Hurricane Forecasting: What Happens Behind the Scenes

Whether or not you live in an area that could be impacted by tropical storms or hurricanes, you probably follow the forecasts. After all, when storms like Michael, FlorenceIrma, Maria and Harvey are tracking toward inhabited areas, the suspense is gripping and the media coverage unavoidable. These forecasts are particularly useful to disaster recovery companies, which pay close attention so that they can be prepared for what’s ahead and take the necessary measures to be able to assist their customers.

If you’re like most people, you probably also wonder about the accuracy of forecasts. Why does the track and intensity of the storm seem to change so often? And what types of technology do researchers and scientists use to make their predictions and forecasts?

As you might expect, today’s researchers and forecasters have a lot of highly specialized technology at their disposal. This post will give you a better understanding of some of the key technology and tools that are used in making their predictions and forecasts, along with factors that can impact accuracy.

Gathering Data for Near-term Storm Forecasts
During the hurricane season, forecasters rely on a host of space, air, sea and land-based tools to gather data and track storms once they have formed, including the following:

  • Satellites—weather satellites can include a variety of sensors and instruments that measure things like sea and land surface temperatures, near-surface winds, as well as moisture structure and precipitation patterns in cloud formations, which provide key insights about a storms intensity and track.1
  • Weather radar—weather surveillance radar and Doppler weather radar help forecasters identify and study areas of precipitation, providing them data about the intensity and type of precipitation in a given area.2
  • Reconnaissance aircraft—The National Oceanic and Atmospheric Administration (NOAA) has a fleet of turbo propeller planes that it can fly into developing storms. The aircraft include specialized microwave radiometers that measure surface wind and rainfall, and they also release specialized probes, called dropwindsondes, that transmit a variety of weather data back to the plane, including data about wind direction and speed as well as pressure.3
  • Buoys—The NOAA also relies on the National Data Buoy Center (NDBC), a network of buoys and automated network stations, to gather data on everything from temperatures, to wave characteristics, to barometric pressure and wind speed.4
  • Ships—the World Meteorological Organization’s Voluntary Observing Ship (VOS) program, relies on trained crews to provide timely weather observations during voyages and even deploy probes to measure water temperatures.5

By now you get the idea. A lot of different technologies are used to gather massive amounts of data about storm variables.

Crunching all of the Data
Gathering data is only the beginning of the forecasting process. Once researchers and scientists have the data they need on hand, the challenge is to quickly figure out what’s most likely to happen. And that’s where things get complicated.

To make predictions, researchers primarily rely on numerical forecast models that are run on supercomputers.6 In essence, the models run the data through mathematical equations that are designed to mimic physical laws and the behavior of the atmosphere.7 Of course, given the complexity of the atmosphere, even with all of the computing power available nowadays, there is still plenty of room for error. For example, the data about what’s happening in the atmosphere is never “complete” and scientists’ understanding of atmospheric physics and the models themselves are imperfect. In other words, there is still significant guesswork or tradeoffs that have to be made in modeling approaches.8 That’s part of the reason there is often still a sizable margin of forecasting error in the days and hours leading up to landfall, as we saw with hurricane Michael's intensity and tracking predictions.9

Steadily Improving Accuracy
If you look at the numbers over the years, however, hurricane forecasting has improved by leaps and bounds. For example, between 2000 and 2016, the average error in storm tracking forecasts was reduced from 150 miles to 70 miles.10

Moreover, as scientists continue to improve models and gain access to more detailed and timely data—as well as increased computer power—the margins of error will only get smaller. Whether or not there will ever come a time when forecasting takes the guesswork out of evacuation and disaster planning remains to be seen. In the meantime, it’s important to always take worst-case-scenario forecasts seriously and prepare your business accordingly. Having contact information for disaster recovery services is a great first step.

1 Hurricane Science: Satellites,, 2015.
2 Weather radar, Wikipedia.
3 Hurricane Hunters: Flying Into Storm’s Core, National Oceanic and Atmospheric Administration (NOAA), May 2017.
4 National Data Buoy Center, Wikipedia.
5 Ship-based Observations,, 2015.
6 NHC Track and Intensity Models, National Hurricane Center, February 2017.
7 Hurricane and Tropical Cyclones, Weather Underground, 2017.
8 Ibid.
9 Hurricane Michael and why it’s so hard to predict storm intensity, Vox, October, 2018.
10 What Goes Into Hurricane Forecasting? Satellites, Supercomputers And More, NPR, September 2017.