Assessing Extreme Weather Risk

We can’t prevent extreme weather, but if we know the risks, our communities can minimize the damage with better construction and early warnings. That knowledge starts with science.

At Lamont-Doherty Earth Observatory, our scientists have been making breakthroughs in the development of computer models to help communities around the world assess their risks of extreme weather events, including hurricanes, storm surges, extreme rainfall, drought, and tornadoes. Assessing risk requires understanding the complex forces that create extreme weather, as well as reconstructing climate and weather patterns through history and being able to project how those patterns might change in the future.

Simulating Storms

Hurricanes are among the most extreme weather events on the planet, bringing high winds, heavy rain, and destructive storm surge flooding. Scientists have some understanding of the complex forces within a hurricane, but knowledge of how garden-variety clouds first aggregate into the dangerous swirling systems is still being developed.

It’s an area Suzana Camargo and Allison Wing have been working on, with a focus on radiative-convective feedbacks and “self-aggregation” – the spontaneous clustering of individual clouds into larger, more coherent weather systems – to improve computer models of hurricane behavior that are necessary for assessing risk. Camargo has also been working on indices that connect environmental variables to cyclone frequency.

To provide insight into the hurricane and storm surge risk a city might face, Chia-Ying Lee has been developing a hurricane risk model that can help determine how intense hurricanes would become under different climate conditions. She uses climate models to obtain expected changes in the drivers of storm intensity, including a warmer sea surface, cooler upper atmosphere, and effects of large scale circulation. Her model will eventually be coupled with storm surge models to obtain more accurate estimates of the risks of catastrophic storm surge to coastal cities worldwide.

Adam Sobel and Ji Nie are developing another method for modeling extreme precipitation events that picks apart the different factors that make the event extreme. It separates out the role of mountains, for example, or moisture in the air. “Unlike running a model lots of times to provide the probability of an event occurring, you can start with the event and then change the conditions, making it little warmer, a little cooler, adding a little more moisture in the air, and you see how much stronger it gets. Ours is a novel way of doing this that applies to extreme rainfall events,” Sobel said.

The New Frontiers

Hurricane modeling has matured in the 15 years since Sobel and Camargo started in the area. Today, the new area is the modeling of tornadoes and hail, and Lamont scientists are at the forefront here, as well.

Tornadoes are notoriously difficult to model. Even weather forecasts can only model storms that have tornado potential, not the tornadoes themselves. Sobel and Michael Tippett of the Columbia Engineering School have been modeling the conditions that can lead to a tornado, such as high wind shear and high instability (a hot, humid surface and cold upper atmosphere) and working in current climate variables that can affect those conditions to develop seasonal forecasts a month or two out. They watched in 2015 as their risk assessment of tornado activity during El Niño years played out as projected, with more tornado activity in Texas and Florida but less over the normally active plains states. They’re also working on risk assessment for hail, which can do significant damage to crops.

“It’s important that the research we’re doing be able to help answer the applied, practical questions that are relevant to public planning and risk management,” Sobel said.

Several Lamont projects that you’ll read about in this Annual Report are directly translating science into impacts on society, the economy, and resources. Studies led by Park Williams and Benjamin Cook have explored the drought history of the western United States over more than 1,000 years, showing the severity of recent droughts and why California’s drought has been driven by more than natural variability. Justin Mankin assessed the loss of snowpack and the risk to heavily populated basins dependent on snow melt for water and irrigation. Richard Seager and colleagues studied the drought in Syria and its connections to civil unrest. Maureen Raymo is estimating sea level rise from earlier episodes of high sea level preserved in the geological record.

Learning from the Past

Climate history is critical for all of this work. To understand what is changing in our Earth and atmospheric systems and to design computer models that can accurately reflect those changes, we need a window into the past.

Using tree ring analysis and sediment cores, Lamont scientists have been opening that window by reconstructing drought, temperature and rainfall extremes through time, revealing patterns of natural climate variability and historic risk trends. Ed Cook and an international team of colleagues published their third drought atlas in 2015, this one covering Europe and the Mediterranean over the past 2,000 years. For studying hurricane history, Raymo and a group of students began testing a technique this past year for reconstructing the history of hurricane strikes over thousands of years using sediment cores from freshwater ponds normally protected by dunes.

Scientists from across Lamont and Columbia University are working on many other aspects of extreme weather. Some are addressing its connections to public health, others to urban planning, and disaster risk management. In 2015, Lamont launched the Initiative on Extreme Weather and Climate to pull together these and other scientists from institutions around the world for collaborative work that can expand integrated scientific assessment and develop comprehensive, risk-sensitive engineering solutions.