With its record-breaking 30 hurricanes and tropical storms in 2020, hundreds of people died in the United States. There were also tens of millions of dollars of damage. One important question is: Will this be the future?Although most climate scientists agree that hurricane intensity, at least in terms rainfall, will increase with global warming, there is uncertainty about the future frequency and severity of hurricanes. Climate models today offer a range possible futures. Some predict an increase in North Atlantic hurricane frequency while others predict a decrease. These contradictory results raise the question of whether these models are capable of forecasting hurricane frequency, or missing a vital process.Peter Huybers, Professor of Earth and Planetary Sciences at Harvard John A. Paulson School of Engineering and Applied Sciences(SEAS), said that "to understand whether these models can be trusted we must see if they reproduce the past." While models today do a great job simulating hurricanes over the past 40 years, data and models tend to diverge as time goes by. This begs the question, "If models don't reproduce the long-term history and hurricanes, can we trust their long term predictions?"Huybers and a group of scientists published a paper in Science Advances that showed these models can reproduce the long-term history hurricanes. However, this is only possible if historical sea surface temperatures are corrected. This research highlights the importance to understand sea surface temperature patterns. It suggests that better understanding these patterns could help reconcile model predictions and improve our understanding about how climate change will affect future hurricane frequency.This paper expands on previous research by Huybers and his colleagues who identified biases in historical sea surface temperature measurements. They also developed a comprehensive method to correct these data. The paper was published in 2019 and led to better understanding of the ocean's warming over time. The researchers use the same correction here to model historical hurricane frequency.The formation of hurricanes is influenced by the sea surface temperature."The frequency and severity of Atlantic hurricanes is dependent on the sea surface temperatures, especially the warmth of subtropical North Atlantic," Duo Chan, who was a graduate student at SEAS, and the first author of this paper.The subtropical North Atlantic can be relatively warm and it causes more atmospheric convection. This leads to more Atlantic hurricanes. Hurricane formation rates drop when the subtropical North Atlantic becomes relatively cold, partly because winds that shear apart protostorm systems cause them to be less active.Today's climate models attempt to replicate past hurricane seasons. They generally forecast too few between 1885-1900, and too many between 1930-1960. However, these models rely on historical sea surface temperatures which indicate a subtropical Atlantic that was relatively cool at the beginning of the 20th Century and a warmer Atlantic by the middle of the century.Huybers & Chan have shown in past research that historical sea surface temperatures are subject to systematic errors. The corrected sea surface temperatures reveal a subtropical North Atlantic that is warmer from 1885-1920 and cooler between 1930 and 1960. These adjustments have brought hurricane frequency into line with observations.Huybers stated that "our corrections to sea-surface temperature patterns were independently created, and significantly improves the skill with which models reproduce historic hurricane variations." These results improve our confidence in both historical sea surface temperatures as well as our models, and give us a firmer basis to study how climate change will impact hurricane frequency moving forward.###Gabriel A. Vecchi of Princeton University and Wenchang Yang from Harvard University co-authored the research, which was partially supported by Harvard Global Institute.