NSSL research meteorologist Adam Clark will present his work on improving severe weather forecasts during NOAA Science Days in June.
Clark, a Presidential Early Career Award for Scientists and Engineers 2014 winner, is presenting on the 2016 Spring Forecasting Experiment, which included a new framework for evaluating convection-allowing models known as the Community Leveraged Unified Ensemble.
Improved technology has allowed faster and more detailed experimental weather forecast models to be used in the NOAA Hazardous Weather Testbed, leading to model improvements and increased collaboration with other academic and government research agencies. One by product of the increased collaboration was that an increasing number of convection-allowing models, or CAMs, were ingested into the HWT. This was good, but also created a problem — the models designed by different agencies could not be compared because of too many independent variables.
“In 2016, we coordinated across all the different agencies that contributed to this experiment and we decided we’re going to have everyone abide by a set of rules— a bunch of criteria for how everyone will run their modeling systems,” Clark said. The rules included using the same area to run their models, the same resolution, and the same amount of detail to depict storms.
“That way we were able to control as many variables as we could so we could say more about why the different systems worked they way they did,” Clark said. CLUE collaborators then designed experiments testing different aspects of the model configurations.
Collaboration with forecasters is key to the CLUE experiments.
“What drives what we do is being able to work with the forecasters and get their take,” he said. “To design a system that is useful, you have to get feedback from the end user, which is forecasters.”
Last year was the first for CLUE to be used in the Spring Forecasting Experiment. With a goal to make forecasts better, Clark said there would be another this spring, with hopes of building on previous years.