Now there is one thing that is widely known among meteorologists, young and old-- the hardest season to forecast precipitation for is the Summer. Why, you ask? Well, the precipitation tends to form from small-scale thunderstorms. While there are preferential locations where these storms form such as over higher terrain or near converging river valleys, they still may depend on small-scale atmospheric variables that are harder to predict. Also, while a storm may be wreaking havoc in Nassau, the Hamptons are likely staying quite dry and sunny. Weather models have a lot of trouble predicting small-scale, thunderstorms. A whole type of model made specifically for looking at small-scale storms is called a mesoscale model. Recall that DREAMS stands for Doppler Radar for Education and Mesoscale Studies. There's that word again! When making the forecast for Wednesday, June 26, we began by looking at the larger-scale model output to get an idea of what the big picture was and if there would be any key players such as a front or low pressure system. Unfortunately, there was no larger forcing that could give us an idea of what time or how strong any storms would be. Therefore, we had to rely on some mesoscale models. It's important to not just look at one, because each model has something slightly different about it whether it is how it calculates the physics of raindrops or the motions of air right near the surface. There's never one model that always "wins" which is why it helps to remain as unbiased as possible and keep an open mind. Here's some information about some models that we looked at. The North American Model (NAM) has a 4-km horizontal resolution (that's roughly 2.5 miles) that is operationally run by the government. It showed some weak storms moving through the area in the late afternoon. The Coastal Meteorology and Atmospheric Prediction Group at the School of Marine and Atmospheric Sciences at Stony Brook University (SBU) actually runs their own mesoscale model called the Weather Research and Forecasting (WRF) model. You could run it, too, if you had a Linux machine with very good processors! The SBU WRF is run twice-daily and generates images to a webpage. The webmaster/graduate student Matthew Sienkiewicz will gladly take suggestions about new types of plots if you happen to have any (see his site in the comments at the end of the post). It also has 4-km horizontal resolution but because of the nature of the model setup, it provides a slightly different picture. The SBU WRF had some storms moving across Long Island during the evening hours.
Because there was uncertainty regarding the timing and location of any convection, the decision about where and when to go was put off until the morning, when fresh model output would be available and the atmosphere settled down from the night before. However, even in the morning there was a large amount of uncertainty. The decision was made to stay close to home and to set up the DOW at Cedar Beach in Mount Sinai with a view looking north over the Long Island Sound. The DOW left SBU at 3 PM and headed east.
|The pod in place at Cedar Beach.|
- For weather model data, please visit this site: http://mag.ncep.noaa.gov/ (or many others, like twisterdata.com)
- For the SBU WRF model plots, please visit this site: http://dendrite.somas.stonybrook.edu/LI_WRF/index.html