Q. Virtual Forecasting System

From SCOOP Wiki

Jump to: navigation, search

This figure shows the work flow for the generation of analytic wind products and for the operation of CH3D-SSMS within the SURA SCOOP forecasting system, run through the Virtual Appliance. For an accurate forecast of storm surge and inundation, knowledge of the atmospheric state at the future time is required. Thus, the creation of an ensemble of forecasted atmospheric states is essential. The first step in doing this is collecting data through several data acquisition models. Forecasted position, minimum pressure, maximum wind speed, and translational velocity are then compared with “best track” data sets. This is done to find errors and probabilities, from which an ensemble of forecasted atmospheric states can be created. The forecasted track and intensity are obtained from the Automated Tropical Cyclone Forecast (ATCF) File / “Deck” Data, while the official historical track and intensity data is obtained from the National Hurricane Center (NHC) Atlantic basin hurricane database (HURDAT). These two sets of data are then compared. Minimum pressure and maximum wind speed values are compared directly at specific forecast hours to determine forecast errors. Position errors are determined in a very similar way. The minimum pressure and maximum wind speed errors are analyzed, while the mean translational velocity during the measurement sampling period is used to form a probabilistic “rotation angle”. All of the data is combined and a probabilistic ensemble of forecasted sets is created. The forecast tracks are then converted to analytic wind and pressure fields, which are in turn supplied to the storm surge models. Each track represents an equal probability of occurring, so the resulting simulated storm surge for each the models of the ensemble set is then combined into a single forecast product. This product then shows the probability of flooding exceeding a certain critical value.



Personal tools