Estimating Biases in Tropical Cyclone Climatology Due to the Lack of Two-Way Air-Sea Coupling in High-Resolution Prescribed SST Experiments

Both observational and regional modeling studies have demonstrated that tropical cyclones (TCs) induce a cold wake in sea surface temperatures which can persist for days or weeks after a storm's passage. This anomaly serves as a negative feedback on storm intensification by reducing surface enthalpy fluxes, especially for stronger, slow moving systems. Traditionally, high-resolution atmospheric general circulation models (AGCMs) have used prescribed sea surface temperatures (SSTs) as a surface boundary condition due to the computational expense and potential mean state biases associated with coupling to a prognostic, 3-D ocean model. However, as horizontal resolutions, and therefore simulated TC intensities, increase in global models, the potential for biases to be introduced by the lack of surface energy closure grows larger.

In this presentation, a highly-simplified slab ocean parameterization that has been developed for use in conjunction with the Community Atmosphere Model (CAM) is described. The parameterization simulates TC-induced SST cooling due to both surface fluxes and mixing processes. The scheme allows for computationally-efficient simulations where air-sea feedbacks can be explored without the need for a fully-dynamic ocean model. Two sets of high-resolution (25km) model runs will be discussed. One utilizes the traditional AGCM framework of prescribed, climatological SSTs, while the other is coupled to the interactive slab ocean. The design of the slab ocean configuration allows for the direct isolation of the impact of two-way air-sea coupling on mean TC intensity curves. Further discussion will highlight changes in storm genesis and lifetime that arise with the different ocean realizations, as well as the effect of TC-TC interactions via cold wake crossing. More generally, the acceptability of using fixed SSTs at increasing resolution in global climate models will be addressed.