How Increasing Resolution Changes the CONUS Water Cycle
Simulations of the water cycle and its processes are a useful tool for water management, but biases in the models limit the value of those simulations for decision-making. Increasing model resolution to capture more fine-scale circulation features is one way to reduce model biases. Increasing model resolution, however, does not improve all aspects of the simulated water cycle, and much work is still needed to understand which aspects of the water cycle respond positively to increasing resolution and which do not. Researchers show that increasing resolution in the Energy Exascale Earth System Model (E3SM) leads to a slowdown of the water cycle over the continental United States (CONUS). The researchers demonstrated which water budget terms improved from this slowdown and which were degraded.
Understanding the water cycle response to increasing resolution helps us better understand when the value justifies the increased cost. Understanding when to make use of high-resolution earth system models provides researchers with a better understanding of how best to use these tools to test scientific hypotheses.
The water cycle is an important component of the earth system, and it plays a key role in many facets of society, including energy production, agriculture, and human health and safety. Researchers used the Energy Exascale Earth System Model version 1 (E3SMv1) to run with low-resolution (roughly 110~km) and high-resolution (roughly 25~km) configurations to evaluate the atmospheric and terrestrial water budgets over the CONUS at the large watershed scale.
At higher resolution and during the warm season, all terms in the water budget become smaller: precipitation, evapotranspiration, moisture convergence, and runoff. The most pronounced changes with resolution to the water cycle come from reductions in precipitation and evapotranspiration. The reductions in evapotranspiration lead to improvements over nearly the entire CONUS, while other terms show mixed results when increasing resolution. Additional exploratory metrics with expected resolution sensitivity, including precipitation and streamflow extremes, storm events, and snowpack, show modest improvements in E3SMv1.