1:45 PM - 3:15 PM
[O11-P122] researching the Sprite (Ionosphere Lightning Discharge) Phenomenon
Keywords:sprite, Machine Learning, meteorology
Motive
In one year from 2023, an observation device that can capture images of the entire sky simultaneously using a security video camera, which was created and used by the Meteor Research Group of our school's
To identify the detailed conditions under which the sprites occur, we conducted a preliminary investigation of the time and direction of occurrence of the sprites in the video data to date and analyzed the weather conditions at the locations where the sprites occurred.
Backgrounds
According to a preceding study (Reference 2), sprites are generated by upward positive polarity lightning strikes, and each type of sprite is known to have different characteristics such as color and shape (Figures 1-3),but the details have not yet been clarified. In this study, we focused on the meteorological conditions at the time of occurrence of sprites and investigated the detailed occurrence conditions.
Research method
The meteor team has been capturing omnidirectional video data since 2023 using an instrument capable of capturing the entire sky (Figure 4), which was produced using five ATOM Cam 2security cameras.
Reference(s).
In this study, we attempted to analyze sprites using this data in the following three steps. Manually check the captured images and cut out the frames in which sprites are observed. Analyze the direction and time of day in which the sprites appear and identify the ground lightning strike that may have caused the sprites in the video. Analyze the correlation between weather conditions and sprite occurrence by tabling the shape of the sprite, presence of precipitation, and atmospheric pressure pattern,etc.
Result
The meteor team used an all-sky imaging device (Figure 4) made with five ATOM Cam 2 security cameras, and (Figure 5) shows 8 sprites confirmed to have occurred in the past year with clear photographs and the results of their analysis.
The following results were obtained from this survey. We observed several sprites that were difficult to observe. We observed a series of sprites, which is unprecedented in the past.
Consideration
The results obtained in this study were discussed based on the contents of a previous study, “Sprite Observation Handbook 2005” (Reference 4), focusing on the following points. The discharges that cause sprites are presumed to be so-called “field lightning. The fact that these events occur frequently in the Sea of Japan during the winter season. There were several cases, such as on February 15, 2024, where several occurrences could be confirmed in a short period of time. Based on these facts, we hypothesized that lightning strikes from relatively short cumulonimbus clouds, which bring snow to the Sea of Japan in winter, are likely to cause sprites.
In addition, there are few observed cases of continuous occurrence of sprites as in (3). Since sprites are a phenomenon occurring in the ionosphere, we believe that there may be ionospheric conditions that tend to cause sprites.
Prospects
All checking of recorded data was done manually. Currently, using Python programming language and OpenCV, a library mainly used for image editing, we have created a program that automatically detects sprites in the observed images through machine learning using approximately 100 images of sprites published on the SonotaCo Network and the school as training data. The program is currently in the trial run phase. This is expected to reduce labor and improve detection accuracy.
In one year from 2023, an observation device that can capture images of the entire sky simultaneously using a security video camera, which was created and used by the Meteor Research Group of our school's
To identify the detailed conditions under which the sprites occur, we conducted a preliminary investigation of the time and direction of occurrence of the sprites in the video data to date and analyzed the weather conditions at the locations where the sprites occurred.
Backgrounds
According to a preceding study (Reference 2), sprites are generated by upward positive polarity lightning strikes, and each type of sprite is known to have different characteristics such as color and shape (Figures 1-3),but the details have not yet been clarified. In this study, we focused on the meteorological conditions at the time of occurrence of sprites and investigated the detailed occurrence conditions.
Research method
The meteor team has been capturing omnidirectional video data since 2023 using an instrument capable of capturing the entire sky (Figure 4), which was produced using five ATOM Cam 2security cameras.
Reference(s).
In this study, we attempted to analyze sprites using this data in the following three steps. Manually check the captured images and cut out the frames in which sprites are observed. Analyze the direction and time of day in which the sprites appear and identify the ground lightning strike that may have caused the sprites in the video. Analyze the correlation between weather conditions and sprite occurrence by tabling the shape of the sprite, presence of precipitation, and atmospheric pressure pattern,etc.
Result
The meteor team used an all-sky imaging device (Figure 4) made with five ATOM Cam 2 security cameras, and (Figure 5) shows 8 sprites confirmed to have occurred in the past year with clear photographs and the results of their analysis.
The following results were obtained from this survey. We observed several sprites that were difficult to observe. We observed a series of sprites, which is unprecedented in the past.
Consideration
The results obtained in this study were discussed based on the contents of a previous study, “Sprite Observation Handbook 2005” (Reference 4), focusing on the following points. The discharges that cause sprites are presumed to be so-called “field lightning. The fact that these events occur frequently in the Sea of Japan during the winter season. There were several cases, such as on February 15, 2024, where several occurrences could be confirmed in a short period of time. Based on these facts, we hypothesized that lightning strikes from relatively short cumulonimbus clouds, which bring snow to the Sea of Japan in winter, are likely to cause sprites.
In addition, there are few observed cases of continuous occurrence of sprites as in (3). Since sprites are a phenomenon occurring in the ionosphere, we believe that there may be ionospheric conditions that tend to cause sprites.
Prospects
All checking of recorded data was done manually. Currently, using Python programming language and OpenCV, a library mainly used for image editing, we have created a program that automatically detects sprites in the observed images through machine learning using approximately 100 images of sprites published on the SonotaCo Network and the school as training data. The program is currently in the trial run phase. This is expected to reduce labor and improve detection accuracy.