Sensitivities of Hydrological Processes to Climate Changes in a Central Asian Glacierized Basin

Zhihua He

doi:10.3389/frwa.2021.683146

Abstract

This study used the WASA (Water Availability in Semi-Arid Environments) hydrological model to simulate runoff generation processes and glacier evolution in the Ala-Archa basin in Central Asia. Model parameters were calibrated by observations of streamflow, satellite snow cover area (SCA) and annual glacier mass balance (GMB). Temperature and precipitation change scenarios were set up by perturbations of the reference measurements in a 20-year period of 1997 to 2016. Seven temperature warming scenarios with an increment of +1°C and six precipitation change scenarios ranging from 70 to 130% of the reference precipitation were used to investigate the sensitivities of hydrological processes to climate changes in the study basin. Results indicate that: (1) Annual runoff increased with rising temperature (T) and precipitation (P) at rates of 76 mm/+1°C and 62 mm/+10%P, respectively. Glacier area was more sensitive to T changes than to P changes. The total glacier area in the basin decreased with T warming at a rate of −0.47 km2/+1°C, whilst increasing with rising P at a rate of 0.16 km2/+10%P. (2) The basin runoff switched from rainfall and groundwater-dominated to ice melt-dominated with warming T, while the dominance of rainfall and groundwater were strongly enhanced by rising P. Proportion of rainfall in the total water input for runoff generation decreased with T warming at a rate of −0.5%/+1°C, while increasing with P increases at a rate of 1.2%/+10% P. Ice melt proportion changed with T and P increases at rates of 4.2%/+1°C and −1.8%/+10%P, respectively. Groundwater contribution to total runoff decreased by −2.8% per T warming of 1°C, but increased by 1.5% per P increase of 10%. (3) The maximum P changes (±30%) could only compensate the effects of T warming of 0.5 to 2.5°C. Increase of annual runoff forced by T warming lower than 2.2°C could be compensated by decrease caused by the maximum P decrease of −30%. Decrease of glacier area caused by 1°C warming cannot be compensated by the maximum P increase of +30%. The combined input of 20% increase of P and T warming of 6°C resulted in 90% increase of annual runoff, and 8% reduction of glacier area. The results inform understandings of the hydrological responses to potential climate changes in glacierized basins in Central Asia.

Highlights

Glaciers in Central Asia high mountains store large volumes of freshwater and provide important meltwater sources for downstream water supply, especially in dry years (Chen et al, 2019)
The model parameters were automatically calibrated by the εNSGAII algorithm using three different observations including streamflow measured at the basin outlet, satellite daily snow cover area (SCA) in the melting period from March to August, and observed annual glacier mass balance (GMB) of the Golubin glacier
The simulated runoff agreed well with observation in the modeling period (Figure 2A), indicated by the NSE and lnNSE values of 0.73 and 0.85, despite that summer peak flows in years such as 2010–2013 and 2015–2016 were underestimated

Summary

INTRODUCTION

Glaciers in Central Asia high mountains store large volumes of freshwater and provide important meltwater sources for downstream water supply, especially in dry years (Chen et al, 2019). The WASA (Model of Water Availability in Semi-Arid Environments) hydrological model was originally developed by Güntner and Bronstert (2004) for runoff simulations in semiarid regions, and was subsequently extended by Duethmann et al (2013, 2015) to represent snow and glacier melt runoff processes in the Central Asian glacierized basins. The model parameters were automatically calibrated by the εNSGAII algorithm using three different observations including streamflow measured at the basin outlet, satellite daily SCA in the melting period from March to August, and observed annual GMB of the Golubin glacier. Considering the large uncertainty of climate change projections from multiple climate models and emission scenarios, seven temperature (T) changes scenarios by linearly perturbing the observed time series of T in the reference period of 1997– 2016 by up to 7◦C were set up to represent the potential T warming in the study basin by the end of this century. In cases that rainfall could not satisfy the simulated ET, the difference between ET and rainfall were subsequently subtracted from snowmelt and ice melt

RESULTS

Limitations

CONCLUSIONS

DATA AVAILABILITY STATEMENT