March 19, 2010 | A new study has verified that the Weather Research and Forecasting model (WRF) can depict seasonal snowfall in Colorado with a high degree of accuracy. The research could be especially useful to scientists studying the impacts of climate change on water resources in the western United States.
Until now, regional climate modeling systems have struggled to accurately simulate seasonal snowfall and snowpack. The headwaters region of the Colorado River, a critical water resource for Colorado and the Southwest, is particularly challenging for climate models to depict due to complex terrain.
The researchers, led by NCAR scientist Roy Rasmussen, used the advanced research version of WRF (ARW) to perform simulations of snowfall between November 1 and May 1 for four past winters, all in the 2000s. They compared the model results with data from SNOTEL (SNOwpack TELemetry) sites. The Colorado River headwaters region contains more than 100 of these sites, which measure accumulated snowfall.
The team found that the ARW simulations and SNOTEL data showed very good agreement when resolutions below 6 kilometers (3.7 miles) were used, showing that ARW is capable of reliably simulating snowfall over a full winter during a variety of conditions. Simulations run at 2 km (1.2 mi) suggest that other global and regional models underestimate high-elevation snowpack and overestimate low-elevation snowpack. At higher resolutions (18 and 36 km, or 11 and 22 mi), ARW underestimated SNOTEL snowfall by 20–40%.
“Recent advances in the WRF model have enabled us to produce more accurate simulations of snowfall,” Rasmussen says. “These high-resolution simulations provide the precipitation fields needed to properly simulate snowpack and runoff in the complex terrain of the Colorado Rockies.”
WRF was created through a partnership that includes NOAA, NCAR, and more than 150 other organizations and universities in the United States and abroad.