"With this particular study aircraft-based Doppler radar information was consumed in to the system," stated Fuqing Zhang, professor of meteorology, Penn Condition. "Our forecasts were similar to or much better than individuals produced by operational global models."
Zhang and Erin B. Munsell, graduate student in meteorology, used The Pennsylvania Condition College real-time convection-enabling hurricane analysis and predicting system (WRF-EnKF) to evaluate Hurricane Sandy. While Sandy made landfall around the Nj coast around the evening of March. 29, 2012, case study and forecast system started monitoring on March. 21 and also the Doppler radar data examined covers March. 26 through 28.
The scientists in comparison The WRF-EnKF forecasts towards the National Oceanic and Atmospheric Administration's Global Forecast System (GFS) and also the European Center for Medium-Range Weather Predictions (ECMWF). Besides the opportunity to effectively assimilate real-time Doppler radar information, the WRF-EnKF model includes high-resolution cloud-enabling grids, which permit the presence of individual clouds within the model.
"Our model predicted storm pathways with 100 km -- 50 mile -- precision four or five days in front of landfall for Hurricane Sandy," stated Zhang. "We had accurate forecasts of Sandy's intensity."
The WRF-EnKF model also runs 60 storm forecasts concurrently being an ensemble, each with slightly varying initial conditions. This program operates on NOAA's devoted computer, and also the analysis ended around the Texas Advanced Computing Center computer due to the enormity of information collected.
To evaluate the Hurricane Sandy forecast data, the scientists divided the 60 incurs groups -- good, fair and poor. This method could isolate questions within the model initial conditions, that are at their peak on March. 26, when 10 from the forecasts recommended that Sandy wouldn't make landfall whatsoever. By searching only at that area of the model, Zhang indicates the errors occur due to variations within the initial steering level winds within the tropics that Sandy was baked into, rather than a mid-latitude trough -- a place of relatively low atmospheric pressure -- in front of Sandy's path.
"Although the mid-latitude system doesn't strongly influence the ultimate position of Sandy, variations within the timing and placement of their interactions with Sandy result in considerable variations in rain fall predictions, especially regarding heavy precipitation over land," the scientists report inside a recent problem from the Journal of Advances in Modeling Earth Systems.
By 2 days before landfall, the WRF-EnKF model was precisely predicting the hurricane's path with landfall in southern Nj, as the GFS model predicted a far more northern landfall in New You are able to and Connecticut, and also the ECMWF model forecast landfall in northern Nj.
Hurricane Sandy is a great storm to evaluate because its path was unusual among Atlantic tropical storms, that do not usually turn northwest in to the mid-Atlantic or Colonial. While the 3 models did a reasonably good job at predicting facets of this hurricane, the WRF-EnKF model was very promising in predicting path, intensity and rain fall.
NOAA is presently evaluating using the WRF-EnKF system in storm conjecture, along with other scientists are utilizing it to calculate storm surge and risk analysis.
