Real-time Hurricane Forecast Products in support of HFIP Transition to Operations

- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner
Monday, 3 February 2014: 4:00 PM
Room C106 (The Georgia World Congress Center )
Paula T. McCaslin, NOAA, Boulder, CO; and D. A. Zelinsky and T. Quirinno

The Hurricane Forecast Improvement Program (HFIP) provides the basis for NOAA and other agencies to coordinate hurricane research needed to significantly improve guidance for hurricane track, intensity, and storm surge forecasts. It also engages and aligns the inter-agency and larger scientific community efforts towards addressing the challenges posed to improve hurricane forecasts. The ultimate goal of HFIP is to transition new research and development into NOAA operations.

In support of the HFIP Transition to Operations (T2O) work, a webpage was developed that displays hurricane forecast guidance products that are operationally available, along with experimental products being tested for T2O in a consolidated fashion, see www.hfip.org/products. The experimental products are provided by various HFIP contributing organizations into one centralized location. The webpage showcases deterministic and ensemble model output, ATCF tracker data, probability products, and displays ensemble track data on Google maps.

The work is carried out by a team comprised of ten individuals representing HFIP funded groups. The team developed a common method for displaying experimental forecast model output for evaluation and comparison. This involved creating common image generating scripts, developing a file naming convention, and creating a common location for all HFIP model image output where the images would be backed up to a mass storage device and downloaded to the HFIP web server. The team's willingness to work together sharing ideas has made the resulting product appear seamless.

NOAA's supercomputing system called 'Jet' is being used to run most of the forecast models. Forecast model output is gathered on a disk partition where each organization has a named directory to stage a set of model output data for transfer to a specified web server, namely hfip.org hosted at NCAR. Members of this team have created GrADS image generating scripts and wrapper scripts to run on multiple model output formats, GRIB and NetCDF, containing a particular subset of forecast parameters specific to tropical cyclones as requested by NHC forecasting specialists. Images are presented in a common, user-friendly format that is easily accessible to our primary customer, NHC forecasters.