Impact of Adjusting the Deep Convection Parameterization on HAFS Intensity Forecast Improvement

Optimizing the deep convection entrainment coefficient (c_o) of the scale-aware Simplified Arakawa-Schubert (SAS) deep convection scheme played a pivotal role in the operational implementation of the Hurricane Analysis and Forecast System version 1 configuration A (HAFSv1A). The increase of c_o from its default value of 0.1 to 0.15 improved the overall storm intensity forecasts, especially for longer forecast lead times. In this study, the impacts of c_o on the improvement of HAFSv1A forecasts will be documented and analyzed, in order to understand the underlying physical mechanisms that improve tropical cyclone (TC) intensity forecasts, focusing on a tropical cyclone science perspective.

Motivation for adjusting c_o is related to the over-intensification issue in one of the past HAFS real-time experiments, which was also found in the operational Hurricane Weather Research and Forecasting model. Specifically, the intensity of certain TCs was over-predicted, when these storms encountered moderate to strong vertical wind shear (VWS) or existed in dry environments. Although the adjustment of c_o to 0.15 somewhat improved the intensity forecast for major hurricanes, the large improvement in intensity forecasts stemmed from several over-intensified TC cases. Even though forecast outcomes exhibited variability across cycles, the over-intensification was mitigated in multiple cycles of certain over-predicted TC cases, consequently reducing wind speed errors. In conclusion, mitigating the over-intensification issue contributes to improving the intensity forecasts of HAFS-A.

One plausible explanation for this improvement is that the increased c_o reduces excessive convection in VWS and dry environments. Under moderate or strong VWS, dry air attempts to infiltrate the core region of the TC. By augmenting the entrainment rate of the deep convection scheme, excessive convection is suppressed, thus ameliorating the problem of TC over-intensification.