| Author | Cao Tien Hiep |
| Call Number | AIT Thesis no.WM-26-02 |
| Subject(s) | Climatic changes--Mathematical models--Evaluation
Hydrological forecasting--Thailand--Evaluation
|
| Note | A thesis submitted in partial fulfillment of the requirements for the degree of Master of Engineering in Water Engineering and Management |
| Publisher | Asian Institute of Technology |
| Abstract | This study proposes a data-driven framework to project future hydrological components and drought risks across Thailand's five distinct climatic regions. To reduce the uncertainty, 35 downscaled General Circulation Models (GCMs) from the NEX-GDDP-CMIP6 were evaluated to identify the most accurate models tailored to each specific region. These optimal regional models were then used as direct inputs into a Convolutional Long Short-Term Memory (ConvLSTM) deep learning model to predict future evapotranspiration, surface soil moisture, and water runoff. These high-resolution hydrological projections were subsequently used to quantify future droughts characteristics under different climate scenarios through standardize indices using a non-parametric approach.The findings provided the rank of 35 GCMs for each region, establishing the most reliable meteorological inputs for local assessments. The ConvLSTM framework successfully linked climate variables to hydrological responses and provided robust predictive performance, accurately reproducing the country's key water patterns, particularly for soil moisture. Furthermore, the analysis had shown that meteorological, agricultural, and hydrological droughts do not always behave in the same way, making a multi-index approach highly justifiable to avoid an incomplete picture of regional risk. Consequently, the multi-index projections provided future droughts frequency, duration, severity, and change across the country. These projections indicate that future drought trends will contrast sharply between regions: the Central and Eastern regions are projected to face significantly longer and more severe droughts, while the North, Northeast, and South are expected to see changing severities with shorter drought durations in the near-future and mid-future periods.This research contributes a comprehensive evaluation of NEX data across distinct sub-regions, significantly improving the reliability of future regional climate information. Furthermore, it introduces a highly accurate, data-driven methodology that generates continuous and reliable hydrological data. Ultimately, by providing detailed projections of future drought hazards, this study equips policymakers with the critical insights needed to develop robust adaptation strategies and enhance regional preparedness. |
| Year | 2026 |
| Type | Thesis |
| School | Faculty of Civil and Environmental Engineering (2026) |
| Department | Other Field of Studies (No Department) |
| Academic Program/FoS | Water Engineering and Management (WEM) |
| Chairperson(s) | Natthachet Tangdamrongsub |
| Examination Committee(s) | Shrestha, Sangam;Shanmugam, Mohana Sundaram;Khadka, Dibesh |
| Scholarship Donor(s) | Loom Nam Khong Pijai (Greater Mekong Subregion) Scholarships |
| Degree | Thesis (M. Eng.) - Asian Institute of Technology, 2026 |
