Skip to main content
U.S. flag

An official website of the United States government

Publication Date
1 October 2021

Hydrological Analysis in Watersheds with a Variable-Resolution Global Climate Model (VR-CESM)

A global climate model with areas of refined resolution is a viable tool for watershed-scale hydrological analyses.
Print / PDF
Powerpoint Slide
The Merced River in Yosemite National Park (USA) feeds Lake McClure, which provides hydropower and water to residents and agriculture in California’s central valley. Fine resolution is needed to reliably simulate hydrology in such complex watersheds.

The variable-resolution community Earth system model (VR-CESM) is a global climate model with refined resolution in specific regions. This study checks VR-CESM water flow against stream flow observations and another model in four undammed watersheds in the western U.S. Future changes in water flow are estimated in these and four watersheds in eastern China. VR-CESM estimates water flow well after. Future water flow may be higher in winter and lower during spring in warmer, snow-dominated watersheds. Elsewhere, changes in precipitation drive changes in water flow. Year-to-year variation in water flow may increase in the western U.S. and eastern China.


VR-CESM bridges the gap between detailed watershed models used for longer-term planning and global climate models that estimate future changes. VR-CESM can be used to estimate potential changes in stream flow in particular areas, without the extra step of changing coarse global data to finer regional data. This will motivate the improvement of water modeling in global models, including dams and human water usage, which will extend the scope of hydrological analyses. Very detailed models will continue to be needed for short-term water management, but models like VR-CESM have the potential to revolutionize longer-term water resource planning.


A major goal of the study is to show that a global model can make reliable stream flow estimates for use in water resource planning. We use VR-CESM in this study to model historical and future climate and hydrology at fine resolution in two specific regions with one global simulation. Most of the globe has one-degree resolution while eastern China and the western U.S. has 1/8-degree resolution (~14km). We have shown that VR-CESM accurately estimates water flow in undammed watersheds, without having to downscale global data to regional data. Calibrating the model to monthly stream flow data is important to achieve model accuracy. Seasonal water flow is estimated to shift in some watersheds and year-to-year variations in water flow are estimated to increase in both the U.S. and China. Improvements to human water management in the model will make VR-CESM a reliable new tool for regional water planning.

Point of Contact
Alan Di Vittorio
Lawrence Berkeley National Laboratory (LBNL)
Funding Program Area(s)
Additional Resources: