Wednesday, 10 January 2018
Exhibit Hall 3 (ACC) (Austin, Texas)
Regional climate model simulations have been shown to be sensitive to model setup and configuration, including boundary conditions. To configure the Weather Research and Forecasting Model (WRF) over the Lake Champlain Basin, ERA-Interim Reanalysis data at 0.75º spatial resolution are used to force hindcast simulations for model evaluation and optimization. The long-term goals of this work include downscaling General Circulation Model (GCM) data under various climate change scenarios to 4 km using three one-way nests. Because most GCMs are of coarser resolution, usually on the order of 1.5°, problems may arise when the resolution of boundary conditions is suddenly changed. To examine the impact of varying boundary condition resolutions on model performance, we compared simulated precipitation obtained using ERA-Interim Reanalysis input conditions at 0.75º (ERA-0.75) and at 1.5º of resolution (ERA-1.5) before and during Tropical Storm Irene. The magnitude of precipitation differences due to differing boundary resolutions was compared to output from four additional perturbation simulations, each of which had the same input conditions as ERA-0.75, but with slightly different outer domain sizes and shapes to gauge internal model variability. Upon analyses of cumulative and daily rainfall across the region, ERA-1.5 did not differ significantly from ERA-0.75 and stayed within the range of the perturbations. The same conclusion was reached when examining rainfall during the top four events in the time period.
- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner