February 22, 2011 | Last year, a team of NCAR scientists verified that the Weather Research and Forecasting model (WRF) can be used to depict seasonal snowfall in Colorado with a high degree of accuracy. The team used an advanced research version of the model to perform simulations of snowfall for four past winters, all in the 2000s, and compared the model output with observational data, finding good agreement at resolutions below 6 kilometers (3.7 miles). (More about the study can be found here.)
The team has now taken the research a step further by using WRF to forecast future snowfall in the Colorado headwaters region. Because it is expensive to run high-resolution models such as WRF for decades into the future, the researchers took a different approach. They added a climate signal to WRF’s daily simulations of the present, warming and moistening the atmosphere to simulate climate conditions predicted for 2045–2055.
The results predict enhanced snowfall on the order of 10–25% in the Colorado headwaters region due to the availability of more moisture in the atmosphere. (The model runs did not simulate changes in storm track, so these results would likely be an upper limit to the expected increase.) Higher elevations experience increased snowfall, according to the study, with lower elevations experiencing enhanced melting; changes in peak snow mass are near zero due to these compensating effects. Total wintertime runoff is above current levels, showing evidence of earlier melting of the snowpack.
The study was able to capture region-specific aspects lacking in global climate models, such as reduced precipitation in the core headwaters region due to the rain shadow effect of mountains upstream and the details of different headwater river catchment basins. In addition, WRF’s ability to model elevation accurately enabled the team to quantify how the microphysical changes from an elevated freezing level may increase precipitation.