As global warming changes weather patterns, Arthur DeGaetano works on methods to anticipate big rainfall events in the northeastern United States.

Tomorrow’s Forecast: Improved Accuracy

by Jackie Swift

Climate change is here to stay, bringing with it changing weather patterns. According to current models, the northeastern United States can expect an increase in extreme rainfall events. The remnant of Hurricane Ida that hit New York City in the summer of 2021, dropping more than seven inches of rain in less than 24 hours and flooding the subway system, is an example of the impact these types of storms can have.

“In this part of the world, I think the effects of climate change are going to manifest themselves in hydrology — in flooding and things like that,” says Cornell professor Arthur T. DeGaetano, Earth and Atmospheric Sciences. “We’ll see this particularly along the coast. The storm sewers in New York City, for example, drain into the Hudson River and into Long Island Sound. As sea level rises, that water will build up, and we will have more of these events on the landscape, where water coming off the pavement or coming down the river exacerbates the flooding.”

DeGaetano’s research focuses on applied climatology, the development of methods and data sets to produce accurate climate information that a wide range of decisionmakers — from governmental departments to engineering firms — can use to make informed choices. He is also the director of the federally supported Northeast Regional Climate Center, which seeks to enhance the use and dissemination of climate information to a wide variety of sectors in the northeastern region of the United States.

Identifying Weather Trends

With funding from the National Oceanic and Atmospheric Administration (NOAA), DeGaetano and his lab are addressing the need for accurate information on current rainfall, mainly in New Jersey and Pennsylvania. “Any engineering design for a drainage system or green infrastructure in a city, for example, is based on the historical observed rainfall record,” he explains. “NOAA has developed atlases of rainfall values, but they are based on data through the year 2000. Prior to that, they used data through the year 1960, so these atlases are not updated very often.”

Analyzing regional rainfall data for the past two decades, DeGaetano and his colleagues created mathematical adjustment factors that can be used to multiply rainfall numbers in a standard NOAA atlas, thereby updating the published value to reflect current trends. “With the changing climate, we’re trying to answer the question, What has the trend been through time?” DeGaetano says. “If you’re building a drainage system right now, you don’t want to base your decisions on data last collected in 2000. What does the situation look like today, with 20 more years of data?”

Rethinking Time-Honored Methods

The researchers also addressed questions related to the best methods of rainfall analysis in a changing climate. As part of that, they are championing the idea of automating the process so that data can be incorporated as soon as they come in rather than waiting for a fixed interval. “That may sound easy,” DeGaetano says. “But many aspects come into play, so it’s not as straightforward as it seems.”

“If you’re building a drainage system right now, you don’t want to base your decisions on data last collected in 2000.”

To begin with, the researchers looked at the priority given to data from weather stations with the longest possible rainfall records, often a hundred years or more. Given that rainfall climatology has been changing dramatically through time, they questioned whether the oldest data were relevant. “Where is the trade off?” DeGaetano asks. “You’re looking at extremes, so you do need a fairly long data record, but you might not want the longest record possible because those data collected very early in the period don’t reflect current climatology. You’re almost guaranteeing that you’re going to be too low because you’re using rainfall values at the very low end of the trend.”

DeGaetano and his colleagues found that data covering the past 70 years yields the most accurate analysis of current climate trends. In addition, they concluded that the time-honored practice of including data from weather stations that cover only short, discrete periods of time — say the 40 years from 1920 to 1960 — should be discontinued. “When you’re doing extreme value analyses, two things work against you: a short period of time and a data series that has a past trend that no longer represents current dynamics,” he explains. “Using a fairly small period of record — half of it observed well in the past when rainfall wasn’t as extreme as it is now — is a double whammy.”

Anticipating Future Rainfall

Along with employing historical data to identify ongoing rainfall trends, the DeGaetano lab uses climate model projections to anticipate future extreme rainfall. In a multistep process called downscaling, researchers take data from climate models, which cover large spatial areas, and make the data relevant on a more local level, DeGaetano explains. “From these downscaled data we’re able to apply many of our same procedures and analyses to give people indications of what extreme rainfall will look like in the future, so they can adjust their engineering standards to reflect those future conditions,” he says.

This process is especially important because climate experts anticipate that in the next 50 years rainfall amounts will increase from 25 to 30 percent in the Chesapeake Bay region, from the Southern Tier of New York down to Virginia. “When you have a 25 percent increase in rainfall, that means that a big storm event you can expect to occur once every 100 years will now be occurring every 40 or 50 years,” DeGaetano says. “So you will see these big, big events that cause a lot of flooding to occur almost twice as often as you would expect them.”

Currently, DeGaetano and his colleagues are turning their attention to how well climate models are able to simulate the seasonality of rainfall. In the northeastern United States, for example, big rainfall events tend to happen twice a year — in the fall during hurricane season and in the spring when mid-latitude storms, such as late-season nor’easters, can transport tropical moisture into the region. “These two types of weather patterns cause the most extreme rainfalls in this region, so the question is, how well do our climate models replicate that twice-a-year climatology?” DeGaetano says. “Do they only get the values of storms that occur in the spring, for instance?”

Needs-Driven Research

DeGaetano’s research is driven by the needs of stakeholders, and with that in mind, he emphasizes the accessibility of his findings to those who posed the questions in the first place. Often that means translating his work to an interactive website such as the one connected to the Northeast Regional Climate Center. Much of his research is also applied to agricultural questions, such as when to spray for a certain pest or when to irrigate, he explains.

“I started in grad school looking at how weather affects agriculture,” he says. “And I still do a lot of work in that area today. A lot of the data we have in our data sets are used quite extensively by the agricultural community, not only in New York but throughout the Northeast.”

Originally published on the Cornell Research website. All rights are reserved in the images. If you’d like to reproduce the text for noncommercial purposes, please contact us.

Cornell research moves quickly. Keep up with our monthly e-newsletter.

--

--

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store