Unit Affiliation: Ocean and Climate Physics, Lamont-Doherty Earth Observatory (LDEO)
Hurricanes are the most devastating natural disasters facing many coastal communities in the US and around the world. Since these storms, also called tropical cyclones (TCs), develop in warm air over warm sea surface temperatures (SSTs), it is natural to ask wether they will pose a greater threat as the world warms. Basic thermodynamics implies that the strongest hurricanes in a cooler climate are weaker than their counterparts in warmer climates, and this strengthening with warming has now been detected in the observed record. But the strengthening of the strongest storms may be just one of a number of changes in TC activity that result from warming. One question here is how TC activity will be influenced by the pattern of ocean surface warming, as observations and simulations both show an uneven pattern of tropical ocean warming accompanying greenhouse gas increases.
Some clues to how the regional distribution of SST increases is likely to affect TC activity can be found in El Nino events(also called El Nino/Southern Oscillation events, or ENSOs), in which the equatorial Pacific warms between the west coast of South America and the dateline. Prior work by the Principal Investigators (PIs) and others shows that the influence of El Nino events on TCs is different depending on whether the El Nino-induced warming is greatest in the central or eastern equatorial Pacific. In particular they found that all El Nino events can suppress TCs in the Atlantic, but the central Pacific "flavor" of El Nino is substantially more effective at Atlantic TC suppression. The El Nino-TC connection can thus serve to guide expectations as to how the regional pattern of future SST warming might affect TC activity. It may also have direct implications if future warming affects either the frequency and intensity of El Nino events or the relative occurrence of the central and eastern Pacific flavors.
Work performed under this award examines the El Nino-TC relationship and its implications for climate change using a combination of climate model simulations, machine learning techniques, and hybrid dynamical-statistical models. One goal of the project is to develop "emergent constraints" that connect the El Nino-TC relationship found in climate models to the relationship found in the observed record. If models show a strong connection between the El Nino-TC relationship in present-day climate and future TC change, then future TC change projections from models which correctly simulate the present-day El Nino-TC relationship are more credible. The physical mechanisms through which El Nino events influence TC activity are also explored.
The work has societal value given the risks posed by TCs for coastal communities and the value of guidance as to how that risk might change in the future. The project also builds a research network spanning three institutions from the eastern seabord to Hawaii. In addition to its scientific value the network helps to overcome the geographical isolation of the University of Hawaii, a minority-serving institution. The project also supports a graduate student in Columbia University's Bridge to the PhD program, which helps college graduates from underrepresented groups with the transition to graduate school. Finally, the PIs are developing an online class titled "Machine learning applications across weather and climate" which will take advantage of tools and results from the project.