Do we have the Necessary Ingredients for Hurricane Intensity Forecasting?

Shuyi S. Chen
U of Miami


Rapid intensification and decay in hurricanes continue to be a major challenge for hurricane prediction. The lack of skill in forecasts of storm structure and intensity may be attributed to deficiencies in the current prediction models: insufficient grid resolution, inadequate model physics such as PBL/surface and microphysics parameterizations, the lack of full coupling to a dynamic ocean, and the lack of observations and/or data assimilation for accurate model initial conditions. In recent years, major efforts have taken place in research and operational communities to address these model deficiencies and have shown somewhat limited improvements in intensity forecasts, which vary from case to case. A fundamental question remains unanswered: what is the predictability of hurricane intensity (e.g., measured by peak winds)? How model error and initial condition error affect intensity prediction? The model error and initial condition error are not easy to separate. A better understanding of these questions will help guide the overall effort of improving intensity forecasts. Ensemble forecasts using a stochastic kinetic-energy backscatter scheme have shown some promise to shed new light on these issues. Although the results are very preliminary, this talk is intended to provide food for thought and invoke a discussion on hurricane intensity forecasting at a fundamental level.