Adaptation of the Advanced Hurricane WRF for driving a storm surge prediction model

1. INTRODUCTION

A wide spectrum of tropical cyclone surface wind fields has been used to drive storm surge prediction models, ranging from parametric wind models, to steady-state dynamic PBL models, to inner-core kinematic analyses, to sophisticated non-hydrostatic NWP models (Cardone and Cox 2009). Though parametric wind models have distinct operational advantages that maximize the number of hours of forecast utility (direct coupling to ocean models, extreme computational efficiency, minimal data I/O), they cannot reproduce far-field winds that generate precursor surges with a high degree of accuracy or replicate unbalanced/fine-scale features such as supergradient inflow or spiral rainbands. On the other hand, extreme care must be exercised when interpolating NWP model gridded wind fields to storm surge model finite-element mesh nodes in both space and time. Otherwise, errors can be introduced that cause along-track elliptical distortions in the shape of the isotachs, resulting in an artificially weak representation of the storm, especially for fast-moving tropical cyclones. To obtain more accurate storm surge predictions, these wind generation methods are now being combined. GWAVA parametric wind model fields and H*Wind analyses (Powell et al. 1998) are being fused through an objective analysis technique and assimilated into the Advanced Hurricane WRF model. The winds are simulated at high horizontal resolution (1 km), output at high frequency (10 minutes), then interpolated to the ADCIRC coastal ocean models grid domain to run storm surge simulations.

2. NUMERICAL MODEL BLEND

The GWAVA (gradient wind asymmetric vortex analysis) wind model (Mattocks and Forbes 2008) is based on Holland (1980), with the added feature that the radius of maximum winds varies azimuthally around the cyclone to capture asymmetry in the shape of the storm. A cross-isobar inflow angle and a directional surface roughness parameterization that modulates the wind speed at a given location based on the types of land cover encountered upwind are applied to represent surface friction. These parametric winds, generated on-the-fly from NHC forecast advisory/best track information in a computationally efficient manner, are available at exact analytical resolution. Thus, they can be directly coupled to an atmosphere/ocean/climate model at every time step and grid point while the model is running. GWAVAs numerics and physics have been extensively upgraded since the model was first utilized in ADCIRC for generating real-time, event-triggered forecasts of storm surge beginning with the 2006 hurricane season. Hindcasts of recent storms demonstrate that this wind forcing produces realistic estimates of storm surge (Fig. 1).

The Advanced Hurricane WRF (AHW) (Davis et al. 2008) is a moving-nest, vortex-tracking version of WRF-ARW that includes drag saturation at high wind speeds and a one-dimensional columnar, mixed-layer ocean model to more accurately simulate vertical momentum/heat exchange. A snapshot of a wind field from an AHW simulation of Hurricane Ike is shown in Fig. 2. This simulation incorporates temporally varying 1-km SPoRT MODIS SST analyses, which significantly improves the simulation of central pressure at landfall when compared with the WRF-default 50-km real-time global SST analyses (Fig. 3).

ADCIRC (Luettich et al. 1996) is a finite-element hydrodynamic model used to simulate wind-driven storm surge, tides, riverine flow and inundation. The unstructured triangular grid includes all waters in the western Atlantic, Caribbean and Gulf of Mexico. Several high-resolution (30 m) grid meshes are available, with computational points draped across inlets and waterways, aligned with shoreline and elevation contours.

3. ANTICIPATED RESULTS

It is well documented that numerical model predictions of hurricanes suffer from significant track and intensity errors (Franklin 2008). One of the most widely used techniques developed to remedy these problems is the insertion of a three-dimensional bogus symmetric Rankine or Holland vortex into the initial state of the atmosphere. The vorticity, geopotential height and velocity perturbations associated with the previously analyzed location of the tropical cyclone are removed in a process known as vortex relocation prior to insertion of the new idealized bogus vortex into the flow field. However, as noted by Rhome et al. (2004), this axisymmetric spinup procedure can inadvertently destroy the environmental wind shear upstream, allowing a storm to intensify unrealistically.

It is anticipated that the assimilation of fused GWAVA/H*Wind analyses into AWH will restore detailed mesoscale structure in the numerically simulated wind fields and mitigate errors in storm track/intensity, while avoiding the problems caused by vortex relocation/insertion. When employed as forcing to drive ADCIRC, these enhanced wind fields should improve the quality of storm surge simulations, as measured by predicted maximum water surface elevation and RMS error in station hydrographs.

REFERENCES

Cardone, V.J. and A.T. Cox, 2009: Tropical cyclone wind eld forcing for surge models: critical issues and sensitivities, Natural Hazards, 51, 29-47.

Davis, C., W. Wang, S.S. Chen, Y. Chen, K. Corbosiero, M. DeMaria, J. Dudhia, G. Holland, J. Klemp, J. Michalakes, H. Reeves, R. Rotunno, C. Snyder, and Q. Xiao, 2008: Prediction of landfalling hurricanes with the Advanced Hurricane WRF model. Mon. Wea. Rev., 136, 1990-2005.

Franklin, J.L., 2008: National Hurricane Center forecast verification, 28th Conference on Hurricanes and Tropical Meteorology, AMS, 28 April - 2 May 2008, Orlando, Florida, 4pp.

Holland, G.J., 1980: An analytical model of the wind and pressure profiles in hurricanes. Mon. Wea. Rev., 108, 1212-1218.

Luettich, R., J. Hudgins and C. Goodall, 1996: Initial results from a combined tide and storm surge forecast model of the U.S. East Coast, Gulf of Mexico and Caribbean Sea. 15th Conference on Weather Analysis and Forecasting, AMS, Norfolk, VA, August 1996, 547-550.

Mattocks, C. and C. Forbes, 2008: A real-time, event-triggered storm surge forecasting system for the state of North Carolina, Ocean Modelling, 25, 95-119.

Powell, M., S. Houston, L. Amat, and N. Morisseau-Leroy, 1998: The HRD real-time hurricane wind analysis system. J. Wind Engineer. and Indust. Aerodyn., 77-78, 53-64.

Rhome, J.R., R.D. Knabb and M. Bender, 2004: An examination of Hurricane Debby (2000) with the 2000 and 2003 Versions of the GFDL hurricane prediction system, 26th Conference on Hurricanes and Tropical Meteorology, AMS, May 3-7, 2004, Miami, FL, 2pp.