A new algorithm from researchers in Austria offers a fix for weather radar images distorted by Wi-Fi, reports the MIT Technology Review.
Weather radar, also known as weather surveillance radar is a type of radar first developed shortly after World War II to detect precipitation, In the half-century that followed, weather radar has developed greatly and now contributes to agriculture alerts and flood warnings, atmospheric research and weather monitoring and prediction.
Weather radar is especially critical in aviation safety, where air traffic controllers must keep track of the movements of storms and other weather phenomenon, evaluate windshear and monitor ice formation in order to ensure aviation safety. However, interference caused by reflections, ground clutter, atmospheric turbulence and rocky or mountainous terrain can distort the radar image, preventing controllers from accurately gauging weather conditions.
Now, researchers at Joanneum Research in Graz, Austria have developed an image analysis algorithm that focuses on identifying and removing various kinds of interference from the radar image, pixel by pixel, while simultaneously filling in gaps caused by uneven terrain, which prevents weather-related echoes from reaching the receiver.
“Improvement of WXR [weather radar] images—by combining them with other sensors in order to complete concealed zones and eliminate interfering signals without losing important meteorological information—is essential for an accurate prediction of weather phenomenon and atmospheric conditions,” lead author Harald Ganster and his colleagues said in a new paper hosted on the Cornell University Library website.
The new approach relies on the unique characteristics of different kinds of interference, says Ganster, such as Wi-Fi interference, which appears in radar images as a straight line. The new algorithm maps these straight lines into vertical ones, which are then easily identified and removed based on the fact that there are no naturally occurring structures in weather radar images that are straight.
The algorithm also fills in the image gaps with meteorologically reasonable values, using data taken from space by the Meteosat Second Generation satellite. While the satellite images do not provide the same level of detail as their weather radar counterparts, they do show large variations in weather in the missing regions. The algorithm then minimizes the difference between the satellite and weather radar images as best as possible.
Austrian air traffic control officials are currently evaluating the new system for future use.
