Researchers have developed a new approach to enhance the capability of Earth System Models (ESMs) in simulating the impacts of small-scale land surface differences. By incorporating a subgrid structure and downscaling atmospheric variables, the study shows significant improvements in modeling snowfall, snow water equivalent, and runoff, particularly in regions with mountainous landscapes. This breakthrough paves the way for more accurate predictions of water cycles and resources, ultimately supporting effective water management plans. Earth system models and hydrology are the key focus areas.
Earth System Models As Disruptor
Although Earth system models (ESMs) are powerful tools to simulate and project changes in the climate, these ESMs have been traditionally operating on coarse grid resolutions of about 50-200 kilometers. The ability to obtain high-frequency, spatially fine-grained constraints on land surface heterogeneity has been a perennial challenge in the realm of climate modeling.
To solve this problem, researchers propose a new method that adds a ‘subgrid’ component to the ESMs. It appropriately inserts the cost of downscaling atmospheric variables (e.g., precipitation, temperature) from the atmosphere grid to a much finer scale subgrid topographic units. These improvements in E3SM Land Model (ELM) implementation have dramatically increased representation of the effects of fine-scale land surface contrasts.
Using Fine-Scale Topography to Uncover the Magic
They found that the new land surface subgrid and downscaled atmospheric variables have profound impacts on key land surface processes when analyzing the ELM simulations. The simulation of snowfall, snow water equivalent, and runoff was greatly improved—particularly in mountainous areas (where the land is harder to model and high-precision data is often lacking) and cool-season maximum precipitation regions.
That improvement is underscored in the paper by the model’s capacity to re-create actual snow water equivalent (any liquid water contained within the snow) at the Snow Telemetry (SNOTEL) sites throughout the western U. S.. An ability of ELM to realistically represent the observed frequency of snow water equivalent at 83% of the SNOTEL sites, a significant improvement in the model skill, has such unprecedented potential for understanding regional and global scale water cycling and its future evolution.
Optimising water resources management
But the conclusions of this study reach far beyond climate modeling. Given the importance of land-atmosphere interactions on hydrologic processes in mountain areas, improvements in ELM can have a profound influence on streamflow and water resources management by providing more accurate representation of these critical land surface process simulations.
Improved predictions of when and where snow will fall, how much water is held as snow on the ground (snow water equivalent), and when that snowmelt will run off are essential for better understanding of the workings of the water cycle—allowing us to more effectively manage our precious freshwater resources. Ultimately, this can result in more reliable and long-lasting water resource management plans that will improve conditions for the dependents of these vital resources—communities and ecosystems.
