UM Rosenstiel School Meteorology and Physical Oceanography graduate student Falko Judt was awarded first prize in the Outstanding Student Poster competition at the American Meteorology Society’s 94th Annual Meeting in Atlanta, Georgia. Falko shares the research behind his poster, titled “Tropical Cyclone Intensity Forecasting: Predictability and Uncertainty.”
“The main goal of my research is to better understand and therefore better predict the intensity of hurricanes. Computer model forecasts are getting better with forecasting where a hurricane will go (its track), but we’re still having a hard time telling in advance how strong a storm might become (intensity).
One school of thought is that it is very difficult to accurately forecast intensity for more than 2 days because small-scale and not well understood processes in the inner regions of a storm. In the inner core, energy from the ocean surface is converted to strong winds in thunderstorm clouds that surround the eye, forming the hurricane rain bands. Knowing the exact arrangements and processes going on in these clouds is, according to this hypothesis, very important for forecasting changes in a hurricane’s intensity.
On the other hand, some scientist argue that these small-scale processes going on within the hurricane don’t matter too much, and the really important factors determining the strength of a storm are properties of the large-scale environment surrounding it. For example, high ocean temperatures, high atmospheric humidity values and low wind shear are known to lead to intensification of hurricanes.
My research was testing these hypotheses by adding stochastic (random) perturbations to state-of-the-art computer models (ensemble forecasts) predicting the intensity of Hurricane Earl of 2010, which is a storm that already happened, so we can compare the forecasts to what was observed.
In one ensemble, tiny perturbations with the size of clouds were added to the model, and another ensemble had larger perturbations on the scale of real weather systems — about a few hundred to thousands of miles. We found that very small perturbations do not change the properties of the hurricane predictions, and even larger-scale perturbations only cause some fluctuations in their forecasted intensity.
We therefore concluded that ultimately the environmental conditions determine how strong a storm will be. This is good news because we have better knowledge about and data from a hurricane’s environment.
Based on this research presented in my poster, we might even be able to tell how strong a storm will become 7 day in advance with the help of improving computer models and better observations from the surrounding atmosphere. The predictability limit seems to be longer than what had been expected.”
– Falko Judt