9:00 AM - 10:30 AM
[MTT37-P03] An attempt of super multi-point intensive observation of micro-pressure fluctuation
Keywords:micro-pressure fluctuation, surface boundary layer, tsunami early warning
1. Background of the study
Lamb waves are atmospheric waves that are trapped on the ground and propagate horizontally at the speed of sound, and are excited by tsunamis and volcanic eruptions. Since Lamb waves arrive before tsunamis (typically about 200 m/s), detecting Lamb waves by a micro-pressure measurement network could be useful for early warning of tsunamis. Imada and Nakajima (2022) developed an inexpensive observation system that can be installed in large numbers. The measurement interval is 35 ms and the measurement accuracy is 0.5 Pa, and the system may detect thunders, atmospheric turbulence, etc. Here we present a preliminary report on an observation conducted distributing a large number of the micro-barometers in a narrow area. In addition to the analysis of the sensor noise and the demonstration of data acquisition, we discuss the observed atmospheric disturbance in the surface boundary layer.
2. Created micro-pressure measurement system
We made 100 micro-barometers that consist of a capacitive MEMS pressure sensor, DPS 310, and a microcomputer, M5 Stack ATOM LITE, and transmit data via a Wi-Fi-connected UDP. The data is stored on a small Debian GNU/Linux server wired to a Wi-Fi access point, which serves as an NTP server to provide the time reference of the barometers.
3. Field observation
On the night of January 12, 2023, we performed a field observation at a flat square near the Science and Technology Library of Kyushu University. 100 barometers were placed at 10-times-10 grid points at intervals of about 0.3 m, and a Wi-Fi access point was placed about 5 m away from this area, enabling observation for about 1 hour while receiving almost all the data (1714 to 5 per minute). The relative time error between barometers is estimated to be within 35 ms.
4. Preliminary results
At present, we are analyzing the data for 9 minutes during which the data acquisition was particularly good. Summary of the preliminary results is as follows:
4.1 Bias compensation
We corrected the bias of the sensors using the average of data for 100 seconds prior to the analysis period, assuming that the pressure is homogeneous averaged for a reasonable length of period. However, as described below, the averaging period have proven to be insufficient.
4.2 Noise of sensor
Spectral analysis was performed on the pressure time series of each barometer and compared with ensemble average of the power spectra from the 100 sensors. The component with a period shorter than about 3 seconds varies widely among barometers, and the component is thought to be sensor noise.
4.3 Atmospheric disturbance
First, we examined an animated map of 1-second moving average data of the deviations from the 100-point average. There, highly persistent deviations are noticed at some points, suggesting the incompleteness of the bias correction. Therefore, we decided to look at fluctuations of shorter temporal scales, so that the deviations of the 3-second moving average from the 10-second moving average were plotted at each time and animated. There we found plane-wave-like moving pressure fluctuations whose wavelength seems to be several meters.
5. Future direction
Further analysis including data from time intervals other than those analyzed here will be carried out, and the results will be presented on the day.
Reference
E. Imada, and K. Nakajima (2022): Development of a micro-pressure observation system for the observation of atmospheric lamb waves excited with the motion of the ground, Japan Geoscience Union Annual Meeting 2022.
Acknowledgements
This research was supported by a grant from JSPS Scientific Research Grants JP22K18872.
Lamb waves are atmospheric waves that are trapped on the ground and propagate horizontally at the speed of sound, and are excited by tsunamis and volcanic eruptions. Since Lamb waves arrive before tsunamis (typically about 200 m/s), detecting Lamb waves by a micro-pressure measurement network could be useful for early warning of tsunamis. Imada and Nakajima (2022) developed an inexpensive observation system that can be installed in large numbers. The measurement interval is 35 ms and the measurement accuracy is 0.5 Pa, and the system may detect thunders, atmospheric turbulence, etc. Here we present a preliminary report on an observation conducted distributing a large number of the micro-barometers in a narrow area. In addition to the analysis of the sensor noise and the demonstration of data acquisition, we discuss the observed atmospheric disturbance in the surface boundary layer.
2. Created micro-pressure measurement system
We made 100 micro-barometers that consist of a capacitive MEMS pressure sensor, DPS 310, and a microcomputer, M5 Stack ATOM LITE, and transmit data via a Wi-Fi-connected UDP. The data is stored on a small Debian GNU/Linux server wired to a Wi-Fi access point, which serves as an NTP server to provide the time reference of the barometers.
3. Field observation
On the night of January 12, 2023, we performed a field observation at a flat square near the Science and Technology Library of Kyushu University. 100 barometers were placed at 10-times-10 grid points at intervals of about 0.3 m, and a Wi-Fi access point was placed about 5 m away from this area, enabling observation for about 1 hour while receiving almost all the data (1714 to 5 per minute). The relative time error between barometers is estimated to be within 35 ms.
4. Preliminary results
At present, we are analyzing the data for 9 minutes during which the data acquisition was particularly good. Summary of the preliminary results is as follows:
4.1 Bias compensation
We corrected the bias of the sensors using the average of data for 100 seconds prior to the analysis period, assuming that the pressure is homogeneous averaged for a reasonable length of period. However, as described below, the averaging period have proven to be insufficient.
4.2 Noise of sensor
Spectral analysis was performed on the pressure time series of each barometer and compared with ensemble average of the power spectra from the 100 sensors. The component with a period shorter than about 3 seconds varies widely among barometers, and the component is thought to be sensor noise.
4.3 Atmospheric disturbance
First, we examined an animated map of 1-second moving average data of the deviations from the 100-point average. There, highly persistent deviations are noticed at some points, suggesting the incompleteness of the bias correction. Therefore, we decided to look at fluctuations of shorter temporal scales, so that the deviations of the 3-second moving average from the 10-second moving average were plotted at each time and animated. There we found plane-wave-like moving pressure fluctuations whose wavelength seems to be several meters.
5. Future direction
Further analysis including data from time intervals other than those analyzed here will be carried out, and the results will be presented on the day.
Reference
E. Imada, and K. Nakajima (2022): Development of a micro-pressure observation system for the observation of atmospheric lamb waves excited with the motion of the ground, Japan Geoscience Union Annual Meeting 2022.
Acknowledgements
This research was supported by a grant from JSPS Scientific Research Grants JP22K18872.