GovLoop

How Big Data Helps Predict the Weather – Plus Your Weekend Reads

How many times a day do you check the weather? Whether it is on your phone, computer, radio or TV – the need for weather data is nearly constant. In the past few years the scientists at NOAA have pioneered numerous hurricane and wind speed prediction models that have become the bedrock for the nation’s weather forecasters.

The results from the models allow us to make critical decisions about evacuations and other preparations for severe weather events, helping protect billions of dollars in property and save thousands of lives. The man behind the models are Mark DeMaria.

He is a Supervisory Meteorologist at NOAA’s Center for Satellite Applications and Research in Fort Collins, Colorado. His work has made him a finalist for the Service to America Medals – the Oscars for federal employees.

DeMaria told Chris Dorobek on the DorobekINSIDER program that while there is still a lot of room to grow, forecasting models have come along way in the last 10 years.

“The ability to track where a storm is going to go have improved tremendously over the past 10-15 years. The accuracy is twice what it was 15 years ago. The skill of the 4-day weather forecast is what the 2-day cast was a few years ago,” said DeMaria. “The new models give you more lead time and allow forecasters to eliminate some of the false alarms.”

Why Are Storms So Hard to Predict?

“The atmosphere varies from very small things like wind gusts to larger scale things like hurricanes, to things that cover the whole plant like the jet stream. All of those things interact with each other and you need to model them at the same time to get an accurate weather forecast,” said DeMaria.

Why Is NOAA Necessary? We have the Weather Channel?

“There is a bit of a misunderstanding because most people’s interface into weather is the media. The average person dosen’t understand the tremendous amount of infrastructure that goes behind those weather forecasts. TV stations don’t own satellites, they don’t have the budget to maintain them. They don’t run their own computer models, that in itself is a gigantic enterprise that really only the federal government can and is willing to support.”

Big Data

“You need these tremendously large data sets to forecast. We use a grid of points that covers the entire globe, where individual stations are only a few miles apart. Plus there are a bunch of data sets from the ground all the way up to 50 kilometers in the atmosphere. How fast the wind is blowing, what the temperature is, the moisture level, we need that data from pretty much everywhere in the atmosphere over the whole earth. If you can’t measure those stats constantly then you are going to have errors in your models.”

Room to Grow

“There is still considerable room to improve. One of big limitations right now is the ability to forecast whether a storm is going to get stronger. Superstorm Sandy was pretty well behaved near the coast so that wasn’t an issue in that case. But there are other storms that can rapidly intensify or weaken and the ability to forecast that even a day or two in advance is pretty limited right now,” said DeMaria.

No Silver Bullet

“There isn’t a silver bullet that we will discover and then suddenly we will able to predict everything perfectly. It is a bunch of different processes interacting together. We need to be able to study the oceans better and measure the atmosphere better.”

Tornados Vs. Hurricanes

“When it come to predicting these storms it is all about the scale.”

You can find all our Sammies interviews here.

Weekend reads

Want More GovLoop Content? Sign Up For Email Updates

Exit mobile version