Methane Detection Through Satellite Imaging: Analyzing and Comparing Methane Concentration in Turkmenistan and Finland
Ristioja, Fanni (2023-05-30)
Methane Detection Through Satellite Imaging: Analyzing and Comparing Methane Concentration in Turkmenistan and Finland
(30.05.2023)
Julkaisu on tekijänoikeussäännösten alainen. Teosta voi lukea ja tulostaa henkilökohtaista käyttöä varten. Käyttö kaupallisiin tarkoituksiin on kielletty.
avoin
Julkaisun pysyvä osoite on:
https://urn.fi/URN:NBN:fi-fe2023060652698
https://urn.fi/URN:NBN:fi-fe2023060652698
Tiivistelmä
Remote sensing by satellites is a growing field of study and has been applied to environmental research, e.g., in detecting greenhouse gases such as methane. Methane has recently been a major concern for industry due to its global warming potential, which is greater than that of carbon dioxide. This thesis studies methane enhancements in natural gas fields in Southern Turkmenistan using satellite data. These results are then compared to ones from Southern Finland, for which a similar analysis was done in order to see how differences in the areas of interest affect results.
The satellite that was used for this thesis was Sentinel 5 Precursor (S5P) and the instrument that detected methane is called Tropospheric Monitoring Instrument (TROPOMI). S5P dataset is open source and was thus chosen for the analysis. Several plots were done that visualized the data and some additional analysis was done for those plots. Two large methane enhancements were observed in Turkmenistan and they were saved into a GeoJSON file so that the data can be easily distributed and visualised. Some analysis was done using wind data from the time of observation since the two methane enhancements seemed to be spread in certain directions. The wind was not very strong and the wind direction differed from the direction of dispersion; therefore, it was not straightforward to conclude that this caused the observed spread.
There were much poorer quality data available for the area of interest in Finland than in Turkmenistan. The quality of the dataset is determined by the quality attribute (qa), which is usually set to 0.5; pixels with a quality attribute of 0.5 or higher are treated as reliable data points that are not contaminated by, e.g., clouds, and are taken into account. Any pixels with a quality attribute less than 0.5 were discarded. In order to see how much of an effect the quality attribute has, the same analysis was done again for both areas of interest with a quality attribute ≥ 0.25. This was particularly interesting for the area of interest in Finland, which had significantly less data than Turkmenistan when using a higher quality attribute. With a lowered qa the increase in the amount of data points in Finland was significant. This also revealed a new pixel that had higher methane concentration than the others in the area, although it was still a minor increase compared to the enhancements in Turkmenistan. This pixel was further examined, but it did not lead to significant results. However, the experiments showed that a more relaxed qa filtering could prove useful for higher latitude observations.
The satellite that was used for this thesis was Sentinel 5 Precursor (S5P) and the instrument that detected methane is called Tropospheric Monitoring Instrument (TROPOMI). S5P dataset is open source and was thus chosen for the analysis. Several plots were done that visualized the data and some additional analysis was done for those plots. Two large methane enhancements were observed in Turkmenistan and they were saved into a GeoJSON file so that the data can be easily distributed and visualised. Some analysis was done using wind data from the time of observation since the two methane enhancements seemed to be spread in certain directions. The wind was not very strong and the wind direction differed from the direction of dispersion; therefore, it was not straightforward to conclude that this caused the observed spread.
There were much poorer quality data available for the area of interest in Finland than in Turkmenistan. The quality of the dataset is determined by the quality attribute (qa), which is usually set to 0.5; pixels with a quality attribute of 0.5 or higher are treated as reliable data points that are not contaminated by, e.g., clouds, and are taken into account. Any pixels with a quality attribute less than 0.5 were discarded. In order to see how much of an effect the quality attribute has, the same analysis was done again for both areas of interest with a quality attribute ≥ 0.25. This was particularly interesting for the area of interest in Finland, which had significantly less data than Turkmenistan when using a higher quality attribute. With a lowered qa the increase in the amount of data points in Finland was significant. This also revealed a new pixel that had higher methane concentration than the others in the area, although it was still a minor increase compared to the enhancements in Turkmenistan. This pixel was further examined, but it did not lead to significant results. However, the experiments showed that a more relaxed qa filtering could prove useful for higher latitude observations.