Close Window

University of Wisconsin-Madison: "Evaluation of new data and modeling techniques for forecasting"

Final Report

The purpose of this project was to develop a cooperative research program between the University of Wisconsin-Madison (UWM) and the NWS Forecast Office at Sullivan, Wisconsin that would provide the WSFO with advanced meteorological information to aid in the forecast process. NWS personnel would then evaluate these data and help improve their usefulness by identifying strengths and weaknesses and suggesting improvements. Those types of data available through the modernized weather service program were emphasized, including output from numerical models and imagery and derived products from weather satellites.

One of the main areas of collaboration was in forecast model product evaluation. The collaboration involved the operational use and evaluation of a locally run nonhydrostatic mesoscale model. A system was developed to execute the University of Wisconsin Nonhydrostatic Modeling System (UW-NMS) operationally and provide products which could be of use to field forecasters. As part of this development, the mesoscale model was modified to accept input from the Eta model for initial fields and large-scale boundary conditions. Extensive work was then begun to create forecast fields of precipitation, pressure, temperature, humidity, and winds at observation levels. Forecasting problems singled out for evaluation included lake effect storms, quantitative precipitation forecasting on the river basin scale, and severe storm forecasting. Model results were available to the forecast office over the Internet (, and near the end of the project, the 27-km grid had been increased to include more of the Chicago, Duluth and La Crosse forecast areas. The model physics, specifically the cumulus parameterization scheme, were also upgraded using information extracted from the monitored performance of the operational simulations. This work has also involved a COMET graduate student fellow (Brad Hoggatt).

In addition to lake-effect snow forecasting and research, a quantitative precipitation forecast (QPF) verification scheme was developed in conjunction with the Sullivan NWS office. By comparing observed precipitation amounts for each of the river basins in Wisconsin with operational UW-NMS QPF for each basin, the researchers quantitatively determined model performance for all types of atmospheric flow patterns.

Three additional areas were the focus of the research efforts with the Space Science and Engineering Center:

1) GOES-derived product imagery. NWS forecasters examined and are provided feedback on the university's derived-image products, which were created from GOES radiances. These hourly products (precipitable water vapor, stability, and surface skin temperature) depicted the evolution of important atmospheric and surface structures. At the time the project ended, these products appeared to have limited use in the operational setting. Other additional products were under development as part of the NWS Lake Effect Snow Project and were being evaluated by the NWS Sullivan office.

2) Satellite-derived cloud algorithms to support ASOS. The university completed a demonstration of the utility of Visible Infrared Spin Scan Radiometer Atmospheric Sounder cloud parameters inferred from multispectral observations for augmenting conventional ground observations. The university provided the forecast office with the three-hourly Sounder Site Specific Satellite Cloud product on RAMSDIS. This is a more detailed product than is normally available through the Family of Services and includes all height and cloud amount information generated for each GOES sounder field of view.

3) GOES sounder products. The NWS provided evaluations of the utility of new products developed from the GOES sounder.