Climate change alters hydrological cycles and streamflow regimes at the local, regional and global levels. In this study, we aimed to assess the change in water balance change and hydrological extremes in the Prek Thnot River Basin of the Lower Mekong in Cambodia through a hydrological model (SWAT) under the two climate change scenarios (RCP2.6 and RCP8.5) following three different GCMs. An ensemble of 3 GCMs included GFDL-CM3, GISS-E2-R-CC and IPSL-CM5A-MR models and was applied to a well-calibrated SWAT model through climate change factors. Annual precipitation under RCP2.6 likely decreases by 0.1–0.5% for the near future (2021–2040) and mid-future (2051–2070) and decreases by 0.2–1.3% under RCP8.5. The decrease in precipitation will lead to reductions in water yield by 1–4% (RCP2.6) and 2–5% (RCP8.5). However, peak flow is expected to increase, while the low flow was projected to decrease (1–2% for RCP2.6 and 8–9% for RCP8.5). The study further found that high flow events will increase in both magnitude and frequency. The finding highlights water resources management issues in the Prek Thnot River Basin, including the frequency of future flood events.

  • We estimated the climate change impact on water balance and hydrological extremes.

  • Annual water balance in the Prek Thnot River Basin is projected to decrease in the future.

  • Future streamflow is expected to increase in the wet season while decreasing in the dry season.

  • Peak flows have an increasing trend while low flows have a decreasing trend in the future.

Graphical Abstract

Graphical Abstract
Graphical Abstract

Climate change is one of the major global threats that strongly affect the environment, ecosystem, and human society. In recent decades, the study on climate change impact on pervasive global warming driven by anthropogenic emissions of greenhouse gases (GHGs) has been given ample attention worldwide (IPCC 2014). In the past few decades, intensification of precipitation and increases in temperature due to global warming have been observed in Asia (Tan et al. 2021a, 2021b; Xu et al. 2021). These changes significantly impact different hydrological systems and consequently increase regional water hazards such as floods and drought (Chattopadhyay & Jha 2016; François et al. 2019). Moreover, climate change alters hydrological cycles and streamflow regimes at the regional and global levels (IPCC 2014; Martínez-Retureta et al. 2021), especially at the basin scale (Zhang et al. 2016). The predicted effects are precipitation patterns, increased intensity of extreme weather events, and glacier retreats (IPCC 2014). They consequently alter the river discharge regimes and directly affect ecosystems, water security. Hence, quantifying the climate change impact on water balance components and hydrological extremes is essential to watershed management as well as to the development of adaptation strategies, more effective mitigation and greater resilience against water hazards (Bhatta et al. 2019; Blanco-Gómez et al. 2019; Oeurng et al. 2019; Touch et al. 2020; Tan et al. 2021b; Yun et al. 2021).

A set of scenarios known as the Representative Concentration Pathways (RCPs), published in the Fifth Assessment Report (AR5) of the Intergovernmental Panel on Climate Change (IPCC) has recognized the significant indications of climate change: global temperature rises due to greenhouse emissions and water cycle changes (IPCC 2014). IPCC (2014) reported that global ocean and land temperatures have increased by 0.85 °C throughout 1880–2012. The global average temperature increase is projected to be between 0.3 °C and 4.8 °C, and the global mean sea level may rise by 0.26–0.82 m by the end of the 21st century (Stocker 2014). Across Southeast Asia, the mean annual temperature by the end of the 21st century is expected to rise from 0.8 °C in the lowest emissions scenario to 3.2 °C in the highest emissions scenario. At the same time, a moderate increase in precipitation is also projected for this region from 1% to 8% by 2100 (IPCC 2013). However, the change in global future precipitation is not uniform (both an increasing and/or decreasing trend) in terms of regions and seasons. The RCPs integrate several scenarios of policy-level interventions, adaptations and vulnerability mitigation practices (Van Vuuren et al. 2011). Specifically, IPCC uses General Circulation Models (GCMs) for future climate change projections based on different Representative Concentration Pathways (RCPs) scenarios. GCM projections indicate a substantial variability of future precipitation and temperatures in terms of trends and extremes (Reshmidevi et al. 2018). Many studies have been applied and tested together with a range of viable GCM models and emission scenarios allowing a comprehensive understanding of the impacts of climate change on water resources management (Islam et al. 2005a; Touch et al. 2020), water availability (Islam et al. 2005b, 2007b; Oeurng et al. 2019), and water scarcity including virtual water trading (Islam et al. 2007a; Gosling & Arnell 2016).

The effective development of water resource use and protection under changing environmental conditions requires hydrological models that can simulate water balance components and hydrological extremes under different change scenarios. Hydrological models are the principal tools used to explore the potential effects of climate change on water resources. The present study used the semi-distributed model, Soil and Water Assessment Tool (SWAT) model (Arnold et al. 1998), to assess water balance components. SWAT has been successfully and widely used around the world to simulate basin hydrology under different climate change scenarios (Senent-Aparicio et al. 2017; Ji & Duan 2019; Zhang et al. 2019) for understanding the effects of future development and management activities. Many studies have applied and tested SWAT performance in different Mekong River basins for climate change impact assessment on water availability and even extend to water quality (Oeurng et al. 2016; Shrestha et al. 2016; Ang & Oeurng 2018; Oeurng et al. 2019; Sok et al. 2020, 2022; Touch et al. 2020; Chim et al. 2021). Moreover, the SWAT model was used to simulate the water balance in different physiographic regions of the West Seti River Basin of Nepal and to analyze the impact of climate change on hydrological extremes (Budhathoki et al. 2021). In Thailand, studies of climate change impacts on water balance and hydrological extremes were in Bang Pakong-Prachin Buri river basin (Okwala et al. 2020) and flow regimes in the Chao Phraya River Basin (Ligaray et al. 2015).

Several previous studies have been carried out in the Prek Thnot River Basin by concerned authorities and stakeholders. Most of these studies were concentrated on flood management, agricultural development, irrigation and drainage system, land-use change, and integrated water resources management (Lee et al. 2014; Fujimoto & Tomotsugu 2019; Sothea et al. 2019; UNDP 2019; Khorn et al. 2020a, 2020b; Koem & Tantanee 2020; Sagara 2021). However, none of them has focused on the overall change in water balance components and extreme hydrological events in future under the climate change scenarios.

Thus, this study aims to bridge the knowledge gap regarding climate change and its impacts on water balance components and extreme hydrological events utilizing the SWAT model's most recent data and research approach in the Prek Thnot River Basin. Furthermore, this study also analyzed the changes in extreme hydrological events due to the changing climate, which affects the water resources development projects in the basin. Three different GCMs with two different emission scenarios, RCP 4.5 and RCP 8.5, were used during the study for a baseline of 2000–2011 and two future periods: near-term future from 2021 to 2040 or 2030s and medium-term future from 2051 to 2070 or 2060s. These GCMs were selected based on previous applications for similar purposes within the region (Oeurng et al. 2019; Heng et al. 2020; Touch et al. 2020; Chim et al. 2021; Sok et al. 2022).

Study area

The Prek Thnot River Basin covers five provinces (Koh Kong, Kampong Speu, Kampot, Takeo, Kandal, and Phnom Penh) in the southwest part of Cambodia, with a coverage area of approximately 5,591 km2 (Figure 1). The geographical coordinates range from latitudes 11 °00′ to 12 °10′ N and longitudes 103 °45′ to 105 °15′ E, which is a representative inland river basin in the plate and plateau regions, and a part of Lower Mekong River Basin (Mekong Delta). Generally, the basin's elevation varies from 0 to 1,815 m above the mean sea level. Prek Thnot River, the main river in the basin, stretches 232 km (Fujimoto & Tomotsugu 2019; Sothea et al. 2019). Prek Thnot River originates from the Cardamom Mountain (1,177 meters high) in the southwest part of the country and flows through Kampong Speu and Kandal provinces in the west-east direction (Koem & Tantanee 2020). Finally, it drains to the outlet at the Bassac River, a tributary of the Lower Mekong River at the south of Phnom Penh. The climate is influenced by the annual monsoon cycle with large spatial and temporal variability of alternating two specific seasons of the tropical zone. The seasonal distribution is divided into the rainy season (May-October) and the dry season (December-April). The annual average temperature ranges from 21 °C to 35 °C. The annual rainfall ranges from 800 mm to 2,600 mm, with the average annual rainfall being approximately 1,225 mm, and more than 90% of precipitation occurs during the rainy season. Agriculture is the dominant area, covering about 85% of the total basin area, particularly during the dry season. And water use for agriculture is the main water use in the Prek Thnot River Basin, especially rice farming during the dry season.
Figure 1

Geographical location map of the Prek Thnot River Basin.

Figure 1

Geographical location map of the Prek Thnot River Basin.

Close modal

The Prek Thnot River Basin experienced maximum flood with discharge up to 1,371 m3/s in 1991 and 1,304 m3/s in 2000 (JICA 2012). In August 1991, floods affected more than 150,000 people in Kompong Speu and Kandal provinces, inundated over 15,000 hectares of paddy fields and damaged thousands of households and main roads connecting the provinces to the Phnom Penh capital (UNDP 2019). Severe floods occurred again in 2000, 2001, 2002, 2008, 2011 and 2013 (UNDP 2019). Within the Prek Thnot River Basin in 1995, drought damaged over 4,000 hectares of crops in Kampong Speu province (MoE 2001). Chhinh & Millington (2015) conducted a study on drought monitoring for rice production in the Prek Thnot River Basin in Kampong Speu province, where droughts between 1994 and 2006 damaged more than 1,000 hectares of paddy rice.

Hydro-meteorological data used in the study

Recorded streamflow and rainfall at daily scale data used in this study were obtained from the data recorded by the Department of Hydrology and River Works (DHRW) of the Ministry of Water Resources and Meteorology (MOWRAM), Cambodia. This study has eight rainfall and one streamflow gauge station along the main Prek Thnot River (Table 1).

Table 1

Recorded rainfall, streamflow, and its descriptive statistic were used in this study

Name of StationRecord usedTimestep of MeasurementMonthly (mm) & (m3/s)
PeriodMinMeanMaxSTDV
Rainfall Record Used 
Kampong Speu 1997–2011 Daily 106 232 76 
Kong Pisey 1997–2011 10 93 243 72 
Oral 1997–2011 16 98 251 72 
O Taroat 1997–2011 91 195 66 
Peam Khley 1997–2011 92 208 64 
Phnom Srouch 1997–2011 11 98 233 67 
Prey Pdao 1997–2011 101 228 68 
Trapeang Chor 1997–2011 13 94 242 71 
Streamflow Record Used 
Peam Khley 2000–2011 Daily 47 192 54 
Name of StationRecord usedTimestep of MeasurementMonthly (mm) & (m3/s)
PeriodMinMeanMaxSTDV
Rainfall Record Used 
Kampong Speu 1997–2011 Daily 106 232 76 
Kong Pisey 1997–2011 10 93 243 72 
Oral 1997–2011 16 98 251 72 
O Taroat 1997–2011 91 195 66 
Peam Khley 1997–2011 92 208 64 
Phnom Srouch 1997–2011 11 98 233 67 
Prey Pdao 1997–2011 101 228 68 
Trapeang Chor 1997–2011 13 94 242 71 
Streamflow Record Used 
Peam Khley 2000–2011 Daily 47 192 54 

SWAT conceptual model

The SWAT model is a physically-based, semi-distributed, agro-hydrological and continuous hydrological model developed for water resources managers to decide the most appropriate strategy or solution by considering the impact of different management practices on streamflow and non-point source pollution, which is a model developed jointly by the United States Department of Agriculture-Agricultural Research Services (USDA-ARS) and Agricultural Experiment Station in Temple, Texas (Arnold et al. 1998). SWAT has undergone continuous improvement over the past few decades (Gassman et al. 2007; Arnold et al. 2012) and its applicability and credibility have been verified in Southeast Asia (Tan et al. 2019). The model has some privileges in predicting climate change that contributes to water-related and hydrological processes over long-term simulation periods, continuous-time, lumped parameter, deterministic, river basin scale model, and originated from agricultural models. The SWAT model can simulate various time steps, including annual, monthly, and daily scales (Neitsch et al. 2011). The SWAT model can be used in small or large catchments by discretizing them into sub-basins, further separated into hydrological response units (HRUs) with homogeneous land use soil type and terrain slope class. The hydrological processes simulated by SWAT are based on the following water balance equation (Equation (1)):
(1)
where SWt is the ultimate soil water content (mm H2O), SWo is the initial soil water content on day ith (mm), t is the time (days), Rday is the amount of precipitation on day ith (mm H2O), Qsur is the amount of surface runoff on day ith (mm H2O), Ea is the total of evapotranspiration on day ith (mm H2O), Wseep is the total of water infiltrating the vadose zone from the soil profile on the day (mm H2O), and Qgw is the total of return flow on day ith (mm H2O).

SWAT model input

The SWAT model was set up to cover about 5,600 km2 from the most upstream of the Prek Thnot River Basin (Figure 1). After the preparation of forcing data, four significant steps must be done to set up the model: (i) watershed discretization and sub-watershed characteristics derivation, (ii) define the HRU, (iii) run the model and analyze the parameter sensitivity; and (iv) calibrate and validate the SWAT model, including uncertainty analysis. Each data input is obtained from different sources, which is summarized in Table 2.

Table 2

Summary of data and their corresponding sources in the SWAT model

Data TypesDescriptionSpatial ResolutionTemporal ResolutionData Sources
Topography map DEM 30 m  Shuttle Radar Topography Mission (SRTM) (http://srtm.csi.cgiar.org.) 
Landuse/land-cover (LULC) map Landuse classification 250 m×250 m 2002 MRC 
Soil map Soil types 250 m×250 m 2002 MRC 
Meteorology Observed rainfall 8 stations Daily, 1997–2011 DHRW of MOWRAM 
Hydrology Observed Streamflow 8 stations Daily, 2000–2010 DHRW of MOWRAM 
Climate Weather Gridded climate data 0.25° Daily, 1997–2011 Global Weather Data (http://globalweather.tamu.edu
General Circulation Models (GCMs) data Climate Change Scenarios RCP2.6&8.5 Change factor in the subbasin Monthly, 2030 and 2060s MRC 
Data TypesDescriptionSpatial ResolutionTemporal ResolutionData Sources
Topography map DEM 30 m  Shuttle Radar Topography Mission (SRTM) (http://srtm.csi.cgiar.org.) 
Landuse/land-cover (LULC) map Landuse classification 250 m×250 m 2002 MRC 
Soil map Soil types 250 m×250 m 2002 MRC 
Meteorology Observed rainfall 8 stations Daily, 1997–2011 DHRW of MOWRAM 
Hydrology Observed Streamflow 8 stations Daily, 2000–2010 DHRW of MOWRAM 
Climate Weather Gridded climate data 0.25° Daily, 1997–2011 Global Weather Data (http://globalweather.tamu.edu
General Circulation Models (GCMs) data Climate Change Scenarios RCP2.6&8.5 Change factor in the subbasin Monthly, 2030 and 2060s MRC 

The climate data used in this study was obtained from Global Weather Data for SWAT (globalweather.tamu.edu). The study used the DEM with a resolution of 30 m from the National Aeronautics and Space Administration of Shuttle Radar Topography Mission (SRTM) on the ASTER-GDEM official website (http://srtm.csi.cgiar.org.) (Figure 2(a)) (SRTM 2015). The SRTM DEM was generated on the near-global scale by the United States Geological Survey (USGS). The elevation distribution varies from 0 m to 1,815 m, representing the topographic condition of the SWAT model set-up for the Prek Thnot River Basin. The Mekong River Commission (MRC) established and validated LULC and soil map layers in 2002 for the whole Mekong River Basin, in which the Prek Thnot River Basin is included. Therefore, LULC and soil map in 2002 from MRC are the applicable and reliable data for this study. The study's LULC map distribution in the Prek Thnot River Basin was derived from satellite imagery and field data collected in 2002 with 250 m horizontal resolution (Figure 2(b)). The soil map in 2022 was developed from base maps with a 250 m horizontal resolution (Figure 2(c)). For this specific study, a 5% threshold value for land use, 10% for soil, and 5% for slope were used. This study determined the HRU distribution by assigning multiple HRUs to each sub-basin (Shawul et al. 2013). The watershed was discretized into small 46 sub-basins, which is equal to 788 HRUs from 18 land uses, 17 soils and five slope classes (0–3%, 3%–8.5%, 8.5%–26%, 26%–60.5%, and >60.5%).
Figure 2

Spatial maps of the Prek Thnot River Basin: DEM (Top), land use distribution (bottom left), and soil type distribution (bottom right).

Figure 2

Spatial maps of the Prek Thnot River Basin: DEM (Top), land use distribution (bottom left), and soil type distribution (bottom right).

Close modal

Model calibration and validation

SWAT is a physically-based, semi-distributed, agro-hydrological simulation model that operates on a sub-daily to annual scale time step on a watershed scale, and it simulates the overall water balance for each HRU. SWAT was developed to predict the impact of management on water, sediment, and agricultural chemical yields in ungauged catchments (Arnold et al. 2012). The calibration was performed by using the manual and a semi auto-calibration procedure (SWAT-CUP), which was applied using a sequential uncertainty fitting algorithm (SUFI-2) detailed by Abbaspour (2014) with the comparison of observed data and literature review information for overall hydrology components. In order to calculate potential evapotranspiration, the Penman-Monteith method was selected. The parameters controlling groundwater behavior in the model and dependent on spatial data have been calibrated with literature daily (Table 3). The parameters were calibrated/validated in the gauge station, namely Peam Khley. The calibration results showed the importance of parameters, such as Soil_AWC, Soil_K, and ALPHA_BF (groundwater parameter) in the studied flow of the analyzed in Prek Thnot River Basin (Table 3). The parameter CN2 is significantly related to the runoff quantity and depends on soil utilization. Soil_K and Soil_AWC are related to the quantity of soil-water relationships in various soil types of the region. The model performance was measured by comparing the simulated with the observed components using the Nash–Sutcliffe model efficiency factor (NSE) as the objective function. The Nash-Sutcliffe efficiency (NSE), percent bias (PBIAS), the ratio of mean squared error to the standard deviation of the measured data (RSR), and coefficient of determination (R2) are frequently used measures in hydrological modeling studies (Moriasi et al. 2015).

Table 3

Calibrated values of SWAT parameters for the Prek Thnot River Basin

ParameteraDescription of parametersLiterature rangeCalibrated value
v_ALPHA_BF.gw Baseflow alpha factor (days) 0–1 0.975 
a_GW_DELAY.gw Groundwater delay (days) 0–500 1.75 
v_GW_REVAP.gw Groundwater ‘revap’ coefficient 0.02–0.20 0.0596 
v_GWQMN.gw Threshold depth of water in the shallow aquifer required for return flow to occur (mm) 0–5,000 570 
a_RCHRG_DP.gw Deep aquifer percolation fraction 0–1 0.2 
a_REVAPMN.gw Threshold depth of water in the shallow aquifer for ‘revap’ to occur (mm) 0–5,000 103.75 
v_ESCO.hru Soil evaporation compensation factor 0–1 0.595 
r_CN2.mgt SCS runoff curve number 35–98 [54.12–88.85]2 
r_CH_K2.rte Effective hydraulic conductivity in main channel alluvium (mm/h) −0.01 to 500 386 
r_SOL_AWC.sol Available water capacity of the soil layer (mm H2O/mm soil) 0–1 [0.33–0.49]3 
r_SOL_K.sol Depth soil surface to bottom of layer (mm/hr) 0–2,000 [76.41; 769.43] 
r_SOL_Z.sol Saturated hydraulic conductivity (mm/hr) 0–3,500 [390.50; 2603.60] 
ParameteraDescription of parametersLiterature rangeCalibrated value
v_ALPHA_BF.gw Baseflow alpha factor (days) 0–1 0.975 
a_GW_DELAY.gw Groundwater delay (days) 0–500 1.75 
v_GW_REVAP.gw Groundwater ‘revap’ coefficient 0.02–0.20 0.0596 
v_GWQMN.gw Threshold depth of water in the shallow aquifer required for return flow to occur (mm) 0–5,000 570 
a_RCHRG_DP.gw Deep aquifer percolation fraction 0–1 0.2 
a_REVAPMN.gw Threshold depth of water in the shallow aquifer for ‘revap’ to occur (mm) 0–5,000 103.75 
v_ESCO.hru Soil evaporation compensation factor 0–1 0.595 
r_CN2.mgt SCS runoff curve number 35–98 [54.12–88.85]2 
r_CH_K2.rte Effective hydraulic conductivity in main channel alluvium (mm/h) −0.01 to 500 386 
r_SOL_AWC.sol Available water capacity of the soil layer (mm H2O/mm soil) 0–1 [0.33–0.49]3 
r_SOL_K.sol Depth soil surface to bottom of layer (mm/hr) 0–2,000 [76.41; 769.43] 
r_SOL_Z.sol Saturated hydraulic conductivity (mm/hr) 0–3,500 [390.50; 2603.60] 

aThe qualifier (r_) states to relative change, i.e., the existing parameter must be multiplied by (1+ the value found in calibration), (v_) refers to the value of the current parameter must be replaced by the value found in calibration, and (a_) means that absolute modification (i.e., the calibrated value must be added to the original value of the parameter). 2 Varies by HRU. 3 Varies by soil layer.

Analysis of water balance components

Water balance in the SWAT model plays a critical role and contains all the processes happening within the basin (Neitsch et al. 2011). The principal water balance components are precipitation, evapotranspiration, surface runoff, lateral flow, groundwater contribution to streamflow (Baseflow), and water yield (Ayivi & Jha 2018; Budhathoki et al. 2021). Equations (2) and (3) show water yield and baseflow.
(2)
(3)

Analyzing hydrological changes

Two techniques were employed to analyze different aspects of hydrological changes. First, annual, seasonal, and monthly streamflow statistics were calculated to understand the river's flow regime changes. The second technique, the Q5 and Q95, was calculated to analyze the high-flow and low-flow conditions. The high flow condition exceeding 5% of the time (Q5) and low flow condition exceeding 95% (Q95) were calculated to evaluate the changes in flood and drought characteristics, respectively. The flow duration curve was used to identify the Q5 and Q95. These results were analyzed at the basin scale for 12 years (2000–2011). Hydrological changes were calculated under individual scenarios and ensembles, i.e., average changes from three different GCMs under both RCPs in two-time horizons.

Analyzing extreme streamflow

The annual maximum and minimum daily streamflow have been extracted from the simulated streamflow data sets to analyze changes in frequency, magnitude, and timing of extreme streamflow over different time slices aspects of flow changes under GCMs for both RCPs compared to the baseline period were employed with further analysis for floods and droughts. Data have been tested to check independence, stationarity, and homogeneity, as these are the assumptions of the frequency analysis. Parametric frequency analysis has been performed on the annual maximum series using the Consolidated Frequency Analysis (CFA) software package and on the annual minimum series using the Low-Flow Frequency Analysis (LFA) software package, both made available by Environment Canada (Pilon & Jackson 1988; Pilon & Harvey 1994). The probability distributions of CFA used to fit these datasets were the Generalized Extreme Value Distribution (GEV) for the annual maximum series and the Weibull distribution for the annual minimum series, as these distributions are recommended for fitting such data series by several statistical hydrology-related textbooks (Mohammed et al. 2017). The GEV distribution's Probability Density Function (PDF) with variable x is given by:
(4)
where a, b, and c are the scale, shape, and location parameters.

The range of x depends on the value of b; it is bounded by c+(a/b) from above for b>0, i.e., bounded from below for b<0, i.e., the shape parameter b determines which extreme value distribution is represented. Depending on the value of b, Equation (2) corresponds to the Fisher-Tippett distribution types I, II, and III.

Several probability distributions apply to LFA, with the Weibull distributions being the most widely used (Nathan & McMahon 1990). The f(x) distribution is a reverse form of the Extreme Value Type III (EV3) distribution; Equation (5) is defined by the following probability density function given by:
(5)
where =lower boundary, =characteristic drought, and =shape parameter. The Weibull distribution may give unsatisfactory estimates because e can be negative or above the observed annual minimum.

Selected climate change scenarios and GCMs

Climate change scenarios were developed using different Global Circulation Models (GCMs), emission scenarios, time horizons, and locations (MRC 2015, 2017). The most important source of uncertainty in flow, irrigation demand and sediment modeling attributed to the different GCMs have been previously identified (Shrestha et al. 2016; Oeurng et al. 2019; Heng et al. 2020; Touch et al. 2020; Sok et al. 2022). Hence, the GCMs selection for climate change scenarios is an essential model development procedure. This study selected the future climate change scenarios based on the MRC (2015) that recommended three different GCMs (GISS-E2-R-CC, IPSL-CM5-MR and GFDL-CM3). For each GCM, we focused on two emission scenarios of Representative Concentration Pathways (RCPs), namely RCP2.6 and RCP8.5 were suitable for the widespread impacts of climate change. RCP 2.6 is a very low scenario in which radiative forcing peaks by the mid-21st century and gradually declines to 2.6 W/m2 by 2100 (Taylor et al. 2012). RCP8.5 is a scenario of high greenhouse gas emissions increasing throughout the twenty-first century before reaching an 8.5 W/m2 radiative forcing level by 2100 (Riahi et al. 2011). Moreover, RCP 8.5 is a plausible and accurate representation of the concentrations of atmospheric carbon that would be reached on the business-as-usual path. The difference between RCP 2.6 and 4.5 remains relatively small until the end of the century. Therefore, these RCP scenarios were selected based on their projects, which are designed to synthesize impact projections in the river flow, water temperature, precipitation, agriculture, biome, health, and infrastructure sectors at different levels of global warming. Furthermore, RCP2.6 and RCP8.5 are commonly selected in Mekong River Basin and Cambodia (Heng et al. 2020; Kayiranga et al. 2021; Wang et al. 2021; Sok et al. 2022). Based on an MRC (2017) study, two time horizons, 2030s (near-term future 2021–2040) and 2060s (medium-term future 2051–2070), were considered in this study, as the MRC is using these time horizons to maximize the amount of uncertainty.

The Department of Climate Change and Adaptation Initiative (CCAI) of the Mekong River Commission was obtained and downscaled climate change data sets (IPCC 5th Assessment Report). This dataset contains a monthly SWAT model (change factors) for precipitation, temperature, and climate weather (solar radiation and relative humidity). MRC CCAI uses SIMCLIM software to downscale the climate dataset. SimCLIM downscales GCM outputs using pattern scaling and a bilinear interpolation method. MRC CCAI conducts the change factor technique to quantify projected climate changes because it is the most helpful and straightforward technique to generate scenarios depending on multiple GCMs, emission scenarios, sensitivities, time horizons, and locations (MRC 2015, 2017). The detailed information in future climate scenarios used for this study was summarized in Table 4.

Table 4

Future climate scenarios used for this study

GCMs (Model ID)Emission ScenarioPattern of ChangeSpatial Resolution (longitude x latitude)
Geophysical Fluid Dynamics Laboratory Climate model version 3 (GFDL-CM3) RCP2.6 RCP8.5 Wetter overall 2.5°×2° 
Institute Pierre-Simon Laplace Coupled Model, version 5A, coupled with NEMO, mid resolution (IPSL-CM5A-MR) RCP2.6 RCP8.5 Increased seasonal variability 2.5°×1.27° 
Goddard Institute for Space Studies Model E2, coupled with the Russell Ocean model, with the carbon cycle (GISS-E2-R-CC) RCP2.6 RCP8.5 Drier overall 2.5°×2° 
GCMs (Model ID)Emission ScenarioPattern of ChangeSpatial Resolution (longitude x latitude)
Geophysical Fluid Dynamics Laboratory Climate model version 3 (GFDL-CM3) RCP2.6 RCP8.5 Wetter overall 2.5°×2° 
Institute Pierre-Simon Laplace Coupled Model, version 5A, coupled with NEMO, mid resolution (IPSL-CM5A-MR) RCP2.6 RCP8.5 Increased seasonal variability 2.5°×1.27° 
Goddard Institute for Space Studies Model E2, coupled with the Russell Ocean model, with the carbon cycle (GISS-E2-R-CC) RCP2.6 RCP8.5 Drier overall 2.5°×2° 

Streamflow calibration and validation of the SWAT model

The graphical results of daily streamflow simulation performance during the calibration and validation periods are shown in Figure 3. The year 2006 was not incorporated in the validation period since no historical data daily streamflow was observed. The streamflow during calibration periods is better matched with observation flow than the validation. Statistical indicators of the model performance of streamflow simulation for calibration and validation periods are presented in Table 5.
Table 5

Model performance evaluation of daily streamflow during calibration and validation period in the Peam Kley station

Statistical indicatorsCalibration (2000–2006)Model PerformanceValidation (2007–2010)Model Performance
NSE 0.65 Satisfactory 0.48 Unsatisfactory 
R² 0.66 Satisfactory 0.60 Satisfactory 
PBIAS −4.79% Very good −11.85% Satisfactory 
RSR 0.59 Satisfactory 0.72 Good 
Statistical indicatorsCalibration (2000–2006)Model PerformanceValidation (2007–2010)Model Performance
NSE 0.65 Satisfactory 0.48 Unsatisfactory 
R² 0.66 Satisfactory 0.60 Satisfactory 
PBIAS −4.79% Very good −11.85% Satisfactory 
RSR 0.59 Satisfactory 0.72 Good 
Figure 3

Observed and simulated daily streamflow for the Prek Thnot River Basin: (a) calibration (2000–2005) and (b) validation (2007–2010).

Figure 3

Observed and simulated daily streamflow for the Prek Thnot River Basin: (a) calibration (2000–2005) and (b) validation (2007–2010).

Close modal
The results of the calibration periods (2000–2005) and the validation periods (2007–2010) showed the statistical indicator with the NSE values of 0.65 and 0.48 and PBIAS values of −4.79% and −11.85%. Based on the results of NSE and PBIAS values, it would suggest that the streamflow model generally reached satisfactory performance (Moriasi et al. 2015). Since the PBIAS values are a negative sign, the model is underestimated for both the calibration and validation periods shown in Figure 4, according to Moriasi et al. (2015). In addition, peak discharge every year was practically underestimated. The overall performance of the model calibration and validation replicated the observed flows for an independent dataset reasonably well.
Figure 4

Correlation between observed and simulated daily river discharge of Prek Thnot River Basin: (a) calibration period (2000–2005) and (b) validation period (2007–2010).

Figure 4

Correlation between observed and simulated daily river discharge of Prek Thnot River Basin: (a) calibration period (2000–2005) and (b) validation period (2007–2010).

Close modal

Water balance components in the Prek Thnot River Basin

Estimation of baseline water balance and components was essential to understanding the detailed information in the Prek Thnot River Basin. It is perfect for analyzing and quantifying the diverse components of hydrological processes occurring within the study area to deal with water management issues. The SWAT model estimated relevant water balance components in addition to the daily streamflow in the Prek Thnot River Basin. Within the validated model result, hydrological components included the precipitation, potential evapotranspiration, actual evapotranspiration, lateral flow, percolation, groundwater, surface runoff, and water yield of the baseline period from 2000 to 2011. The contribution was made by each sub-basin (Figure 5).
Figure 5

Water balance components on the Prek Thnot River Basin: Precipitation, Potential Evapotranspiration, Actual Evapotranspiration, Lateral flow, Percolation, Groundwater, Surface runoff, and Water yield of baseline period from 2000 to 2011.

Figure 5

Water balance components on the Prek Thnot River Basin: Precipitation, Potential Evapotranspiration, Actual Evapotranspiration, Lateral flow, Percolation, Groundwater, Surface runoff, and Water yield of baseline period from 2000 to 2011.

Close modal

The basin is dominated by the tropical monsoon season, distinguished into two seasons, the rainy season from May to October and the dry season from November to April. The monthly contribution of water balance components is high in October during the wet season and contributes low value in February (Table 6). The annual average rainfall during the study period in the Prek Thnot River Basin was 1,169 mm; 72% (848 mm) of the average rainfall was withdrawn by actual evapotranspiration, 23% (272 mm) returned flow to the streamflow, while another 8% recharged to deep aquifer groundwater (Table 7). Water yield of 272 mm has contributed to surface runoff (proportion of 16%), lateral flow (proportion of 48%), and groundwater (proportion of 35%). These results revealed that actual evapotranspiration contributed to the basin's most considerable amount of water loss.

Table 6

Monthly average water balance components (precipitation, potential evapotranspiration, actual evapotranspiration, surface runoff, lateral flow, and water yield) of the Prek Thnot River Basin for the baseline period from 2000 to 2011

PRECIPPETAETSURQLAT_QWYLD
Monthly(mm)
Jan 16.2 148.1 48.1 1.1 1.3 6.1 
Feb 12.8 140.3 54.4 0.6 0.8 3.4 
Mar 59 150.3 70.0 3.2 3.7 7.8 
Apr 106.4 139.0 68.4 4.6 6.7 10.4 
May 100.7 144.9 73.6 2.3 8.6 11.6 
Jun 102.6 128.7 70.3 2.5 7.1 9.7 
Jul 128.3 122.7 75.3 8.7 12.3 22.2 
Aug 143.4 115.7 76.7 9.3 12.2 22.6 
Sep 167.1 104.3 79.4 14.5 19.2 39.6 
Oct 242.7 110.7 86.8 41.6 45.0 104.1 
Nov 70.5 119.1 84.1 6.4 11.8 25.8 
Dec 19.6 134.3 61.2 1.2 2.5 8.5 
PRECIPPETAETSURQLAT_QWYLD
Monthly(mm)
Jan 16.2 148.1 48.1 1.1 1.3 6.1 
Feb 12.8 140.3 54.4 0.6 0.8 3.4 
Mar 59 150.3 70.0 3.2 3.7 7.8 
Apr 106.4 139.0 68.4 4.6 6.7 10.4 
May 100.7 144.9 73.6 2.3 8.6 11.6 
Jun 102.6 128.7 70.3 2.5 7.1 9.7 
Jul 128.3 122.7 75.3 8.7 12.3 22.2 
Aug 143.4 115.7 76.7 9.3 12.2 22.6 
Sep 167.1 104.3 79.4 14.5 19.2 39.6 
Oct 242.7 110.7 86.8 41.6 45.0 104.1 
Nov 70.5 119.1 84.1 6.4 11.8 25.8 
Dec 19.6 134.3 61.2 1.2 2.5 8.5 

PRECIP, precipitation; AET, actual evapotranspiration; PET, potential evapotranspiration; SURQ, surface runoff contribution to streamflow; LAT_Q, lateral flow; and WYLD, the net amount of water that contributes to streamflow (SURQ+LAT_Q+groundwater contribution to streamflow – transmission losses).

Table 7

Annual average water balance components of the Prek Thnot River Basin for the baseline period (2000–2011)

Water Balance ComponentAmount (mm)
Precipitation 1,169 
Evapotranspiration 848 
Surface runoff 96 
Lateral flow 131 
Groundwater (Shallow Aquifer) flow 85 
Groundwater (Deep Aquifer) flow 34 
Groundwater contribution to streamflow (Baseflow) 119 
Water yield 272 
Water Balance ComponentAmount (mm)
Precipitation 1,169 
Evapotranspiration 848 
Surface runoff 96 
Lateral flow 131 
Groundwater (Shallow Aquifer) flow 85 
Groundwater (Deep Aquifer) flow 34 
Groundwater contribution to streamflow (Baseflow) 119 
Water yield 272 

Impact of climate change on water balance components

Projected annual change of the water balance components at a basin scale

The impact of climate change on water balance was evaluated based on different GCMs and emission scenarios. The projected annual average water balance components (precipitation, actual evapotranspiration, surface runoff, lateral flow, groundwater flow and total water yield) and its percentage change in Prek Thnot River Basin for three different GCMs under two RCP scenarios in the 2030s and 2060s at the basin scale are represented in Table 8 and Figure 6. Overall, a high percentage change could be expected in almost every water balance component for the medium-term future (2060s) compared to the near-term future (2030s) under both RCP scenarios. The projected annual water balance components will see a decreasing trend in most GCMs (except GFDL-CM3) and time horizons for both RCPs. The GW_Q, WYLD, SURQ, and PERC parameters are ordered from highest to lowest change. While the AET presented the lowest relative change with an increase of 1.3% and 1.8% (near-term future and medium-term future) for RCP2.6 and 5.2% and 1.4% (near-term future and medium-term future) for RCP8.5 and a decrease of 0.5% and 0.7% (near-term future and medium-term future) for RCP2.6 and 2.6%, and 8.4% (near-term future and medium-term future) for RCP8.5. Similar to the study by Sok et al. (2022), the AET also showed less change, an increase and a decrease for the near and medium term with the GCMs under the two RCP scenarios.
Table 8

Projected annual average water balance components and its percentage change compared to the baseline condition (from 2000 to 2011) based on GCMs under emission scenarios RCP2.6 and RCP8.5 in the 2030s and 2060s for Prek Thnot River Basin

ScenarioTime PeriodGCMs ModelPRECIPAETSURQLAT_QGW_QWYLD
(mm/year)
Projected Annual Water Balance Components for Prek Thnot River Basin 
Baseline 2000–2011  1,169 848 96 131 119 272 
RCP 2.6 2030s GFDL 1,192 859 102 136 124 282 
(2021–2040) GISS 1,144 844 87 123 109 253 
 IPSL 1,166 847 95 131 118 270 
2060s GFDL 1,200 863 104 135 125 285 
(2051–2070) GISS 1,136 842 85 125 106 248 
 IPSL 1,165 846 94 131 118 270 
RCP 8.5 2030s GFDL 1,259 892 119 144 137 311 
(2021–2040) GISS 1,072 826 65 108 83 205 
 IPSL 1,157 840 93 130 115 267 
2060s GFDL 1,362 936 140 158 159 360 
(2051–2070) GISS 960 777 46 86 60 147 
 IPSL 1,140 837 89 129 112 265 
Annual Change of Water Balance Components for Prek Thnot River Basin 
RCP 2.6 2030s GFDL 2.0 1.3 6.3 3.4 4.2 3.7 
(2021–2040) GISS −2.1 −0.5 −9.4 −6.1 −8.4 −7.0 
 IPSL −0.3 −0.1 −1.0 −0.3 −0.8 −0.7 
2060s GFDL 2.7 1.8 8.3 2.6 5.0 4.8 
(2051–2070) GISS −2.8 −0.7 −11.5 −4.7 −10.9 −8.8 
 IPSL −0.4 −0.3 −1.8 −0.3 −1.3 −0.9 
RCP 8.5 2030s GFDL 7.7 5.2 24.0 9.8 15.1 14.3 
(2021–2040) GISS −8.3 −2.6 −32.3 −17.5 −30.3 −24.6 
 IPSL −1.0 −0.9 −3.1 −0.9 −3.4 −1.8 
2060s GFDL 16.5 10.4 45.8 20.6 33.6 32.4 
(2051–2070) GISS −17.9 −8.4 −52.1 −34.3 −49.6 −46.0 
 IPSL −2.5 −1.3 −7.3 −1.2 −5.9 −2.6 
ScenarioTime PeriodGCMs ModelPRECIPAETSURQLAT_QGW_QWYLD
(mm/year)
Projected Annual Water Balance Components for Prek Thnot River Basin 
Baseline 2000–2011  1,169 848 96 131 119 272 
RCP 2.6 2030s GFDL 1,192 859 102 136 124 282 
(2021–2040) GISS 1,144 844 87 123 109 253 
 IPSL 1,166 847 95 131 118 270 
2060s GFDL 1,200 863 104 135 125 285 
(2051–2070) GISS 1,136 842 85 125 106 248 
 IPSL 1,165 846 94 131 118 270 
RCP 8.5 2030s GFDL 1,259 892 119 144 137 311 
(2021–2040) GISS 1,072 826 65 108 83 205 
 IPSL 1,157 840 93 130 115 267 
2060s GFDL 1,362 936 140 158 159 360 
(2051–2070) GISS 960 777 46 86 60 147 
 IPSL 1,140 837 89 129 112 265 
Annual Change of Water Balance Components for Prek Thnot River Basin 
RCP 2.6 2030s GFDL 2.0 1.3 6.3 3.4 4.2 3.7 
(2021–2040) GISS −2.1 −0.5 −9.4 −6.1 −8.4 −7.0 
 IPSL −0.3 −0.1 −1.0 −0.3 −0.8 −0.7 
2060s GFDL 2.7 1.8 8.3 2.6 5.0 4.8 
(2051–2070) GISS −2.8 −0.7 −11.5 −4.7 −10.9 −8.8 
 IPSL −0.4 −0.3 −1.8 −0.3 −1.3 −0.9 
RCP 8.5 2030s GFDL 7.7 5.2 24.0 9.8 15.1 14.3 
(2021–2040) GISS −8.3 −2.6 −32.3 −17.5 −30.3 −24.6 
 IPSL −1.0 −0.9 −3.1 −0.9 −3.4 −1.8 
2060s GFDL 16.5 10.4 45.8 20.6 33.6 32.4 
(2051–2070) GISS −17.9 −8.4 −52.1 −34.3 −49.6 −46.0 
 IPSL −2.5 −1.3 −7.3 −1.2 −5.9 −2.6 

PRECIP, precipitation; AET, actual evapotranspiration; SURQ, surface runoff contribution to streamflow; LAT_Q, lateral flow; GW_Q, groundwater contribution to streamflow; and WYLD, the net amount of water that contributes to streamflow (SURQ+LAT_Q+GW_Q – transmission losses).

Figure 6

Projected changes of water balance components in different GCMs (top) and range of projected change represented as an ensemble of selected GCMs outputs (bottom) under RCP2.6 and RCP8.5 scenarios compared with the baseline in the 2030s (left) and 2060s (right).

Figure 6

Projected changes of water balance components in different GCMs (top) and range of projected change represented as an ensemble of selected GCMs outputs (bottom) under RCP2.6 and RCP8.5 scenarios compared with the baseline in the 2030s (left) and 2060s (right).

Close modal

Under the projected climate scenarios, the GISS-E2-R-CC was projected to decrease between 2 and 18% (precipitation), 0.5–8% (actual evapotranspiration), 9–52% (surface runoff), 5–34% (lateral flow), 8–50% (groundwater flow), and 7–46% (water yield) (Figure 6). Likewise, a considerable reduction in recharges to deep aquifers is also predicted. On the other hand, the IPSL-CM5A-MR was also expected to decrease by 0.3–2.5% (precipitation), 0.1–1% (actual evapotranspiration), 1–7% (surface runoff), 0.8–4% (lateral flow), 0.8–6% (groundwater flow), and 0.7–3% (water yield). However, the GFDL-CM3 is projected to increase in range from 2–17%, 1–10%, 6–46%, 3–21%, 4–34%, and 4–32% of precipitation, actual evapotranspiration, surface runoff, lateral flow, groundwater flow and water yield, respectively. Generally, the GFDL-CM3 model exhibited an increase in annual streamflow, while the GISS-E2-R-CC and IPSL-CM5A-MR models revealed a significant decrease for all time slices. In this study, annual precipitation in the Preck Thnot River Basin was projected to increase by 8–17% for the GCMs under the two RCP scenarios, for the near and medium term, respectively, compared to the previous study; it was 5–11% according Sok et al. (2022) in Sen River Basin, the largest sub-basin of Tonle Sap Lake in the Lower Mekong Basin.

Future precipitation at the sub-basin scale

The spatial distribution of the projected changes in the ensemble means precipitation for the Prek Thnot River Basin as calculated using GCMs (GFDL-CM3, GISS-E2-R-CC and IPSL-CM5A-MR), under different emission scenarios RCP2.6 and RCP8.5 was shown in Figure 7 for the near-term future and in Figure 8 for the medium-term future. Considering the different physiographic sub-basin scales, precipitation in all basins will experience an increasing trend in most GCMs (except GISS-E2-R-CC) during both future periods and emission scenarios. Generally, GFDL-CM3 and IPSL-CM5A-MR under both RCPs, both different future periods in the basin will experience an increase in precipitation, whereas, in the GISS-E2-R-CC model, some regions will experience a decrease in precipitation. Projected precipitation changes in the medium-term future will be almost double the near-term future. The annual average change of precipitation in the near-term future is projected to increase with maximum values of 2.92% for RCP2.6 and 8.65% for RCP8.5 (Figure 7). At the same time, the projected decrease for RCP2.6 and RCP8.5 with minimum values of 3 and 9%, respectively. In the medium-term future, precipitation is projected to increase with a maximum of 2 and 20% values for RCP2.6 and RCP8.5, respectively. In contrast, the projected decrease with minimum values of 3% for RCP2.6 and 9% for RCP8.5 (Figure 8).
Figure 7

Changes in annual average precipitation under different future projection in the 2030s.

Figure 7

Changes in annual average precipitation under different future projection in the 2030s.

Close modal
Figure 8

Changes in annual average precipitation under different future projection in the 2060s.

Figure 8

Changes in annual average precipitation under different future projection in the 2060s.

Close modal

Future water yield at the sub-basin scale

The spatial distribution of water yield in the basin is expected to experience an increasing trend in most GCMs (except GISS-E2-R-CC) under different emission scenarios in the near-term future (Figure 9) and the medium-term future (Figure 10). Generally, an annual average change of water yield in the near-term future would increase with maximum projected values of 10 and 23% for RCP2.6 and RCP8.5, respectively. However, it would decrease with minimum projected values of 16% for RCP2.6 and 34% for RCP8.5 (Figure 9). In the medium-term future, annual water yield is projected to increase with maximum values of 8% for RCP2.6 and 48% for RCP8.5. At the same time, the projected decrease had minimum values of 13 and 45% for RCP2.6 and RCP8.5, respectively (Figure 10). Generally, water yield is expected to increase with GFDL-CM3 and IPSL-CM5A-MR under both RCPs and future periods; however, it is predictable to decrease with the GISS-E2-R-CC model in some basins.
Figure 9

Changes in annual average water yield under different future projections in the 2030s.

Figure 9

Changes in annual average water yield under different future projections in the 2030s.

Close modal
Figure 10

Changes in annual average water yield under different future projections in the 2060s.

Figure 10

Changes in annual average water yield under different future projections in the 2060s.

Close modal

Impact of climate change on the flow regime

The future flow changes are evaluated corresponding to the change in meteorological condition in the Prek Thnot River Basin as projected by the three different GCMs and the two RCPs and by comparing the projected streamflow in two future time horizons, the 2030s (2021–2040) and 2060s (2051–2070) to the baseline period (2000–2011). The projected hydrographs and monthly percentage changes of streamflow show a clear trend in changes (Figures 11 and 12). During the dry season (December-April), the streamflow is projected to decline based on GISS-E2-R-CC and IPSL-CM5A-MR, with the exceptions of GFDL-CM3. A significant increase in streamflow would occur in March with the GFDL-CM3 with a value range from 16% to 164%. The considerable reduction of monthly streamflow magnitude from 9% to 56% (GISS-E2-R-CC) and 14% to 64% (IPSL-CM5A-MR) would occur in March and June, respectively. In the wet season (May–October), the flow is expected to reduce for all GCMs (except GFDL-CM3). The increase in flow varies from 1% to 23% for GFDL-CM3. The highest magnitude of flow reduction ranges from 14% to 70% for GISS-E2-R and from 8% to 69% for IPSL-CM5A. Generally, the magnitude changes of streamflow vary depending on the season in which streamflow augmentation occurs in the dry and wet season for two GCMs (i.e., GISS-E2-R-CC and IPSL-CM5A-MR) except GFDL-CM3 for both RCPs and time horizons. These results are parallel with the study of Sok et al. (2022), which found that the peak seems to be high in October, and the monthly flow for the GFDL-CM3 model positively changes more than GISS-E2-R and IPSL-CM5A-MR under both scenarios for the 2030s and 2060s in the Sen River Basin. However, the GISS-E2-R model showed considerable declines every month.
Figure 11

The comparison of baseline and future monthly streamflow in different GCMs (top) and the range of projected change represented as an ensemble of selected GCMs outputs (bottom) under RCP2.6 and RCP8.5 scenarios in the 2030s (left) and 2060s (right).

Figure 11

The comparison of baseline and future monthly streamflow in different GCMs (top) and the range of projected change represented as an ensemble of selected GCMs outputs (bottom) under RCP2.6 and RCP8.5 scenarios in the 2030s (left) and 2060s (right).

Close modal
Figure 12

Projected change of monthly streamflow in different GCMs (top) and range of projected change represented as an ensemble of selected GCMs outputs (bottom) under RCP2.6 and RCP8.5 scenarios compared with the baseline in the 2030s (left) and 2060s (right).

Figure 12

Projected change of monthly streamflow in different GCMs (top) and range of projected change represented as an ensemble of selected GCMs outputs (bottom) under RCP2.6 and RCP8.5 scenarios compared with the baseline in the 2030s (left) and 2060s (right).

Close modal

Changes in hydrological extremes

Climate changes can dramatically change river flow regimes and extreme events such as floods and droughts. The high (Q5, flow equals or exceeds 5% of the time) and low (Q95, flow equals or exceeds 95% of the time) flows were extracted from flow duration curves derived from simulated daily flows. The Q5 and Q95 provide some pieces of evidence to prove that streamflow is expected to increase in magnitude through GFDL-CM3 and IPSL-CM5-MR (except GISS-E2-R-CC) under both RCP scenarios in the 2030s and 2060s (Table 9). The previous studies were conducted using the same GCMs model under RCP2.6 and RCP8.5 scenarios in the 2030s and 2060s in the Sen River Basin (Heng et al. 2020; Sok et al. 2022). Also, Oeurng et al. (2019) and Touch et al. (2020) utilized the same GCMs model but different emission rates (RCP6.0) in all rivers of the Tonle Sap Lake Basin and Pursat River Basin, respectively. The results showed that there are more incredible percentage changes in streamflow in the 2060 than in the 2030s, and the projected changes under RCP8.5 are more extensive than those by RCP2.6.

Table 9

Absolute and percentage changes in future flows, relative to the baseline period for the three GCMs under RCP2.6 and RCP8.5 in the 2030s and 2060s

Percentage changes under RCP2.6
Time HorizonsExceedance ProbabilityBaselineGFDL-CM3
GISS-E2-R-CC
IPSL-CM5A-MR
Years(%)(m3/s)(m3/s)%Change(m3/s)%Change(m3/s)%Change
2030s 944 1,000 884 −6 952 
95 182 183 165 −9 187 
2060s 944 986 894 −5 949 
95 182 185 169 −7 186 
Percentage changes under RCP8.5
RCP8.5
Time HorizonsExceedance ProbabilityBaselineGFDL-CM3
GISS-E2-R-CC
IPSL-CM5A-MR
Years(%)(m3/s)(m3/s)%Change(m3/s)%Change(m3/s)%Change
2030s 944 1,080 14 776 −18 975 
95 182 186 127 −30 189 
2060s 944 1,400 48 609 −35 1,020 
95 182 224 23 71 −61 204 12 
Percentage changes under RCP2.6
Time HorizonsExceedance ProbabilityBaselineGFDL-CM3
GISS-E2-R-CC
IPSL-CM5A-MR
Years(%)(m3/s)(m3/s)%Change(m3/s)%Change(m3/s)%Change
2030s 944 1,000 884 −6 952 
95 182 183 165 −9 187 
2060s 944 986 894 −5 949 
95 182 185 169 −7 186 
Percentage changes under RCP8.5
RCP8.5
Time HorizonsExceedance ProbabilityBaselineGFDL-CM3
GISS-E2-R-CC
IPSL-CM5A-MR
Years(%)(m3/s)(m3/s)%Change(m3/s)%Change(m3/s)%Change
2030s 944 1,080 14 776 −18 975 
95 182 186 127 −30 189 
2060s 944 1,400 48 609 −35 1,020 
95 182 224 23 71 −61 204 12 

For the Q5, in the 2030s, two GCMs (GFDL-CM3 and IPSL-CM5-MR) suggest an increase of 6–14% and 1–3%, whereas GISS-E2-R-CC predicts a decrease of 6 and 18% for RCP2.6 and RCP8.5, respectively (Figures 13 and 14). Moreover, in the 2060s, the Q5 is likely to increase by 4–48% and 1–8% for GFDL-CM3 and IPSL-CM5-MR but decrease by 5 and 35% for GISS-E2-R-CC for RCP2.6 and RCP8.5, respectively. These results indicate that future flood magnitudes in Prek Thnot River Basin would rise slightly for the GFDL-CM3 and IPSL-CM5A-MR models projection while reducing for the GISS-E2-R-CC model in the 2030s and 2060s.
Figure 13

Flow duration curves in different GCMs (top) and ensemble of selected GCMs outputs (bottom) under RCP2.6 and RCP8.5 emission scenarios compared with the baseline in the 2030s (left) and 2060s (right).

Figure 13

Flow duration curves in different GCMs (top) and ensemble of selected GCMs outputs (bottom) under RCP2.6 and RCP8.5 emission scenarios compared with the baseline in the 2030s (left) and 2060s (right).

Close modal
Figure 14

Percentage changes of Q5 and Q95 in different GCMs (top) and range of projected change represented as an ensemble of selected GCMs outputs (bottom) under RCP2.6 and RCP8.5 scenarios compared with the baseline during the future periods in the 2030s (left) and 2060s (right).

Figure 14

Percentage changes of Q5 and Q95 in different GCMs (top) and range of projected change represented as an ensemble of selected GCMs outputs (bottom) under RCP2.6 and RCP8.5 scenarios compared with the baseline during the future periods in the 2030s (left) and 2060s (right).

Close modal

For the Q95, in the 2030s, streamflow is predicted to decrease about 9 and 30% by the GISS-E2-R-CC model and, at the same time, to increase from 1–2% and 3–4% by the other two models for RCPs, respectively. In the 2060s, Q95 is expected to increase and decrease at a higher magnitude than in the 2030s. It would increase from 2 to 23% (GFDL-CM3) and 2–12% (IPSL-CM5A-MR), but it would decrease by about 7 and 61% for the GISS-E2-R-CC model for RCP2.6 and RCP8.5, respectively. The significant decrease in Q95 projected by the GISS-E2-R-CC model indicates that the drought event would significantly increase in the 2030s and 2060s. These projected decreases in Q95 would have implications for future drought for policymakers to set up measures for sustaining ecosystems and biodiversity during the dry season. In this study, the GISS-E2-R-CC model projected a decreasing trend of Q5 and Q95 under most RCP scenarios in most future periods, indicating that both the low and high flows will be lower than their baseline values similar to the previous studies (Oeurng et al. 2019; Heng et al. 2020; Touch et al. 2020; Sok et al. 2022).

Changes in frequency and magnitude of the annual peak flow

The frequency and magnitude of the annual peak flow for Prek Thnot River Basin in the 2030s and 2060s were conducted using the Consolidated Frequency Analysis (CFA) under three different GCMs and two emission scenarios (RCPs). This study simulated daily streamflow under climate change scenarios for 2-, 5-, 10-, 20-, 50-, and 100-year return periods, which correspond to the Sok et al. (2022). Based on Pradhanang et al. (2013), 2-year return periods are responsible for small floods, 10-year return periods are for moderate floods, and 100-year return periods are for large floods. The GCMs ensemble means were used to quantify the average changes in different time slices.

The annual peak flow series of each return period by the parametric frequency analysis is shown in Figure 15. The percentage changes of the annual peak flow corresponding to each return period in the 2030s and 2060s are given in Figure 16, where positive values indicate an increase and negative values indicate a decrease in peak flow. The results show an increasing range in the annual peak flow in the 2030s and 2060s for all return periods and all GCMs (except GISS-E2-R-CC) under emission scenarios (RCPs). The most significant increase is found in the 2060s with the GFDL-CM3 model under RCP8.5 for all return periods. The annual peak flow with a 2-year, 10-year, and 100-year return period is expected to increase by 52, 50, and 45% in the 2060s for the GFDL-CM3 model under RCP8.5, respectively. In this study, the GFDL-CM3 model flood for both scenarios and time horizons for the considered return periods shows greater frequency and magnitude than the baseline equivalent to the previous study (Sok et al. 2022).
Figure 15

Projection of peak flow return period curves with different GCMs (top) and range of projected change represented as an ensemble of selected GCMs outputs (bottom) under RCP2.6 and RCP8.5 scenarios compared with the baseline for the future periods in the 2030s (left) and 2060s (right).

Figure 15

Projection of peak flow return period curves with different GCMs (top) and range of projected change represented as an ensemble of selected GCMs outputs (bottom) under RCP2.6 and RCP8.5 scenarios compared with the baseline for the future periods in the 2030s (left) and 2060s (right).

Close modal
Figure 16

Percentage change of annual peak flow for different return periods with different GCMs (top) and range of projected change represented as an ensemble of selected GCMs outputs (bottom) under RCP2.6 and RCP8.5 scenarios compared with the baseline for the future periods in the 2030s (left) and 2060s (right).

Figure 16

Percentage change of annual peak flow for different return periods with different GCMs (top) and range of projected change represented as an ensemble of selected GCMs outputs (bottom) under RCP2.6 and RCP8.5 scenarios compared with the baseline for the future periods in the 2030s (left) and 2060s (right).

Close modal

Changes in frequency and magnitude of the annual low flow

The frequency and magnitudes of annual low flow in the Prek Thnot River Basin were analyzed using the Low Flow Frequency Analysis (LFA). The low flow frequency analysis describes the hydrological droughts. The return periods are estimated by the parametric frequency analysis performed on the annual low flow series. Figure 17 shows a graphical means to understand low flow events characteristics (frequency, duration, and magnitude). The percentage changes of the annual low flow corresponding to each return period in the 2030s and 2060s are given in Figure 18, where positive values indicate an increase and negative values indicate a decrease in low flow. Unlike the annual peak flow case, the results indicate a decreased uncertainty range in the annual low flow with all return periods for all GCMs (except GFDL-CM3) in the 2030s and 2060s. The low annual flow with a 2-year, 10-year, and 100-year return period is likely to increase by 47, 44, and 40% in the 2060s for the GFDL-CM3 model under RCP8.5. Generally, the frequency and magnitude of future drought demonstrated an increase with the GISS-E2-R-CC and IPSL-CM5A-MR models for RCP2.6 and RCP8.5 in the 2030s and 2060s compared to baseline for all return periods. The most significant increase is found in the 2060s with the GISS-E2-R-CC model under RCP8.5 for all return periods. Similar to Sok et al. (2022), drought magnitude in baseline conditions is higher than in the near-future and mid-future for the GISS-E2-R-CC and IPSL-CM5A-MR models for both scenarios and the considered return periods. In contrast, the GFDL-CM3 model showed greater magnitude than the baseline. Moreover, it can be observed that drought magnitudes are markedly different between RCP8.5 and RCP2.6.
Figure 17

Projection of base flow return period curve with different GCMs (top) and range of projected change represented as an ensemble of selected GCMs outputs (bottom) under RCP2.6 and RCP8.5 scenarios compared with the baseline for the future periods in the 2030s (left) and 2060s (right).

Figure 17

Projection of base flow return period curve with different GCMs (top) and range of projected change represented as an ensemble of selected GCMs outputs (bottom) under RCP2.6 and RCP8.5 scenarios compared with the baseline for the future periods in the 2030s (left) and 2060s (right).

Close modal
Figure 18

Percentage change of baseflow or low flow for different return periods with different GCMs (top) and range of projected change represented as an ensemble of selected GCMs outputs (bottom) under RCP2.6 and RCP8.5 scenarios compared with the baseline for the future periods in the 2030s (left) and 2060s (right).

Figure 18

Percentage change of baseflow or low flow for different return periods with different GCMs (top) and range of projected change represented as an ensemble of selected GCMs outputs (bottom) under RCP2.6 and RCP8.5 scenarios compared with the baseline for the future periods in the 2030s (left) and 2060s (right).

Close modal

It is important to note that changes in flow regime, increasing peak flow, and decreasing low flow could threaten river ecosystems and socio-economic development, particularly in agriculture and aquaculture. However, climate change is not the only driving factor for the water balance components and hydrological extremes. Land-use change and other human activities (e.g., dam, reservoir, withdrawal, diversion, etc.) in the river system also play a crucial role in such hydrological extremes and could be the research interest for further investigations.

This study is the first and most detailed climate change assessment of the Prek Thnot River Basin's hydrology. The SWAT model was applied to assess the impacts of climate change on water balance components and extreme hydrological events in the Prek Thnot River Basin. The model was successfully built with good performance to examine the watershed properly. The hydrological response of the basin was simulated using the daily streamflow for three GCMs under two RCP scenarios for the near future (2021–2040) and mid-future (2051–2070).

Under climate change scenarios, the annual water balance components mostly decrease spatially across the basin. Individual GCMs and scenarios showed considerable differences in patterns and magnitude changes in spatiotemporal precipitation and total water yield. The ensemble means scenario showed decreases from the baseline condition in annual precipitation and total water yield under both RCP scenarios for the near- and mid-future periods.

The results further found the changes in the annual distribution of streamflow of the Prek Thnot River Basin. Climate change profoundly affecting the flow regime would be positively altered (increase compared to the baseline) with the GFDL-CM3 model under both RCPs in all future periods; at the same time, GISS-E2-R-CC and IPSL-CM5A-MR models would be negatively altered (decrease compared to the baseline). Generally, future monthly streamflow was projected to increase during the wet season while decreasing during the dry season.

Through the effect of climate change on hydrological extremes, peak flow would be more substantial in magnitude and frequency with GFDL-CM3 and IPSL-CM5A-MR models under both RCP scenarios for all future periods. However, the low flow with GISS-E2-R-CC and IPSL-CM5A-MR models is projected to decrease significantly in both magnitude and frequency. Under these projected changes, simulated peak flow increased slightly from the near to mid-future under both RCPs, while the low flow decreased. Additionally, extreme peak flow events increased in both magnitude and frequency, and extremely low flows are projected to occur less often under climate change.

This study discussed that the prediction results depend on the climate model used. The increasing peak flow and decreasing low flow could threaten river ecosystems and socio-economic development, particularly in agriculture. However, climate change is not the only driving factor for the water balance components and hydrological extremes. Land-use change and other affected by human impact (e.g., dam, reservoir, withdrawal, diversion, etc.) in the river system also play a crucial role in such hydrological extremes, which can be a part of the further investigation caused by these factors that were not included in the study. This study suggests the careful and efficient selection of structural and non-structural measures and water management plans to mitigate the impacts of climate change in the water resources system.

The authors would like to acknowledge the Cambodia Higher Education Improvement Project (Credit No.6221-KH) of the Ministry of Education, Youth and Sports (MOEYS) for financial support of the study. The authors would also like to thank the Ministry of Water Resources and Meteorology (MOWRAM) and the Mekong River Commission (MRC) of Cambodia for providing the data.

Data cannot be made publicly available; readers should contact the corresponding author for details.

The authors declare there is no conflict.

Abbaspour
K.
2014
SWAT-CUP 2012: SWAT Calibration and Uncertainty Programs – A User Manual
.
Swiss Federal Institute of Aquatic Science and Technology
,
Switzerland
.
Arnold
J. G.
,
Srinivasan
R.
,
Muttiah
R. S.
&
Williams
J. R.
1998
Large area hydrologic modeling and assessment part I: model development 1
.
JAWRA Journal of the American Water Resources Association
34
(
1
),
73
89
.
Arnold
J. G.
,
Moriasi
D. N.
,
Gassman
P. W.
,
Abbaspour
K. C.
,
White
M. J.
,
Srinivasan
R.
,
Santhi
C.
,
Harmel
R.
,
Van Griensven
A.
&
Van Liew
M. W.
2012
SWAT: Model use, calibration, and validation
.
Transactions of the ASABE
55
(
4
),
1491
1508
.
Ayivi
F.
&
Jha
M. K.
2018
Estimation of water balance and water yield in the Reedy Fork-Buffalo Creek Watershed in North Carolina using SWAT
.
International Soil and Water Conservation Research
6
(
3
),
203
213
.
Blanco-Gómez
P.
,
Jimeno-Sáez
P.
,
Senent-Aparicio
J.
&
Pérez-Sánchez
J.
2019
Impact of climate change on water balance components and droughts in the Guajoyo River Basin (El Salvador)
.
Water
11
(
11
),
2360
.
Budhathoki
A.
,
Babel
M. S.
,
Shrestha
S.
,
Meon
G.
&
Kamalamma
A. G.
2021
Climate change impact on water balance and hydrological extremes in different physiographic regions of the West Seti River Basin, Nepal
.
Ecohydrology & Hydrobiology
21
(
1
),
79
95
.
Chhinh
N.
&
Millington
A.
2015
Drought monitoring for rice production in Cambodia
.
Climate
3
(
4
),
792
811
.
Chim
K.
,
Tunnicliffe
J.
,
Shamseldin
A.
&
Bun
H.
2021
Assessment of land use and climate change effects on hydrology in the upper Siem Reap River and Angkor Temple Complex, Cambodia
.
Environmental Development
39
,
100615
.
François
B.
,
Schlef
K. E.
,
Wi
S.
&
Brown
C.
2019
Design considerations for riverine floods in a changing climate@ a review
.
Journal of Hydrology
574
,
557
573
.
Fujimoto
T.
&
Tomotsugu
S.
2019
The civil engineers’ unfinished business: Japan's commitment to the development of the Cambodian Prek Thnot Project
.
広島平和科学
41
,
51
71
.
Gassman
P. W.
,
Reyes
M. R.
,
Green
C. H.
&
Arnold
J. G.
2007
The soil and water assessment tool: historical development, applications, and future research directions
.
Transactions of the ASABE
50
(
4
),
1211
1250
.
Gosling
S. N.
&
Arnell
N. W.
2016
A global assessment of the impact of climate change on water scarcity
.
Climatic Change
134
(
3
),
371
385
.
Heng
B.
,
Oeurng
C.
,
Try
S.
&
Yuzir
A.
2020
Flow regime alteration analysis under climate change in Tonle Sap Subbasin
. In
Paper Presented at the IOP Conference Series: Earth and Environmental Science
.
IPCC
2013
Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change
.
Cambridge University Press
,
Cambrige
,
UK
.
New York, NY, USA
.
IPCC
2014
Climate Change 2014: Impacts, Adaptation, and Vulnerability. Part A: Global and Sectoral Aspects. Contribution of Working Group II to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change: Intergovernmental Panel on Climate Change
.
Islam
M.
,
Agata
Y.
,
Hanasaki
N.
,
Kanae
S.
&
Oki
T.
2005a
Global water resources under future changes: toward an improved estimation
. In
Paper Presented at the AGU Spring Meeting Abstracts
.
Islam
M.
,
Oki
T.
,
Kanae
S.
,
Hanasaki
N.
,
Agata
Y.
&
Yoshimura
K.
2007a
A grid-based assessment of global water scarcity including virtual water trading
.
Water Resources Management
21
(
1
),
19
33
.
JICA
2012
Preparatory Survey for Irrigation and Drainage System Rehabilitation and Improvement Project in the Kingdom of Cambodia, Ministry of Water Resources and Meteorology, Cambodia. Final Report, Volume III
.
Retrieved from
.
Kayiranga
A.
,
Chen
B.
,
Guo
L.
,
Measho
S.
,
Hirwa
H.
,
Liu
S.
,
Bofana
J.
,
Sun
S.
,
Wang
F.
&
Karamage
F.
2021
Spatiotemporal variations of forest ecohydrological characteristics in the Lancang-Mekong region during 1992–2016 and 2020–2099 under different climate scenarios
.
Agricultural and Forest Meteorology
310
,
108662
.
Khorn
N.
,
Ismail
M. H.
,
Kamarudin
N.
&
Nurhidayu
S.
2020a
Land use change using geospatial techniques in Upper Prek Thnot Watershed in Cambodia
.
Pertanika Journal of Science & Technology
28
(
3
),
879
892
.
Khorn
N.
,
Ismail
M. H.
,
Nurhidayu
S.
,
Kamarudin
N.
&
Sulaiman
M. S.
2020b
Impact of land use/land cover change on runoff using SWAT modelling: a case study in Upper Prek Thnot Watershed in Cambodia
.
Authorea Preprints
.
Koem
C.
&
Tantanee
S.
2020
Flash flood hazard mapping based on AHP with GIS and satellite information in Kampong Speu Province, Cambodia
.
International Journal of Disaster Resilience in the Built Environment
12
(
5
),
457
470
.
Ligaray
M.
,
Kim
H.
,
Sthiannopkao
S.
,
Lee
S.
,
Cho
K. H.
&
Kim
J. H.
2015
Assessment on hydrologic response by climate change in the Chao Phraya River Basin, Thailand
.
Water
7
(
12
),
6892
6909
.
Martínez-Retureta
R.
,
Aguayo
M.
,
Abreu
N. J.
,
Stehr
A.
,
Duran-Llacer
I.
,
Rodríguez-López
L.
,
Sauvage
S.
&
Sánchez-Pérez
J.-M.
2021
Estimation of the climate change impact on the hydrological balance in Basins of South-Central Chile
.
Water
13
(
6
),
794
.
MoE
2001
Vulnerability and Adaptation Assessment to Climate Change in Cambodia
.
Ministry of Environment
,
Phnom Penh
,
Cambodia
.
Retrieved from
.
Mohammed
K.
,
Islam
A. S.
,
Islam
G. T.
,
Alfieri
L.
,
Bala
S. K.
&
Khan
M. J. U.
2017
Extreme flows and water availability of the Brahmaputra River under 1.5 and 2 C global warming scenarios
.
Climatic Change
145
(
1–2
),
159
175
.
Moriasi
D. N.
,
Gitau
M. W.
,
Pai
N.
&
Daggupati
P.
2015
Hydrologic and water quality models: performance measures and evaluation criteria
.
Transactions of the ASABE
58
(
6
),
1763
1785
.
MRC
2015
1st Draft Report on Defining Basin-Wide Climate Change Scenarios for the Lower Mekong Basin
.
Mekong River Commission (MRC)
,
Vientiane
,
Laos
.
Retrieved from
.
MRC
2017
Summary of the Basin-Wide Assessments of Climate Change Impacts on Water and Waterrelated Resources in the Lower Mekong Basin
.
Nathan
R.
&
McMahon
T.
1990
Practical aspects of low-flow frequency analysis
.
Water Resources Research
26
(
9
),
2135
2141
.
Neitsch
S. L.
,
Arnold
J. G.
,
Kiniry
J. R.
&
Williams
J. R.
2011
Soil and Water Assessment Tool Theoretical Documentation Version 2009
.
Texas Water Resources Institute
,
College Station, TX
.
Oeurng
C.
,
Cochrane
T. A.
,
Arias
M. E.
,
Shrestha
B.
&
Piman
T.
2016
Assessment of changes in riverine nitrate in the Sesan, Srepok and Sekong tributaries of the Lower Mekong River Basin
.
Journal of Hydrology: Regional Studies
8
,
95
111
.
Oeurng
C.
,
Cochrane
T. A.
,
Chung
S.
,
Kondolf
M. G.
,
Piman
T.
&
Arias
M. E.
2019
Assessing climate change impacts on river flows in the Tonle Sap Lake Basin, Cambodia
.
Water
11
(
3
),
618
.
Pilon
P.
&
Harvey
K.
1994
Consolidated Frequency Analysis
.
Reference manual, Environment Canada
,
Ottawa
,
Canada
.
Pilon
P.
&
Jackson
R.
1988
Low Flow Frequency Analysis Package–LFA
.
Environment Canada
,
Ottawa
.
Pradhanang
S. M.
,
Mukundan
R.
,
Schneiderman
E. M.
,
Zion
M. S.
,
Anandhi
A.
,
Pierson
D. C.
,
Frei
A.
,
Easton
Z. M.
,
Fuka
D.
&
Steenhuis
T. S.
2013
Streamflow responses to climate change: analysis of hydrologic indicators in a New York City water supply watershed
.
JAWRA Journal of the American Water Resources Association
49
(
6
),
1308
1326
.
Reshmidevi
T.
,
Kumar
D. N.
,
Mehrotra
R.
&
Sharma
A.
2018
Estimation of the climate change impact on a catchment water balance using an ensemble of GCMs
.
Journal of Hydrology
556
,
1192
1204
.
Riahi
K.
,
Krey
V.
,
Rao
S.
,
Chirkov
V.
,
Fischer
G.
,
Kolp
P.
,
Kindermann
G.
,
Nakicenovic
N.
&
Rafai
P.
2011
RCP 8.5: A scenario of comparatively high greenhouse gas emissions
.
Climatic change
109
(
1
),
33
57
.
Sagara
J.
2021
Surface Water Resources Assessment of the Tonle Sap and Mekong Delta River Basin Groups: Improving Climate Resilience, Productivity, and Sustainability (2071–7202)
.
Shawul
A. A.
,
Alamirew
T.
&
Dinka
M.
2013
Calibration and validation of SWAT model and estimation of water balance components of Shaya mountainous watershed, Southeastern Ethiopia
.
Hydrology and Earth System Sciences Discussions
10
(
11
),
13955
13978
.
Shrestha
B.
,
Cochrane
T. A.
,
Caruso
B. S.
,
Arias
M. E.
&
Piman
T.
2016
Uncertainty in flow and sediment projections due to future climate scenarios for the 3S Rivers in the Mekong Basin
.
Journal of Hydrology
540
,
1088
1104
.
Sok
T.
,
Oeurng
C.
,
Ich
I.
,
Sauvage
S.
&
Miguel Sánchez-Pérez
J.
2020
Assessment of hydrology and sediment yield in the Mekong River Basin using SWAT model
.
Water
12
(
12
),
3503
.
Sothea
K.
,
Sovann
S.
,
Sothy
H.
,
Sakhon
P.
&
Samseiha
U.
2019
Assessment of hydrology for agricultural development based on climate change impacts in Prek Thnot River Basin, Cambodia
.
Development and Climate Change in the Mekong Region
10
,
215
232
.
SRTM
2015
Source of the 30 m Resolution, ASTER-GDEM (Advanced Spaceborne Thermal Emission and Reflection Radiometer Global Digital Elevation Model)
.
doi:earthexplorer.usgs.gov
.
Stocker
T.
2014
Climate Change 2013: the Physical Science Basis: Working Group I Contribution to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change
.
Cambridge University Press, Cambridge
.
Tan
M. L.
,
Gassman
P. W.
,
Srinivasan
R.
,
Arnold
J. G.
&
Yang
X.
2019
A review of SWAT studies in Southeast Asia: applications, challenges and future directions
.
Water
11
(
5
),
914
.
Tan
M. L.
,
Juneng
L.
,
Tangang
F. T.
,
Chung
J. X.
&
Radin Firdaus
R.
2021a
Changes in temperature extremes and their relationship with ENSO in Malaysia from 1985 to 2018
.
International Journal of Climatology
41
,
E2564
E2580
.
Taylor
K. E.
,
Stouffer
R. J.
&
Meehl
G. A.
2012
An overview of CMIP5 and the experiment design
.
Bulletin of the American Meteorological Society
93
(
4
),
485
498
.
UNDP
2019
New Report Provides Foundation for Better Flood Management in Cambodia's Prek Thnot River Basin
.
Van Vuuren
D. P.
,
Edmonds
J.
,
Kainuma
M.
,
Riahi
K.
,
Thomson
A.
,
Hibbard
K.
,
Hurtt
G. C.
,
Kram
T.
,
Krey
V.
,
Lamarque
J.-F.
,
Masui
T.
,
Meinshausen
M.
,
Nakicenovic
N.
,
Smith
S. J.
&
Rose
S. K.
2011
The representative concentration pathways: an overview
.
Climatic Change
109
(
1
),
5
.
doi:10.1007/s10584-011-0148-z
.
Wang
S.
,
Zhang
L.
,
She
D.
,
Wang
G.
&
Zhang
Q.
2021
Future projections of flooding characteristics in the Lancang-Mekong River Basin under climate change
.
Journal of Hydrology
602
,
126778
.
Xu
F.
,
Zhou
Y.
&
Zhao
L.
2021
Spatial and temporal variability in extreme precipitation in the Pearl River Basin, China from 1960 to 2018
.
International Journal of Climatology
42
(
2
),
797
816
.
Yun
X.
,
Tang
Q.
,
Li
J.
,
Lu
H.
,
Zhang
L.
&
Chen
D.
2021
Can reservoir regulation mitigate future climate change induced hydrological extremes in the Lancang-Mekong River Basin?
Science of The Total Environment
785
,
147322
.
Zhang
Y.
,
You
Q.
,
Chen
C.
&
Ge
J.
2016
Impacts of climate change on streamflows under RCP scenarios: a case study in Xin River Basin, China
.
Atmospheric Research
178
,
521
534
.
This is an Open Access article distributed under the terms of the Creative Commons Attribution Licence (CC BY 4.0), which permits copying, adaptation and redistribution, provided the original work is properly cited (http://creativecommons.org/licenses/by/4.0/).