The HWRFx Modeling System: Coupling to a One-Dimensional Ocean Model with Multiple Vertical Turbulence Closures

To correctly forecast the potential contribution of the ocean to tropical cyclone (TC) intensity, the ocean component of coupled forecast models must accurately predict the pattern and rate of SST cooling relative to the storm center. The Hurricane Weather Research and Forecasting Experimental model (HWRFx) is an offshoot of the NCEP HWRF model that was developed by OAR specifically to study and improve forecasts of track, intensity, and structure with emphasis on rapid intensity change. Since the initial incarnation of HWRFx was not coupled to an active ocean, our eventual goal is to couple the atmospheric model to the HYbrid Coordinate Ocean Model (HYCOM) along with a surface gravity wave model. As an interim step during this development, a one-dimensional ocean model has been incorporated into HWRFx as a module containing one-dimensional subroutines called separately at each grid point on both the parent domain and the movable nests. These subroutines were extracted from three-dimensional HYCOM code. This approach will enable future twin forecast experiments using the 1-D and 3-D versions of HYCOM to be compared for a large number of storms to document conditions where 3-D ocean dynamics may significantly impact SST cooling and TC forecasts. The 1-D ocean model contains multiple state-of-the-art vertical turbulence closures, enabling the impact of these parameterizations on SST cooling and intensity evolution to be quantified. Idealized tests of this version of HWRFx for Hurricane Ike (2008) verify the necessity of incorporating an active ocean. They further demonstrate that although turbulence closure has a small effect on the track forecast, it has a potentially significant impact on intensity. Results from a suite of idealized and realistic experiments using the 1-D ocean model, the latter including a repeat of the 69 cycles of numerical forecasts from hurricane seasons 2005 and 2007 originally conducted using HWRFx without an active ocean, will be presented. The critical importance of high-quality ocean observations, particularly upper ocean profiles, for improving ocean model initialization and evaluating ocean model performance will be emphasized.