3D Visualization of modeled hurricane forecasts and their impacts at landfall

Improved forecasts of hurricane tracks and intensities are needed to prepare coastal communities in order to avoid extensive costs incurred by evacuation and storm damage. The NASA finite-volume General Circulation Model (fvGCM) and the Florida State Superensemble are two models being developed for use in forecasting large-scale weather phenomena. Students working at the NASA Applied Sciences DEVELOP Program are working to visualize outputs from these models in an effort to demonstrate the societal benefits of NASA data usage and to streamline data visualization capabilities for the science community. The finite-volume General Circulation Model (fvGCM) is a global climate model developed at the NASA Goddard Space Flight Center. The model forecasts of the intensity and path of hurricane Ivan during 2004 are investigated. The experiment, referred to as an Observing System Simulation Experiment (OSSE), used the fvGCM model to help assess the impacts of wind measurements from a potential space-based wind lidar. Two forecasts are examined. The first, referred to as the control run, is initialized with data from the nature run; this run is used to simulate results produced using data from existing remote sensors. Next, an additional forecast is produced by initializing the model with the data used in the control run combined with additional measurements. Wind measurements as would be produced by a space-based lidar instrument are the added initialization parameters. Each of these forecasts is then compared to the nature run for accuracy. The effect of the added measurements is then analyzed to improve the model. The DEVELOP Southern United States Disaster Management team created visualizations of each of the forecasts used in the aforementioned OSSE. Three-dimensional visualizations were then produced from each forecast. The model initially produces a 0.36° by 0.25° global dataset. Each data point contains U, V, and &omega wind vector components and humidity for four distinct altitudes. The model also produces a sea-level pressure measurement for each point on the grid. To visualize the forecasts, sea-level pressure measurements for a 126.72° longitude by 46.75° latitude area were converted into a contour isobar map by creating an IDL script. Next the plots were input into ArcGIS 9 as part of the final visualization. To create the 3D surface plot of the pressure, a Matlab script was created to access the same region and convert it into a geo-referenced columned tab-delimited text file. This particular format can be read by the ArcGIS 9 software and converted into a shape (SHP) file. These shape files were then rendered into the final three dimensional surface using the functionality of ArcGIS 9. Bringing both components together with a United States map outline created the final visualization. Prior to the conference, we will visualize additional forecasts from the fvGCM model with additional parameters. We will then visualize damage estimates for storm landfall associated with each of the forecasts by using outputs from the Federal Emergency Management Agency's (FEMA) Hazard United States – Multi-hazard (HAZUS-MH) model. This will in turn demonstrate varying results associated with multiple hurricane forecasts.