A Collaborative Approach for Tropical Cyclone Wind Watches and Warnings

The National Hurricane Center (NHC) formally separated tropical cyclone wind and storm surge watches and warnings (WWs) for the U.S. Gulf and Atlantic coasts starting with the 2017 Hurricane Season, in recognition of the fact that these two hazards do not always occur at the same times and/or locations, and because the safety actions required are usually different for the water and wind threats. The storm surge WWs are issued through a collaborative process between NHC and impacted Weather Forecast Offices (WFOs) via AWIPS2, where the first guess fields are determined from a probabilistic storm surge model (P-surge). This paper describes a conceptual framework for a similar process for tropical cyclone wind watches and warnings. NHC runs a wind speed probability (WSP) model that estimates the probabilities of 34, 50 and 64 kt winds out to 120 h. This model could serve as a first guess for the wind WWs, similar to how the P-surge model is used for the storm surge WWs. It will be shown that there is already a close relationship between the output from the WSP model and NHC’s coastal wind WWs, even though there currently is no quantitative procedure relating the two. Recent modifications to the WSP model also make its use for the inland portion of the wind WWs issued by the WFOs more appropriate. The AWIPS2-based NHC-WFO collaboration process currently used for surge could therefore be adapted to wind WWs to collaboratively improve the spatial and temporal consistency of the coastal and inland wind WWs.

The WSP model is primarily statistical, where it uses the past five years of NHC track and intensity errors, and errors from a simple climatology and persistence wind structure model to estimate the probabilities. Long-range plans for transitioning the WSP model to use information from dynamical model ensemble systems will also be presented.