could provide very important information for water resource management.

Highlights

Uncertainty analysis is important when simulating runoff because it is difficult for hydrological models to reflect the hydrological process perfectly
The runoff could decrease if future precipitation increases less than 4.5% or decreases
It was found that the impact of climate change can be regarded as the main source of the runoff uncertainty during the flood period, while the impact of the downscaling methods can be regarded as the main source during the non-flood season

Summary

Introduction

Uncertainty analysis is important when simulating runoff because it is difficult for hydrological models to reflect the hydrological process perfectly. Outputs of the Global Climate Models (GCMs) are the most effective data to represent future climate conditions and are widely used to study future hydrological responses to the impact of climate change [1,2,3]. Hydrological models are driven by future climate data from GCMs, and the projected future runoff is compared with historically measured runoff to analyze the runoff’s response to climate change. The GCMs output with a low-resolution and need to be downscaled to a regional scale, which is suitable for hydrological modeling [4,5,6,7]. The uncertainties of climate change (including the choice of GCMs and emission scenarios) and the uncertainties of downscaling methods are regarded as the main uncertainties of the hydrological model inputs in studies on climate change’s impact on runoff [10]

Results

Discussion

Conclusion