Desktop Storm Chasing: Using Digital Images To Verify Weather Conditions

When most people think of storm chasing, they think of passionate weather folks driving across the Southern Plains in search of a massive tornado. While this type of storm chasing can bring us some great information about how tornadoes form along with images of the damage they create, let’s focus on a different kind of storm chasing. This type of chasing involves using digital images to get a real-time view of the current weather in any given area.

Filling the gaps with digital imagery

Digital imagery is everywhere nowadays. One can log onto the internet from Alaska and view a live stream of a beach in Florida. My mom can live stream Old Faithful erupting in Yellowstone National Park, one of her favorite things to see. The aforementioned weather folks can live stream the storms and tornadoes they chase. Many state departments of transportation (DOTs) have installed cameras in their fleet of snowplows which can showhow a snow event is affecting the roadway innear real time. Traffic cameras also help to provide this information. All of these images help to fill in the gaps in the weather observation network across the country. This visual data can help determine the current weather situation, which in turn can lead to a better forecasts, especially in the short term.

Ground truth

Meteorologists are constantly trying to verify what is happening on the surface when a weather event moves over a given area. The main problem with this, is that there are a limited number of observation sites around the country (and world) to verify what (if anything) is happening at ground level. Many weather observations are taken at airports, but how many airports are in a given state or area. Most states have also set up a Road Weather Information System (RWIS)network, which gives surface weather conditions and can also give pavement conditions (if set up), but these are still limited in area coverage. There are other observation networks set up for agriculture, snowfall in the mountains, etc. but with some of these systems, the data is not immediately available. So meteorologists turn to camera images (streaming or still) to help determine how a given weather event is affecting the surface.

The surface observations are used in conjunction with radar and satellite to determine if precipitation is reaching the ground or possibly if the radar is picking up some false echo. Once a meteorologist determines what is happening on the surface, they are able to use that information to help decide what weather model may be handling the current situation more accurately.

A model for the future

Roadway and bridge images, especially from RWIS cameras and DOT snowplow cameras (if available), help us to verify if our pavement model (HiCAPS) is handling a given weather situation properly. Did the roadway become snow covered? Did it stay dry because the wind blew the snow off the road? Or did it become wet because the pavement temperature was warm enough? These are the questions forecasters are trying to answer by using these images. Roadway images were also used to help with verification of the HiCAPS model as it was being developed and refined. Images were gathered to see how the pavement condition changed during an event and after it was plowed/treated, and how these compared with road condition cross sections produced by the model, allowing for refinement of the model and the associated maintenance recommendations.

The Bottom Line

As webcams become more saturated across a given region in the future, the gaps in the observation network will shrink and we will have a better picture of current surface conditions. Pretty soon we won’t need to go chase down a tornado because there will already be a camera in its path.