Sensitivity of Hurricane Harvey's Texas Rainstorm to the Underlying Soil Moisture Condition

The 2017 North Atlantic hurricane season brought with it devastating socioeconomic impacts. Six tropical cyclone (TC) landfall events occurred along the United States Gulf coast, with two of those landfalls associated with category-4 hurricanes. In addition to the damaging wind and storm surge that accompanied these landfalling hurricanes, extremely heavy rainfall occurred and led to freshwater flooding. The most extreme rainfall and flooding occurred in conjunction with Hurricane Harvey, which made landfall northeast of Corpus Christi, Texas, just after 0000 UTC 26 August 2017 as a category-4 storm. Following landfall, Harvey was sustained at tropical storm intensity for the next 84 hours as it remained nearly stationary over southeast Texas and produced over 1200 mm of rainfall in regions from Houston to Beaumont, Texas. The historic inland freshwater flooding in the southeast Texas coastal zone contributed to over 30 fatalities and an estimated $190 billion in damage.

The ability of Harvey to maintain tropical storm strength over southeast Texas for over 3 days after landfall while producing historical rainfall and flooding prompts questions about the role of the underlying land surface condition in Harvey's structural evolution, maintenance, and focus of precipitation. The aim of this presentation is to combine NASA satellite and reanalysis products with an ensemble of high-resolution Weather Research and Forecasting (WRF) model simulations to address three science questions: 1) How did the soil moisture and precipitation evolve as Harvey made landfall and remained nearly stationary for 84 hours? 2) How did the heterogeneously evolving soil moisture/land surface conditions influence the physical processes that drive heavy precipitation? 3) What influence did the evolving underlying soil moisture conditions have on Harvey's intensity after landfall? The WRF ensemble simulations were configured with 4-km horizontal grid spacing, initialized at 1200 UTC 23 August, and utilized three different initial soil moisture conditions: control (GFS analysis soil moisture), saturated, and dry soils. The initial soil moisture conditions were allowed to evolve based on rainfall in one set of simulations and were held fixed in another.

Results show that soil moisture dramatically increased within ~100 km of the Texas coastline in the 84 hours subsequent to Harvey's landfall. The overall rainfall in the Texas coastal zone was historic in all of the WRF simulations, indicating that the rainstorm itself may not have been sensitive to the underlying soil condition. A total column water budget supports this assertion, indicating that the rainfall was driven primary by integrated water flux convergence (IWC) along a coastal baroclinic zone. The IWC was sustained for over 72 hours resulting in the exceptional rains and flooding. The potential influence of underlying soil moisture on Harvey's post-landfall intensity is a bit more complex. The WRF simulations with dry soils produced a weaker vortex overall, but with a larger diurnal variation in vortex intensity compared to the control and saturated soil conditions. The WRF simulations with saturated soils did not produce a more intense vortex after landfall compared to the control simulation. Overall, saturated soils did not have a notable positive impact on postlandfall intensity and dry soils did have a clear negative impact overall in the WRF simulations. This result suggests that the treatment of soil moisture conditions in numerical models may be an important component in the prediction of postlandfall intensity of slow-moving TCs such as Harvey.