Conference Call Attendees: Jonathan Blaes-RAH, Michael Brennan-NHC, John Billet-AKQ, Bob Bright-CHS, Frank Alsheimer-CHS, Reid Hawkins-ILM, & Bryce Tyner-NCSU
The TC Inland Winds group held its monthly conference call on 6/13/12. The call began with a discussion of Bryce’s recent NDFD verification using HWind analyses. The latest, 24, and 48 hour forecasts were analzed for average bias in the forecast, normalized relative to the average analyzed HWind analyzed wind speed at each location. Results continue to suggest a general overprediction in wind speeds for almost all areas and storms. The only exception was Gabrielle (2007), but this storm just grazed the Outer Banks and is not a representative case of a “typical” TC affecting the study region. Many of the storms displayed a higher bias for the latest forecast cycle prior to the analyzed time. Jonathan hypothesized that this is likely due to a increased confidence near the time of landfall in storm track that would lead to impacts from the TC, thus leading to forecasters to increase forecasted wind speeds. It was suggested to analyze the bias in other manners, including separate plots for each analyzed HWind time as well normalization to maximum HWind analyzed wind speeds throughout all forecast cycles, rather than taking an average.
Bryce then updated the group on the land use dataset he plans to use to calculate frictional decay based on wind direction. He is following a model presented at the April Tropical Conference and working with the developers of the model to get the product. The 2001 land use data has been downloaded and he is in the process of programming the model.
Bryce discussed potential alternatives to the Rankine vortex interpolation used in the TCMWindTool. Holland et al. (2010, MWR) present one alternative that requires very little additional input and is likely a better interpolation method for the winds. Holland et al. model is currently being coded by Bryce and will be examined to see if it indeed leads to significant improvements in the initial model wind field for recent landfalling TC cases.
Jonathan concluded the phone call by discussing potential deliverables from the project. They show a progression from most easy to execute to more long term goals.
The next conference call will be held on Wednesday, July 11th at 11 AM