Experimental tool helps improve flash flood forecasts in the Northeast U.S.

Flash flood in Washington, D.C., in July. (Photo by Alek Krautmann, NOAA)

Floods and flash floods kill more people each year than any other severe weather hazard. And a few extra minutes of notice can make a big difference reducing deaths and economic loss. This is why researchers at NOAA’s National Severe Storms Laboratory are partnering with the NOAA National Weather Service Weather Prediction Center to test an experimental flash flood and intense rainfall forecasting tool.

The Warn-on-Forecast System, or WoFS, provides additional information different from what forecasters currently use because it is high-resolution and can update quickly. The weather model focuses on individual thunderstorms and hazards associated with those storms a few hours before they form and as they develop. Ultimately, the new tool will help forecasters issue flash flood warnings earlier.

The Norman-based researchers are collaborating with WPC and several NWS forecast offices to study how they are using WoFS in real-time when making forecast decisions, said Nusrat Yussouf, a research scientist at the University of Oklahoma Cooperative Institute for Mesoscale Meteorological Studies.

“Our evaluation process of research-to-operations back to research helps us improve experimental products,” she said.

This summer the prediction system proved its usefulness. For example, in July when parts of the Northeast and mid-Atlantic were inundated with intense rainfall, WPC forecasters used WoFS as they observed the perfect conditions for flash flooding over the I-95 corridor.

The WoF experimental system showed up to five inches of rain in some areas. The guidance Screenshot of WPC discussion where WoFS utilization is mentioned.provided through WoFS gave forecasters more confidence to use the phrase “flash flooding likely” when they issued area forecasts for parts of Pennsylvania and New Jersey, down to Baltimore, Washington D.C and Virginia. The storms resulted in flooded roads during rush hour, stranded motorists, cancelled and delayed flights, power outages and property damage.

This short-term exploration of the experimental WoFS’s capabilities in NWS operations is valuable for researchers at NSSL and OU CIMMS. Yussouf, whose work supports NSSL, said researchers cannot easily study NWS forecasters’ natural decision-making process in a controlled testbed environment.

“The traditional testbed experiment environments are more controlled with a routine start and end time,” she said. “We’ve created something more organic in operations that gives us insights into how that decision process occurs and how the WoF workflow may look in NWS operations in the future.”

Forecasters provide feedback to researchers throughout the experiment, including products they would like to see and what does or does not work well for them.

Yussouf said the collaboration with WPC is mutually beneficial since the Center focuses on intense rainfall and flash flooding events.

“Our goal is to help provide forecasters more tools to save lives and property,” Yussouf said. “This is one more tool to help them.”

Share this:

Gab at the Lab: Derek Stratman

Derek Stratman, NRC Postdoc


Background:Ph.D. Meteorology, University of Oklahoma (2016)
M.S. Meteorology, University of Oklahoma (2011)
B.S. Meteorology, Valparaiso University (2009)
Experience:Derek was born and raised in Jasper, Indiana, best known for Strassenfest, an annual summer festival celebrating German heritage and culture. He attended Valparaiso University in his home state, earning his bachelor’s degree in meteorology. Then, he moved to Norman to continue his education at the University of Oklahoma. He earned both his Master’s and Ph.D. in meteorology at OU before accepting a National Research Council Postdoc position with NSSL’s Warn-on-Forecast group.
What He Does:Derek began working with the Warn-on-Forecast group in August 2016. His current research is focused on alleviating storm displacement errors in storm-scale forecasts. Previously, he had been an OU graduate research assistant. He worked with NSSL from 2009 to 2011, looking at storm-scale model verification. From 2011 to 2016, he worked with the Center for Analysis and Prediction of Storms on improving storm-scale modeling and data assimilation techniques. He also took part in several field experiments. In 2010, he participated in the Verification of the Origin of Rotation of Tornadoes Experiment 2 (VORTEX2), assisting with mobile mesonet operations and taking surface observations. In 2013, Derek helped coordinate data collection for the Mesoscale Predictability Experiment (MPEX).
Trivia: Derek and his wife recently had their first child. In his free time, Derek enjoys several hobbies, including photography, storm chasing, astronomy, camping/hiking, playing trumpet, and sports.

Share this:

Q&A with Pam Heinselman

Pam HeinselmanNSSL scientist Pamela Heinselman recently transitioned from our Radar Research Development Division to the Forecast Research Development Division, marking a significant shift in her area of focus. Heinselman has been a research scientist with the Lab since March 2009 and received a Presidential Early Career Award for Science and Engineering in 2009 as well. Previously, she was a collaborator with OU’s Cooperative Institute for Mesoscale Meteorological Studies, where her efforts centered on developing phased array radar through experiments in NOAA’s Hazardous Weather Testbed.

While radar research has long been her passion, Heinselman was ready for a new challenge. For many years, her research concentrated on warning and forecast applications of weather radars. Now, she is applying that experience to develop Warn on Forecast, a program aiming to increase tornado, severe thunderstorm, and flash flood warning lead times.

We sat down with her to get her take on how radar and forecasting work together at NSSL.

Q: What inspired you to make the switch to the Warn-on-Forecast group? What do you hope to accomplish in this new role?

A: What inspired me was the opportunity to engage in new challenges, to be immersed in and learn more about this exciting research area, and to contribute to the success of the Warn-on-Forecast program through my skills and experience.

What I hope to accomplish is to work with our in-house scientists and OAR labs and National Weather Service partners at National Centers and local offices to advance and eventually transfer to operations a cutting-edge forecast system that ultimately improves the ability of individuals, families, and communities to protect their lives and property.

Q: How is your position in FRDD related to your work with radar?

A: My position in FRDD is related to my work with radar in several ways. Most importantly, like Phased Array Radar, the Warn-on-Forecast system under development is cutting-edge technology. While Phased Array Radar is introducing adaptive rapid-radar scanning as a potential replacement for the WSR-88D, Warn-on-Forecast is introducing frequently updating, probabilistic high-impact weather forecast guidance as an integral part of a forecasting paradigm shift, known as Forecasting a Continuum of Environmental Threats (FACETS). Another connection is the Warn-on-Forecast program’s exploration of benefits from assimilating legacy and rapid-scan dual-polarization radar data in these model forecasts.

Q: In your Phased Array Radar Innovative Sensing Experiment, you used eye tracking technology to analyze forecaster decision-making. How will the results of this research be useful in developing Warn-on-Forecast?

A: The results of the eye tracking experiment will shed light on forecaster cognitive processes that will aid the development of forecast visualization techniques optimized for the needs of operational forecasters. Additionally, since currently forecasts rely heavily on radar data in their warning decision process, results of the experiment will help to bridge the use of weather radar data with the use of probabilistic forecast guidance in operations.

Q: What do you see as the biggest challenges and opportunities for radar and forecasting research?

A: One of the biggest challenges for radar and forecasting research is finding creative solutions to known technological issues, such as matching co-polar radar cross-sections, reducing model error, and attaining the computational resources needed for forecasting systems with 1-km or smaller grid spacing.

At the same time, one of the biggest opportunities for radar and forecasting research is to revolutionize the frequency and specificity of high-impact weather observations and forecasts to ultimately provide decision makers with more timely guidance that improves their ability to take protective action well in advance of life-threatening events.

Q: How will Warn-on-Forecast address the need for greater lead time and more accurate weather forecasts?

A: Warn-on-Forecast will address the need for greater lead time and more accurate weather forecasts by producing frequently updated, well-calibrated probabilistic 0 to 6 hour convective-scale analyses and forecast guidance that support high-impact forecast and warning operations within NOAA.

Share this:

Gab at the Lab: Kent Knopfmeier

Kent Knopfmeier, Research Associate (NSSL/CIMMS)




Background:M.S. Atmospheric Sciences, Purdue University (2007)
B.S. Atmospheric Sciences, Purdue University (2005)
Experience:Kent is originally from Evansville, Indiana and remained in-state to attend Purdue University, where he earned both his bachelor’s and Master’s degrees in atmospheric science. He came to CIMMS/NSSL on June 15, 2009, as a research associate with the Forecast Research and Development Division’s Warn-on-Forecast group. During his early days with the Lab, Kent was advised by longtime NSSL Research Scientist Dr. Dave Stensrud, who is currently a meteorology professor and department head at Penn State University.
What He Does:Since May 1, 2013, Kent has been the NSSL Liaison to the NOAA Hazardous Weather Testbed. He is the co-developer and manager of NSSL’s Experimental Warn-on-Forecast System. He also manages the HWT’s Experimental Forecast Program website and helps to facilitate the EFP each spring. This year, the Spring Forecasting Experiment will run from May 2 - June 3. Learn more about the EFP (http://hwt.nssl.noaa.gov/efp/ ) and Warn-on-Forecast (http://www.nssl.noaa.gov/projects/wof/ ).
Trivia: Kent is an avid fan of Purdue athletics, and enjoys both watching and playing sports - particularly basketball and golf. With friends, he likes to participate in live team trivia and group card games (especially Euchre). His favorite place to visit is Colorado.
Share this:

Gab at the Lab: Nusrat Yussouf

Nusrat Yussouf, Research Scientist (CIMMS/NSSL)




Background:Ph.D. Computer Science, University of Oklahoma
M.S. Computer Science, University of Oklahoma
B.S. Electrical & Electronics Engineering, BUET, Bangladesh
Experience:After completing her bachelor’s degree in EE from Dhaka, Bangladesh, Nusrat made her long journey across half the world to USA. She joined CIMMS/NSSL as a Research Associate after completing her Master’s from OU in 2002. She completed her Ph.D. while working full time at CIMMS/NSSL. She is currently working as a Research Scientist.
What She Does:Nusrat is part of the Warn-on-Forecast research group in the Forecast Research and Development division at NSSL. Her current research interest is to improve probabilistic forecasts of tornadoes, flash floods, large hails, damaging winds and other hazardous convective weather using ensemble data assimilation and storm-scale numerical weather prediction models. Nusrat enjoys doing research related to Warn-on-Forecast because this work has the potential to save lives.
Trivia: Nusrat has been married for 18 years and has two boys. She likes to travel with her family. She enjoys listening to music, watching movies and cooking Bangladeshi dishes.
Share this:

Gab at the Lab: Corey Potvin

Corey Potvin, Research Scientist (OU CIMMS)




Background:Ph.D. Meteorology, University of Oklahoma (2010)
M.S. Meteorology, University of Oklahoma (2006)
B.S. Meteorology, Lyndon State College (2004)
B.A. Mathematics, Lyndon State College (2004)
A.S. Computer Science, Lyndon State College (2004)
Experience:Corey grew up in Lewiston, Maine, and earned his bachelor’s degrees at Lyndon State College in Lyndonville, Vermont. After many cold winters, he came to Oklahoma to earn his Master’s and Ph.D. He was a Postdoctoral Research Associate with NSSL from 2010-2012, then was hired as a Research Scientist with OU CIMMS. He has also been an Adjunct Assistant Professor at the University of Oklahoma since 2014.
What He Does:Corey’s research focuses heavily on thunderstorms. He studies technique development for storm analysis, data assimilation, and processes/predictability. He is also part of the Warn-on-Forecast team at NSSL. This group examines methods for improving short-term guidance on storm hazards (you can learn more about Warn-on-Forecast here). What Corey most enjoys about working at NSSL is the fascinating, important work and the opportunities for collaborations with so many talented people. He also likes that he is able to advise and mentor students in his role. He also likes networking at conferences and connecting with a wider field of meteorologists.
Trivia: Corey has been married for eleven years and enjoys the beach, discussing religion/philosophy, and playing video games - especially World of Warcraft!
Share this:

NSSL to host 5th Warn-on-Forecast Workshop

WoF logoThe NOAA National Severe Storms Laboratory will host the fifth annual Warn-on-Forecast Workshop April 1-3, 2014 at the National Weather Center in Norman, Okla.  NSSL’s Warn-on-Forecast research project aims to increase accuracy and lead times for warnings of storm-specific hazards through high-resolution weather prediction models.

The three-day event gives researchers an opportunity to share progress reports on a variety of operational and experimental models, techniques, and decision-making tools in support of the Warn-on-Forecast project.

Researchers will share results from models that attempt to use satellite, lightning, targeted observations, and radar data, including phased array radar data to predict individual thunderstorms. They will report on how these data impact the model by using case studies of past events, and show comparisons with what actually happened. The group will also address the challenge of how to predict the birth of a storm, and share results using various new techniques.

Warn-on-Forecast collaborators include NOAA National Severe Storms Laboratory and Earth System Research Laboratory’s Global Systems Division, NOAA National Weather Service and Storm Prediction Center, and The University of Oklahoma’s Center for the Analysis and Prediction of Storms.

Share this:

2012 Warn On Forecast Workshop

The NOAA National Severe Storms Laboratory hosted the third annual Warn-on-Forecast Workshop February 8-9 at the National Weather Center in Norman, Okla.  Warn-on-Forecast is a National Oceanic and Atmospheric Administration research program tasked to increase tornado, severe thunderstorm, and flash flood warning lead times.

The Warn-on-Forecast workshop gives researchers an opportunity to present progress reports and to discuss plans for further research toward improvements in lead time for severe weather warnings.

Lead times are the time between a warning and when weather actually strikes. Trends in yearly-averaged tornado warning lead time suggest the present weather warning process, largely based upon a warn-on-detection approach using National Weather Service Doppler radars, is reaching a plateau and further increases in lead time will be difficult to obtain. A new approach is needed. Warn-on-Forecast is a convective-scale probabilistic hazardous weather forecast system. Guidance is provided by an ensemble of forecasts from numerical weather prediction models. Further research is needed to develop this system.

Warn-on-Forecast collaborators include NOAA National Severe Storms Laboratory and Earth System Research Laboratory, NOAA National Weather Service and Storm Prediction Center, The University of Oklahoma’s Center for the Analysis and Prediction of Storms, and Social Science Woven Into Meteorology.

Share this:

First annual Warn-on-Forecast workshop

The first annual workshop for the Warn-on-Forecast project was held on 23 February 2011 in Norman, Oklahoma, on the University of Oklahoma campus. Warn-on-Forecast is a NOAA research project to create forecasts of severe weather so specific, forecasters will be able to issue a warning based on that forecast before the weather even forms.

The workshop brought together over 60 participants from across the United States to listen to progress reports from all the groups participating in the project.

Focus topics for discussion included a social science research action plan and the benefits of VORTEX2 research to the Warn-on-Forecast project.

These reports indicated that the project is moving forward with research that will lead to improvements in lead time for severe weather warnings.  The project also has the potential to benefit a number of different weather information user communities, including surface transportation, aviation, and renewable energy.

Share this: