S57 Examining Beaver Island Wetland Ecosystems Using Dynamically Downscaled WRF Simulations

Sunday, 7 January 2018
Exhibit Hall 5 (ACC) (Austin, Texas)
Cody Michael Converse, Central Michigan University, Wyoming, MI; and D. B. Kluver

Using the National Center for Atmospheric Research’s Weather Research and Forecasting model (WRF), a dynamically downscaled analysis is performed across three nested domains to observe and document how large-scale weather events influence smaller, more localized meteorological patterns. Beaver Island, the study area, is an island located in northern Lake Michigan, just northwest of Charlevoix. Though it is small (10km x 20km), it is home to wetland ecosystems rich in biodiversity and considerably vulnerable to changes in climate. The innermost WRF domain contains Beaver Island surrounded by northern Lake Michigan (3.5km spatial resolution), the middle domain contains the northern half of Michigan’s lower peninsula and the eastern half of the upper peninsula (7km spatial resolution), and the largest domain contains all of Michigan (28km spatial resolution). The WRF model simulations are run from May 12th, 2016 through August 1st, 2017, and utilize the F0000 forecast hour of the 00Z, 06Z, 12Z, and 18Z daily runs of the National Weather Service’s Global Forecast System (GFS). Verification of the WRF output is provided by cross-comparison with data gathered from in-situ weather stations deployed in wetlands on Beaver Island and nearby Garden Island. Analysis of the hourly WRF output shows measurable differences at the validation sites of both meso-scale and synoptic-scale influences.
- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner