Impacts of Microphysical Uncertainties in All-sky Satellite Infrared and Microwave Radiances Data Assimilation for TC Prediction

Effective assimilation of satellite passive all-sky brightness temperatures (or Tbs, converted from radiances) is vital to improving tropical cyclone (TC) analyses and forecasts. Cloud affected Tbs, including infrared (IR) Tbs from geostationary satellites and microwave (MW) Tbs from low-Earth-orbit satellites, are sensitive to the microphysical characteristics of clouds. Model clouds are, in turn, sensitive to the microphysics parameterization (MP) scheme used to create them. Differences in MP schemes are known to influence the outcomes of forecasted TC tracks and intensities. However, less is known about the impact of the interaction between MP scheme properties and assimilation of satellite all-sky IR and MW Tbs on TC analyses and forecasts.

In this study, a Weather Research and Forecasting model and ensemble Kalman Filter (WRF-EnKF) data assimilation (DA) system is used to compare two MP schemes and their TC analyses and forecasts for the 2017 Atlantic Hurricane Season. The two MP schemes are the WRF single-moment-6-class scheme (WSM6; Hong and Lim 2006) and the hybrid (double-moment cloud ice and rain) scheme by Thompson et al. (2008; hereafter referred to as THO). Consistent with previous studies, ensemble backgrounds of WSM6 experiments tend to produce significantly less cloud than observed and have model-equivalent IR Tbs significantly warmer than observations. Such warm biases result in strengthened vortices in IR_WSM6 (i.e., experiments with WSM6 and assimilation of only IR Tbs) analyses in the early WRF-EnKF cycles. For the IR_THO experiments, the ensemble backgrounds tend to produce wide-spread cloud coverage and relatively cold model-equivalent IR Tbs from deep clouds. The cold biases in Tbs from deep clouds in the backgrounds of members tend to result in vorticity removal. In addition, the IR_WSM6 experiments add more water vapor to the analyses than IR_THO (Fig. 1). Finally, the IR_WSM6 analyses with their stronger vortices and higher water concentrations than for IR_THO tend to lead to stronger storms in the forecasts than IR_THO (Fig. 2).

To date, this investigation indicates that 1) numerical TC forecasts are significantly impacted by MP scheme differences within the satellite data assimilation forecast-and-analysis framework, and 2) better understanding of the interactions between MP scheme properties and IR and MW Tbs would produce insights on weaknesses in the MP schemes and the satellite DA system, leading to a path forward for improving TC prediction.

IRMW_WSM6 and IRMW_THO experiments (i.e., with IR and MW Tbs assimilated together) are currently ongoing. The additional assimilation of MW Tbs is likely to improve the analyzed structures and distributions of hydrometeors of TCs, thus potentially leading to better forecasts.