Assessment of Climate Change Impact on the Streamflow in Mae Ngat Basin, Chiang Mai

Abstract

Background and Objectives: Climate change has significantly affected water resources worldwide, including Thailand. This is especially true for watershed areas, which are vital for agriculture and local water use. One such area is the Mae Ngat watershed in Chiang Mai Province, a tributary of the upper Ping River, where the Mae Ngat Somboon Chon Dam serves as the primary water reservoir. Studies over the past decade have shown that this region has experienced increasing variability in rainfall, in terms of frequency, intensity and timing. This has led in turn to flash floods during the rainy season and droughts during the dry season. Such climatic uncertainty has impacted streamflow, ecosystems, agriculture and the overall efficiency of water resource management in the basin. To address this uncertainty, the present study aimed to assess the impact of climate change on streamflow and inflow into the Mae Ngat reservoir using the Soil and Water Assessment Tool (SWAT) model. The analysis incorporated climate data from the Coupled Model Intercomparison Project Phase 6 (CMIP6), which were spatially downscaled to a daily scale using the NASA Earth Exchange Global Daily Downscaled Projections (NEX-GDDP-CMIP6). This dataset is highly suitable for watershed-scale hydrological impact assessments and supports long-term water management planning under changing climatic conditions.

Methodology: The SWAT model was developed using input data, including topography, land use, soil characteristics and local meteorological records within the Mae Ngat watershed. The basin was subdivided into multiple sub-basins to enhance the spatial accuracy of hydrological process simulation. Model calibration and validation were performed using observed streamflow data from station P.56a during the period 2003–2014. Model performance was evaluated using statistical indicators, yielding R² of 0.80, NSE of 0.80, and PBIAS of -1.31%, indicating satisfactory agreement between the simulated and observed data and the model suitability for future projections. Future climate projections were obtained from 10 Global Climate Models (GCMs) under CMIP6. Four greenhouse gas emission scenarios were applied: SSP1-2.6, SSP2-4.5, SSP3-7.0 and SSP5-8.5. Daily downscaled data from NEX-GDDP-CMIP6 were used. Simulations covered two future periods: 2031–2050 and 2051–2070, with the historical baseline of 2003–2014. Streamflow outputs were analyzed on annual, monthly and spatial scales to assess changes and trends in water availability under different climate change scenarios.

Main Results: The projection results indicate that annual streamflow in the Mae Ngat watershed is expected to increase under all future climate scenarios. Notably, under SSP1-2.6 and SSP2-4.5, streamflow is projected to rise by approximately 54.07%–88.14% and 53.64%–76.32%, respectively. For SSP3-7.0, the projected increases are 35.79% and 70.91%, while SSP5-8.5 shows increases of 60.30% and 78.34% during the two future periods. Spatial analysis reveals that the central and lower parts of the watershed are mostly affected, with significant increases in streamflow that heighten the risk of flooding during the rainy season. Meanwhile, drought conditions remain a concern during the dry season due to the uneven distribution of rainfall throughout the year. Monthly-scale analysis shows that both SSP1-2.6 and SSP2-4.5 result in more evenly distributed increases in streamflow throughout the year, potentially reducing hydrological extremes. In contrast, SSP3-7.0 maintains dry-season streamflow levels close to the historical baseline, implying continued drought vulnerability. Additionally, SSP5-8.5 exhibits significant increases in streamflow during August and September, raising the likelihood of flash flood events. These contrasting seasonal patterns underscore the necessity of developing flexible and adaptive water resource management strategies to cope with both flood and drought risks under changing climate conditions.

Conclusions: Based on the SWAT model simulation using downscaled CMIP6 climate projections, streamflow in the Mae Ngat watershed is projected to increase during the period 2031–2070 under all greenhouse gas emission scenarios. Notably, SSP1-2.6 and SSP2-4.5 show a more consistent and evenly distributed increase in the streamflow throughout the year, while SSP3-7.0 and SSP5-8.5 exhibit higher variability, with increased flows during the rainy season and decreased flows during the dry season. These findings highlight a dual-risk scenario, i.e., flooding and drought, that may compromise the stability of the watershed's hydrological system. The calibrated SWAT model proves to be a reliable and effective tool for long-term water resource planning and management under future climate change conditions.

Practical Application: The results of the present study can be used as a policy-support tool for integrated and adaptive water resource management in the Mae Ngat watershed. Recommended strategies include enhancing reservoir capacity through structural improvements or supplementary storage facilities to handle peak rainfall; proper designs of effective drainage systems to mitigate flash flood risks in downstream areas are also recommended. Seasonal water allocation planning and promotion of water-efficient agriculture during the dry season are also essential. The integration of weather monitoring systems, predictive models and information technology can support strategic decision-making in such terms as crop scheduling, smart irrigation systems and early warning mechanisms. The presented framework and findings can also be applied to other watersheds in Thailand with similar hydrological and climatic characteristics, contributing to national resilience against future climate variability.