Thitimar Chongtaku, a PhD candidate from the Remote Sensing and Geographic Information Systems program under the guidance of her advisor, Prof. Attaphongse Taparugssanagorn, successfully conducted her research in identifying the risk identification due to the heatwaves and published her Research paper in the applied science journal. The research unveils new insights into the growing threat of heatwaves in Thailand, offering a groundbreaking approach that combines remote sensing and machine learning to analyze heat wave patterns over nearly four decades. The research, titled “Integrating Remote Sensing and Ground-Based Data for Enhanced Spatiotemporal Analysis of Heatwaves: A Machine Learning Approach,” addresses a critical knowledge gap and provides valuable tools for managing the impacts of extreme heat.
Figure 1: Thitimar Chongtaku with Dr. Attaphongse Taparugssanagorn
A New Approach to Understanding Heatwaves
The study focuses on heatwave dynamics from 1981 to 2019, utilizing a blend of satellite-derived land surface temperature (LST) data and ground-based air temperature observations. By integrating these two data sources, the research provides a richer, more nuanced picture of heatwave patterns over time and across different regions in Thailand. The aim is to elucidate the complex temperature dynamics in such regions, encompassing urban (Bangkok), suburban (Pathum Thani), and rural (Saraburi) locations. The study focuses on urban, peri-urban, and rural regions within the central region of Thailand. These classifications adhere to national standards and are based on criteria such as administrative divisions, demographic profiles, primary land use, and prevalent occupations. Therefore, the selected study areas are systematically categorized into urban, peri-urban, and rural types according to these variables. Additionally, these areas are situated within a tropical monsoon climate zone, characterized by three distinct seasons: summer (mid-February to mid-April), rainy (mid-April to mid-October), and winter (mid-October to mid-February). The use of remote sensing allows for the collection of large-scale, high-resolution temperature data, while ground-based measurements offer localized precision, making the combined approach highly effective in capturing the complexity of heatwave events.
At the core of the research is the application of advanced machine learning algorithms, which are used to analyze the vast dataset and identify patterns in heatwave frequency, duration, and intensity. Techniques such as regression models, clustering, and time-series analysis enable the study to pinpoint trends that would be difficult to detect through traditional methods. This machine-learning approach allows researchers to filter out noise from the data and focus on critical factors influencing heat wave behavior, such as changes in land use, urbanization, and seasonal variations.
Figure 2: Study-focused areas in Thailand
Revealing the Hidden Threat of Heatwaves
Thitimar’s analysis has uncovered a worrying trend: heatwaves in Thailand are becoming more frequent and intense, particularly in peri-urban and rural regions where the influence of microclimates is most pronounced. These areas, often caught between urban expansion and rural landscapes, are experiencing shifts in temperature patterns that pose a significant risk to both public health and local ecosystems.
The study’s findings highlight how microclimatic variations—localized weather patterns influenced by specific geographical features—can have a substantial impact on heatwave dynamics. For example, areas with dense vegetation or proximity to water bodies may experience less severe heat waves compared to those with extensive concrete structures, which can trap heat. By analyzing these variations, Thitimar’s research underscores the importance of considering local context when developing strategies to address climate-related challenges.
Figure 3: Result of distribution of pixel-wise correlation of air temperature and LST
A Technological Leap in Climate Research
This research represents a significant advancement in the application of remote sensing and machine learning in environmental sciences. Satellite data from sources such as MODIS (Moderate Resolution Imaging Spectroradiometer) and Landsat are used to gather information on land surface temperatures, while ground-based data is sourced from weather stations across Thailand. The integration of these datasets is a complex process that requires sophisticated data fusion techniques, which are facilitated by machine learning models that can handle vast amounts of information with accuracy and speed.
The machine learning algorithms used in the study are particularly adept at detecting subtle trends over long time periods and across diverse terrains. For instance, Thitimar’s approach can distinguish between natural seasonal variations and long-term changes due to global warming. This capability makes the study a powerful tool for monitoring ongoing climate changes and predicting future heatwave trends with greater precision.
Transforming Data into Actionable Insights
The real impact of this research lies in its ability to translate complex data into actionable insights for policymakers and urban planners. By identifying areas most vulnerable to heatwaves, the study serves as a critical resource for guiding climate adaptation strategies and informing public health interventions. The insights derived from the research can help authorities implement heat mitigation measures, such as increasing green spaces in urban areas or establishing early warning systems for vulnerable communities.
Moreover, the research has broader implications beyond Thailand. The methodologies developed in this study can be applied to other regions experiencing similar climate challenges, providing a blueprint for integrating remote sensing and machine learning in the analysis of extreme weather events. This makes Thitimar’s work not only a vital contribution to Thailand’s climate resilience efforts but also a model for other countries aiming to better understand and manage heat waves.
Figure 4: Research Flowchart
The innovative approach presented in this research can be applied across diverse environments to assess and manage heat waves effectively. Policymakers, stakeholders, and urban planners can utilize the findings to develop adaptive strategies, improve public health outcomes, and enhance resilience to climate change impacts. Additionally, the research contributes to advancing the field of remote sensing applications in environmental sciences, offering valuable insights for sustainable development initiatives worldwide.
Figure 5: Thitimar collecting Meteorological data from the Thai Meteorological Department
Aligning with Global Goals for Sustainable Development
Thitimar’s study directly supports the United Nations Sustainable Development Goals (SDGs), especially Goal 13: Climate Action, by providing critical data for mitigating the effects of extreme weather events. It also contributes to Goal 11: Sustainable Cities and Communities, focusing on the unique challenges faced by urban and peri-urban populations. The study’s emphasis on microclimatic variations aligns with the need for tailored, location-specific strategies in addressing the impacts of climate change.
As the global community grapples with the increasing frequency of heatwaves and other extreme weather events, research like Thitimar’s offers a path forward. By harnessing the power of remote sensing and machine learning, AIT’s work equips communities with the tools they need to adapt to the challenges of a warming world.
Figure 6: Retrieving Satellite data
In summary, this study addresses a critical knowledge gap concerning the challenges posed by heatwaves in Thailand, particularly emphasizing the need for comparative analyses across urban, peri-urban, and rural settings. By integrating geospatial analysis, remote sensing, and terrestrial data, this innovative methodology maps heatwave patterns across three socio-economic regions: Bangkok (urban), Pathum Thani (peri-urban), and Saraburi (rural). This approach gives a clearer view of how heat waves affect each region differently, which is essential for better urban planning and public health responses. The study’s unique method brings together various types of data to make heat wave predictions more accurate. Overall, it confirms past findings and adds new insights into how cities and surrounding areas can better adapt to extreme heat.
Read the complete research in the link below:
