Readily available but underused Doppler radar data can help predict the height of the planetary boundary layer - the lowest part of the atmosphere and where weather forms - and, in turn, improve severe weather forecasts, according to scientists at Penn State.
"This boundary layer data has never been used this way," said Yunji Zhang, assistant professor of meteorology and atmospheric science at Penn State and lead author of the study. "It may help us improve the forecasts of rainfall and other kinds of severe weather, and that's always the goal."
The researchers said using the data to estimate the height of the boundary layer and adding this to their numerical weather models improved forecasts in a case study of storms in 2022 storms that brought deadly flash flooding to eastern Kentucky. The team reported their findings in the journal Monthly Weather Review.
Zhang said the work takes advantage of a previous discovery that that National Weather Service's Doppler radar network can be used to estimate the top of the planetary boundary layer by detecting changes in moisture and turbulence that occur at its upper boundary. That finding was made by team member David Stensrud, professor of meteorology and atmosphere science, and his collaborators, at Penn State, the University of Oklahoma, and the National Severe Storm Laboratory.
And because there are more than 150 National Weather Service Doppler radar stations across the country that scan their section of the atmosphere every few minutes, the data provides an advantage over other observations. Ground-based weather stations, for example, miss what is happening in the sky above and satellites and weather ballons can only take measurements when they pass overhead or are launched.
"This data provides valuable information that can help us better observe how the boundary layer evolves during the day, or that can be put into our models to improve forecasts," said Zhang, who is also assistant director of the Penn State Center for Advanced Data Assimilation and Predictability Techniques. "Either way, we should make good use of them, because they are valuable, and they fill a gap that no other observation planforms can do."
Understanding conditions in the boundary layer is particularly important because it is where ingredients of most severe weather live, as well as where air pollutants build up and spread. Zhang said knowing the depth of the boundary layer can reveal details about the conditions inside and the potential for severe weather.
"Environmental conditions like temperature, moisture and wind contribute to boundary layer depth," Zhang said. "So, a better estimation of the boundary layer depth might translate to better estimations of these conditions."
The scientists added this new information to their numerical weather prediction models using data assimilation, a statistical method that can paint the most accurate picture of current weather conditions. This is important, Zhang said, because even small changes in the atmosphere can lead to large discrepancies in forecasts over time.
The researchers applied the model to weather forecasts from a series of overnight storms in July 2022 that brought heavy rain and devastating floods to eastern Kentucky.
Zhang said that using surface observations alone, the team found incorrectly predicted where the heaviest rainfall would occur. Adding the boundary layer height observations resulted in more accurate predictions.
"With better forecasts, we can provide more accurate and timely warnings that help local emergency management, residents and businesses prepare," Zhang said. "If you do not have a good forecast, or the forecast is off by a few hundred kilometers, the preparations are basically off as well."
This work was part of the Consortium for Advanced Data Assimilation Research and Education, or CADRE, a multi-university data assimilation effort funded by the National Oceanic and Atmospheric Administration (NOAA).
"We are working closely with our NOAA collaborators to try and transfer some of our knowledge and technologies in terms of applying these observations in our models to their systems," Zhang said. "This may help improve some forecasts that are produced in a more operational scenario."
In addition to Zhang and Stensrud, Braedon Stouffer, a doctoral candidate, and Lyn Comer, who earned a master's degree in 2023 from Penn State, also contributed.
The U.S. National Science Foundation and NOAA supported this work.
