The Impact of Ensemble Ocean Data Assimilation on Coupled Tropical Cyclone Forecasts

The ocean component of coupled tropical cyclone (TC) prediction system is important for TC track and intensity forecasts, because sea surface temperature (SST) affects the development of tropical cyclones through sensible and latent heat fluxes into the storm. Hence, accurate SST and oceanic upper layer simulations are critical for accurate forecasts

In this study, a local ensemble transform Kalman filter (LETKF) data assimilation system is developed to improve ocean initialization for the HYbrid Coordinate Ocean Model (HYCOM) in the coupled tropical cyclone forecasting models - HWRF (Hurricane Weather and Research Forecast) and HMON (Hurricanes in a Multi-scale Ocean-coupled Non-hydrostatic model). As a preliminary work, a two-week coupled HMON-HYCOM ensemble forecasts with varying physics schemes are performed from GEFS (Global Ensemble Forecast System) initial conditions to generate the HYCOM ocean ensemble. After the two-week spin-up, ocean observations are assimilated in a daily analysis cycle using 4D-LETKF. The assimilated observations include T/S (Temperature and Salinity) profiles from Argo floats, XBTs, and CTDs; in situ SST (sea surface temperature) from buoys, drifters, and ship observations; satellite retrieved SST; and along track SSH (sea surface height) from the satellite (Jason 2 in this study). The ocean analysis from the HYCOM-LETKF system provides the ocean initial condition for coupled TC forecasts.

This data assimilation system is tested for Hurricane Isaac (2012). The ocean initial conditions are verified against satellite and in situ ocean observations. The impact of assimilating ocean observations on the track and intensity forecast of Hurricane Isaac is examined and discussed. Data denial experiments are designed to identify the individual impact of different ocean observation platforms on coupled TC forecasts.