Integration of satellite-derived aerosol optical depth and pm2.5 monitoring data for improving air quality monitoring network in Thailand | |
| Author | Wipaporn Saweangwit |
| Call Number | AIT Thesis no.EV-24-07 |
| Subject(s) | Air quality--Remote sensing--Thailand Air quality management--Technological innovations--Thailand Air--Pollution--Measurement--Technological innovations--Thailand |
| Note | A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science in Environmental Engineering and Management |
| Publisher | Asian Institute of Technology |
| Abstract | Fine particulate matter (PM2.5) is a major concern in several areas in Thailand. Particularly, the Chiang-Rai province is faced with health effects, vista, and tourism from extreme concentration of PM2.5 every year during the dry season. To tackle this matter, information of PM2.5 is crucial to detect sources, dispersion, and concentration to solve the problem in time. Real time data is measured by air quality monitoring stations that are available in only two locations in Chiang-Rai province. By doing so, it is limited to provide data for the whole province and inability to deal with situation in different locations. Thus, the existence of technology as satellite is the key to close this gap since satellite can cover broadly area, and their aerosol optical depth (AOD) product can be used to predict PM2.5 concentration in the area where has no ground-based monitoring. In this study, AOD from two different satellites, i.e., GEMS and Himawari, were applied to develop relationships between AOD and PM2.5 concentrations. The spatial resolution of GEMS and Himawari are 3.8x5 km2 and 5x5 km2 , respectively, with the same temporal resolution as hourly. The data in this study was from July 2022 to October 2023 covering high PM2.5 concentration in Chiang Rai province. AOD-PM2.5 relationship was performed with three model (Linear regression, random forest, XGBoost) to select the best model. The meteorological parameters were employed, and it increased the accuracy of model significantly. Furthermore, the models were improved with Grid-Seach method (CV = 10) to reduce root mean square error (RMSE) and mean absolute error (MAE). As a result, the best model was run on GEMS’s AOD with meteorological parameters (temperature, wind speed, relative humidity) by XGBoost model with R2 0.88, RMSE 41.35 µg/m3 , and MAE 36.74 µg/m3 . Also, RMSE and MAE decreased 7-14% after improvement. Eventually, the best models were applied and classified the relationship between hourly observed PM2.5 and hourly predicted PM2.5. This investigation found that the highest relation of the observed PM2.5 and predicted PM2.5 is on 12.00-13.00 hrs in both different satellites with R 2 of 0.97. The model has high bias when the observed PM2.5 is greater than 100 µg/m3. Spatial distribution of PM2.5 concentration from the two satellites was compared after applied models in Chiang-Rai. Limited number of observations of GEMS affected PM2.5 concentration prediction while Himawari achieved better results during dry season. |
| Year | 2024 |
| Type | Thesis |
| School | School of Environment, Resources, and Development |
| Department | Department of Energy and Climate Change (Former title: Department of Energy, Environment, and Climate Change (DEECC)) |
| Academic Program/FoS | Environmental Engineering and Management (EV) |
| Chairperson(s) | Ekbordin Winijkul |
| Examination Committee(s) | Cruz, Simon Guerrero;Natthachet Tangdamrongsub |
| Scholarship Donor(s) | Her Majesty the Queen's Scholarships (Thailand) |
| Degree | Thesis (M. Sc) - Asian Institute of Technology, 2024 |