Understanding the geographic distribution of tropical cyclone formation for applications in climate models

Abstract

Projections of Tropical cyclone (TC) formation under future climate scenarios are dependent on climate model simulations. However, many models produce unrealistic geographical distributions of TC formation, especially in the north and south Atlantic and eastern south Pacific TC basins. In order to improve confidence in projections it is important to understand the reasons behind these model errors. However, considerable effort is required to analyse the many models used in projection studies. To address this problem, a novel diagnostic is developed that provides compelling insight into why TCs form where they do, using a few summary diagrams. The diagnostic is developed after identifying a relationship between seasonal climatologies of atmospheric variables in 34 years of ECMWF reanalysis data, and TC detection distributions in the same data. Geographic boundaries of TC formation are constructed from four threshold quantities. TCs form where Emanuel’s Maximum Potential Intensity, $$V_{{PI}}$$ , exceeds $$40\,\,{\text{ms}}^{{ - 1}}$$ , 700 hPa relative humidity, $$RH_{{700}}$$ , exceeds 40%, and the magnitude of the difference in vector winds between 850 and 200 hPa, $$V_{{sh}}$$ , is less than $$20~\,\,{\text{ms}}^{{ - 1}}$$ . The equatorial boundary is best defined by a composite quantity containing the ratio of absolute vorticity $$(\eta )~$$ to the meridional gradient of absolute vorticity ( $$\beta ^{*}$$ ), rather than $$\eta$$ alone. $${\beta ^*}$$ is also identified as a potentially important ingredient for TC genesis indices. A comparison of detected Tropical Depression (TD) and Tropical Storm (TS) climatologies revealed TDs more readily intensify further to TS where $${V_{PI}}$$ is elevated and $${V_{sh}}$$ is relatively weak. The distributions of each threshold quantity identify the factors that favour and suppress TC formation throughout the tropics in the real world. This information can be used to understand why TC formation is poorly represented in some climate models, and shows potential for understanding anomalous TC formation behaviour in the real world.