Parameter Regionalization Based Global Hydrological Model Development,Simulation,and Prediction

The global hydrological model is an important tool for the analysis and assessment of spatial distribution and temporal changes of global terrestrial water resources,as well as an important scientific tool for projecting the impact of climate change and human activities on water resources in future.Since it is difficult to calibrate the parameters of the global hydrological model,the uniform parameters are generally used,which limits its application in a large area with no or lack of data.It is an urgent need to improve the accuracy of model simulation by adopting regionalized model parameters associated with land surface heterogeneities.In this paper,based on the analysis of previous studies,the global hydrological model FLEX-Global was developed based on the regionalization of the hydrological model parameters to improve the simulation accuracy,and the performance of the FLEX-Global was verified by comparing with the measured data and with other global hydrological models.This study used the future climate change data from several global climate models and scenarios in the latest 6th International Coupled Model Comparison Program CMIP6 as the driving force for global terrestrial(excluding Greenland and Antarctica)river discharges projections and trend analysis.The main research efforts and findings are as follows:(1)FLEX-Global is a global hydrological model that uses the FLEX hydrological model as a framework,coupled with the MCT-HSC runoff generation module,and uses the global river routing model Ca Ma-Flood for river network routing.The spatial resolution of the model is 0.5°,and the daily time step is used to simulate river discharges on global scale.To test the accuracy of the model,the FLEX-Global model was first validated by measured discharge of 26 major rivers in the Global Runoff Data Center(GRDC),and compared with seven existing global hydrological models.The results of the multi-indicators evaluation showed that the FLEX-Global model with parameter regionalization can improve the simulation accuracy of river discharge,and the application of the runoff generation module MCT-HSC in the FLEX-Global model can also improve the model performance to a certain extent,which was especially true in the tropical climate region,arid climate region and temperate climate region with significant improvement.(2)In future projection with FLEX-Global.In terms of time,the future global runoff was likely showing an increasing trend under the SSP1-2.6 scenario,which may be related to an increasing trend in precipitation and a slow rise in temperature.However,under the SSP5-8.5 scenario,the two climate models selected in this paper drive the simulated global land runoff change trends to show opposite trends,which may be caused by the opposite changes in precipitation in the two climate models.This indicated that there are uncertainties in the global climate models in predicting future climate change.Spatially,under the SSP1-2.6 scenario,the water resources in the middle and low latitude zones,mainly in central North America,central-eastern Africa and East Asia,may increase and the possibility of flooding may increase;in the middle and high latitude zones,mainly in northern Asia,southern Europe and western Africa,the water resources may decrease and the possibility of drought disaster may increase.Under the SSP5-8.5 scenario,the water resources in the high latitude zones and the northern Indian Ocean coast in the low latitude zones,mainly in northern North America,northern Asia and central South America,central-eastern Africa,East Asia and South Asia may increase,and the possibility of flooding may increase;in the middle and low latitude zones,the water resources mainly in south-central North America,northern South America,southern Europe,western Africa,western Oceania may decrease,and the possibility of drought disaster may increase.This may be related to the changes in precipitation in these areas,and the increase in precipitation variability will increase the risk of floods and droughts in some areas.(3)The projected future water resources in the Yangtze River basin may increase under both scenarios.Under the SSP1-2.6 scenario,the discharge in the middle and lower reaches of the basin mainly showed an increasing trend,with a higher probability of flooding,while the discharge in the upper reaches mainly showed a decreasing trend,with a higher probability of drought.Under the SSP5-8.5 scenario,the discharge in the upper reaches of the basin mainly showed an increasing trend,with a higher possibility of flooding;the discharge in the middle and lower reaches mainly showed a decreasing trend,with a higher possibility of drought.The projected future water resources in the Yellow River basin may mainly show an increasing trend under the SSP5-8.5 scenario.Under this scenario,the discharge in most areas of the basin may increase,and the possibility of flooding in the main stream areas may increase,while the possibility of drought in the basin may decrease.However,there were large uncertainties in the water resources changes and the spatial distribution of possible flood and drought areas under the SSP1-2.6 scenario.The two global climate models selected in this paper as the driving data for the simulated runoff changes in the Yellow River basin did not manifest consistent conclusions.The FLEX-Global model developed in this paper solves the problem that most global hydrological models use uniform parameters,improves the performance of global hydrological models in some areas to a certain extent,and makes a useful attempt to improve the performance of global hydrological models.On this basis,the latest CMIP6 climate model was used as the driving data to simulate the future global land river discharge based on the FLEX-Global model,and the temporal and spatial distribution of future water resources change trends were discussed,which can provide references for water resources management,flood and drought disaster prevention and control on a global scale.In future research,it is also necessary to consider the impact of human activities,reduce the uncertainty of the hydrological parameters used to build the model,improve the spatial resolution,and more accurately simulate the spatial distribution and temporal evolution of global terrestrial water resources.