Comparison of Thunderstorm Monitoring by Satellite radiometer, Ground-based Meteorological and Lightning Detection System observations over the European Territory of Russia

A thunderstorm is a very dangerous phenomenon as lightnings can cause fires, damage electric power utilities while accompanied with heavy showers they lead to floods or mudflows, destroy roads, etc. Thunderstorms are recorded at ground-based meteorological stations and in their surroundings with a regularity of 3 hours and have a time lag of several hours; these observations are also subjective as the observer at the station can hear the thunder at the distance up to 20-30 km and miss a single lightning overhead, and there is also a problem of placing the observation into codes of present,  last hour or past weather. Lightning remote sensors continuously detect lightning events in the cloud or between clouds and earth at the distance up to 750 km (probability of detection is up to 80 %), however at the edge of the field of view the real location of the lightning can move up to 10-20 km while the probability of detection decreases up to 20-30 %. Satellite radiometer observations cover great territory, have high regularity (up to 15 min) and a time lag of several minutes, though they determine thunderstorm clouds by indirect signs, and at high satellite viewing angles they can miss small in size and intensity thunder clouds or relocate big thunderstorm clouds (up-to 20-40 km).

In the SRC “Planeta” there was developed a threshold technology that finds thunderstorm clouds while interpreting multispectral IR satellite measurements of brightness temperature  by indirect features and selects 4 classes according to probably and intensity of thunderstorms. In February, 2021 the Arktika-M No. 1 spacecraft was launched into a highly elliptical Molniya-type orbit with multispectral scanning geostationary device MSU-GS-VE as key payload. In February, 2024 the Arktika-M No. 2 was launched. The technology was adopted to the MSU-GS-VE channels 4-10. 

Within the European territory of Russia there are 14 lightning remote sensors (that are a part of the Russian Lightning Detection System (LDS)) and several dozens of meteorological stations. Three kinds of thunderstorm monitoring were compared during May-September of 2021-2024.

On the whole, weather observations at meteostations are the most inconvenient for thunderstorm monitoring because of their subjectivity, discreteness, a large time lag and inability to determine the exact time of a thunderstorm (the difference between the synoptic hour and the thunderstorm event can reach up to 3 hours). As a result, both satellite and LDS data in comparison with ground-based at meteostations have the lowest precision. When comparing satellite and at meteostations data with LSD the accuracy is much better. Satellite classification overestimates the size and the number of weak thunderstorm clouds as it shows clouds that can regenerate into thunder clouds or, on the contrary, are former thunder clouds, while the LSD data nondescript fully the synoptic situation.

So, for the most complete thunderstorm monitoring one should combine various kinds of observations so as they supplement each other. Because of large discreteness and a time lag from the synoptic hour observations at meteorological stations can be used only to analyze recent weather events. For the operative thunderstorm monitoring it is optimal to place LDS records over the satellite classification, so that the user can see the synoptic situation on the screen only about 20 minutes later as the stated time with periodic updates of the information every 15 min.