The ability to predict North Atlantic hurricanes activity months in advance is essential to prevent major impacts on societies and ecosystems of these natural hazards, being hurricanes (and typhoons, in the Pacific ocean) the most intense expression of tropical cyclones. In a study just published on PNAS – Proceedings of National Academy of Sciences of the United States of America, researchers report a strong correlation between the depth-averaged temperature over the upper 40 m in the eastern North Atlantic and September tropical cyclone (TC) activity in the North Atlantic, likely reflecting high ocean temperatures driving reduced trade winds and reduced wind shear in the tropics, essential factors in the genesis of an hurricane. “In this study”, says Enrico Scoccimarro, CMCC researcher at CSP – Climate Simulation and Prediction Division and lead author of the study, “we found out a new predictor of Atlantic hurricane activity, that is the sub-surface ocean temperature (upper 40-m averages) over the eastern Atlantic domain. Our results suggest that not only considering ocean sub-surface temperature instead of sea surface temperature (SST) provides additional predictive skill of hurricane activity, but also that it’s important to include the non-local effect associated with eastern Atlantic Ocean temperature, via a reduction of wind-shear, in seasonal forecasts of tropical cyclone activity. “

The most active seasons occur for lower than normal wind shear conditions over the Main Development Region of North Atlantic (where hurricanes tend to form), which are also driven by reduced trade wind strength, conditions usually associated with higher than normal sea surface temperatures in the eastern Atlantic Ocean. So, the authors noticed that a significant step toward operationally reliable TC activity predictions is gained after including upper ocean mean temperatures over the eastern Atlantic domain. “Ocean temperature in the East Atlantic modulates in fact trade winds intensity while influencing wind shear and therefore hurricane activity; with the new predictor included in our study, we can forecast three months in advance (in June-July) tropical cyclone activity in September, the month with the most TC activity”, Scoccimarro says.

The study is the results of the more than ten-year research on hurricanes and global ocean of the CMCC Foundation divisions CSP – Climate Simulation and Prediction and ODA – Ocean Modelling and Data Assimilation.
CSP Division lead by researcher Silvio Gualdi (among the authors of the study) has been involved in the study of hurricanes on a global scale for over ten years and has recently shifted its focus on the seasonal scale to provide a more reliable forecast of hurricane activity in the North Atlantic Ocean.
 In order to investigate the relationship between the thermal state of the Atlantic Ocean and the tropical cyclone activity in terms of Accumulated Cyclone Energy (ACE), researchers used observed 1980-2015 tropical cyclone records, and a 1/4 degree resolution (~ 25 km horizontal spatial resolution) global ocean reanalysis. This high resolution global ocean reanalysis has been developed by ODA Division lead by Dr. Simona Masina (among the authors of the study). Based on NEMO model (a state-of-the-art modelling framework of ocean related engines), it allows to obtain a very detailed information of the ocean state (temperature, salinity, currents, etc), even more complete than information provided by only observational ocean data (and it was essential to include the upper 40-m ocean average temperature in the study – satellite data for instance provide sea surface temperatures only -). Last but not least, nothing would have been possible without the computational power of the CMCC Supercomputing Center based in Lecce, where models have been run to obtain the data sets used in the study.
“The next step”, Scoccimarro adds, “will be to use the results of this study to implement an hurricane forecasting system up to three months in advance of September, typically the peak of the tropical cyclone season. The idea is to extend the temporal horizon of our forecasts: we want to be able to predict early summer ocean subsurface temperature in the eastern Atlantic domain two months in advance.”

The team of authors was led by researcher Enrico Scoccimarro (CSP). Among the authors, all CMCC researchers, Alessio Bellucci (CSP), Andrea Storto, Silvio Gualdi (CSP Division Director), Simona Masina (ODA Division Director) and Antonio Navarra.
Read more:
 The paper on PNAS “Remote subsurface ocean temperature as a predictor of Atlantic hurricane activity,” by Enrico Scoccimarro et al., 2018