The Srepok watershed in the Central Highland of Vietnam plays an important role in the economic development of the region. Any harmful effects of climate change on natural resources may cause difficulties for social and economic development in this area. The present study aims to predict and evaluate changes of water resources in the Srepok watershed under the impact of climate change scenarios by using the soil and water assessment tool (SWAT) model. The study used observed weather data from 1990 to 2010 for the first period and climate change scenarios A1B and A2 from 2011 to 2039 for the second period and from 2040 to 2069 for the third period. According to the climate change scenarios of the studied watershed, future minimum and maximum daily average temperature will rise in all climate change scenarios and the amount of annual precipitation will fall in scenario A1B and go up in scenario A2. Based on the simulation results, the annual water discharge in scenario A1B decreased by 11.1% and 1.2% during the second and third periods, respectively, compared with the first. In scenario A2, annual water discharge increased by 2.4% during the second period but decreased by 1.8% during the third period.

INTRODUCTION

Climate change has many negative impacts related to natural resources, especially water. This phenomenon leads to changes in precipitation and temperature, which affects the hydrological cycle, and thus changes the streamflow and modifies the transformation and transport characteristics of sediment as well as water pollutants (Tu 2009). A number of studies have been performed on the impact of climate changes on hydrology (Zhang et al. 2007; Kim & Kaluarachchi 2009; Boyer et al. 2010; Bauwens et al. 2011; Al-mukhtar et al. 2014; Teferi et al. 2015). Understanding the responses of hydrological processes to climate changes is important for planning and managing water resources (Zhang et al. 2008).

Water resources depend strongly on weather conditions especially rainfall and temperature (National Research Council (USA) 1998). A study showed that precipitation and temperature play a critical role in the water discharge of the Zamu River basin (Wang et al. 2008). In addition, the impact of climate conditions on surface hydrology should be considered and assessed separately (Li et al. 2009). In general, regional impacts of climate change on hydrology vary from place to place and need to be considered at a local scale (Wang et al. 2013).

Outputs of general circulation models (GCMs) are used to generate the future climate conditions for a study area. Hydrological models have proven effective and can be used to assess the influence of climate variability on water availability (Li et al. 2012). The soil and water assessment tool (SWAT) hydrological model can be used to estimate climate change impact on runoff behavior. For instance, SWAT was used to evaluate the impact of climate change on water resources in the Songhua river basin, the Maumee river watershed and the Zarqa River Basin (Cousino et al. 2015; Hammouri et al. 2015; Abbasa et al. 2016; Li et al. 2016). These studies showed worsening water availability in the future under the impact of climate change (Abbasa et al. 2016). According to Li et al. (2016), the streamflow in the Nenjiang river basin and lower Songhua river basin would be 20.3–37.8% lower than baseline conditions (1980–2009). Cousino et al. (2015) showed that, compared to the baseline scenario (1985–2004), climate change scenarios reduced annual flow from 10% to 24% in the Maumee River watershed. Hammouri et al. (2015) reported that during the main rainy months in Jordan there would be a 40% decrease in surface runoff volume as a result of increasing temperature and decreasing precipitation.

Vietnam has experienced climate change, such as rising air temperatures and variable precipitation. From 1958 to 2007, the average annual temperature increased by 0.5–0.7°C (MONRE 2009). Annual precipitation decreased in Northern Vietnam but increased in Southern Vietnam, and for the entire country, rainfall has decreased on average by 2% over the past 50 years (1958–2007) (MONRE 2009). These changes have significantly affected the availability of water resources in Vietnam. Most studies related to climate change in Vietnam are primarily on climate change scenarios (MONRE 2009) or on outputs from individual GCMs. For example, Kawasaki et al. (2010) used the output from the Japanese Meteorological Agency GCM for IPCC SRES A1B scenario and the hydrological model HEC-HMS (Hydrologic Modeling System) to consider climate change impact on water resources in the Central Highland of Vietnam, and Thai & Thuc (2011) used the MIKE 11-NAM hydrological model and climate change scenarios from the Vietnam Ministry of Natural Resources and Environment (MONRE 2009). That data was downscaled from GCMs by the MAGICC/SCENGEN model to evaluate the impact of climate change on flow in the Hong-Thai Binh (located in North Delta) and Dong Nai river basins (located in the Central Highland and South Delta). The climate scenarios from MONRE (2009) were developed only for Vietnam's seven climate zones with low spatial resolution: Northwest, Northeast, North Delta (Red River Delta), North Central Coast, South Central Coast, Central Highlands and South Delta (Mekong River Delta). Therefore, it is not possible accurately to reflect the specific local details of climate change in Vietnam (MONRE 2010).

Methods of assessing hydrological effects of environmental change include field survey, paired catchment, statistical analysis and hydrological modeling (Li et al. 2009, 2012). Among these approaches, the hydrological modeling method is the most suitable for use in scenario studies. There are many hydrological models which are widely used to assess the impacts of climate changes on hydrology, including the Hydrologic Simulation Program–Fortran, the SWAT, WaTEM/SEDEM and the Water Erosion Prediction Project. The SWAT model was selected for the present study because it is widely used to assess hydrology and water quality in agricultural catchments around the world. Another reason for its selection is its availability and user-friendliness in handling input data (Arnold et al. 1998).

The Srepok river basin, a sub-basin in the Lower Mekong was selected as the study area. Within Vietnam, its basin area is 18,000 km2 and is distributed among four provinces: Gia Lai, DakLak, DakNong and Lam Dong (Figure 1) (Ha 2011). The watershed has two seasons: dry and rainy. The annual rainfall of the basin is 2,112 mm, and average daily maximum and minimum temperatures are 28.7°C and 19.8°C respectively. In the Srepok watershed, the forest covers about 9,720 km2 of the watershed area and agricultural land about 5,060 km2. The agricultural land is mainly under coffee, rubber, vegetables and rice paddy. Currently, there are many critical issues for water resource management in the basin (The Government of Vietnam 2006). These problems range from hydrological variability (including floods and droughts) to environmental degradation (including pollution of waterways and deforestation of catchments), over-exploitation of groundwater, conflicts over the use of water for different purposes and trans-boundary conflicts (inter-provincial and inter-district). So far, few studies have quantified the potential impacts of climate change on hydrology in the Srepok watershed (Van Ty et al. 2012; Khoi 2013). However, comprehending climate change impacts on hydrological conditions is essential to enable more efficient water resources development and to make suitable adaptation plans in this region.
Figure 1

Location of the Srepok watershed.

Figure 1

Location of the Srepok watershed.

The overall objective of the current project is to investigate changes in streamflow and hydrological processes resulting from climatic variation in the Srepok watershed, which is located in the DakLak and DakNong provinces, Vietnam. The specific objectives are: (1) to set up, calibrate and validate the SWAT model in terms of streamflow; (2) to simulate responses of streamflow and hydrological components under climate change scenarios.

MATERIALS AND METHODS

Data collection

In this study, all data and information related to the SWAT model in the Srepok watershed were collected. SWAT requires a spatial dataset including a digital elevation model (DEM), land use and soil maps. Meteorological data, such as daily precipitation, maximum and minimum air temperature, are also required. However, the number of local weather stations is limited and the spatial distribution of the four local weather stations did not strongly represent the watershed. Thus, in addition to locally acquired data, the data from 12 global weather stations were downloaded and used in this research (see below). For the calibration and validation of streamflow simulation, observed water discharge data from 1990 to 2010 was used. Climate change data were downloaded from the Southeast Asia START Regional Center's website. Table 1 displays the sources and types of data collected.

Table 1

Sources and types of data collected for SWAT simulation

Data type Description Sources 
Topography map DEM with resolution 30 m Department of Natural Resources and Environment Dak Lak and Dak Nong provinces ASTER Global DEM 
Land use map (2010) Land use with resolution 30 m Department of Natural Resources and Environment Dak Lak and Dak Nong provinces Global Land Cover Characterization (http://www.globallandcover.com/GLC30Download/index.aspx
Soil map (2005)  Department of Natural Resources and Environment Dak Lak and Dak Nong provinces 
Weather (1990–2010) Daily precipitation, minimum and maximum temperatures, relative humidity, wind speed, solar radiation The Central Highland Region Hydro-Meteorological Centre, Daily Climate Forecast System Reanalysis (CFSR) data (http://globalweather.tamu.edu/
Climate change data (2011–2069) SEASTART-AR4 GCMs A1B, A2 emission scenarios Southeast Asia START Regional Center (http://www.start.or.th/
Water discharge Monthly observed water discharge, period 1990–2010 The Central Highland Region Hydro-Meteorological Centre 
Data type Description Sources 
Topography map DEM with resolution 30 m Department of Natural Resources and Environment Dak Lak and Dak Nong provinces ASTER Global DEM 
Land use map (2010) Land use with resolution 30 m Department of Natural Resources and Environment Dak Lak and Dak Nong provinces Global Land Cover Characterization (http://www.globallandcover.com/GLC30Download/index.aspx
Soil map (2005)  Department of Natural Resources and Environment Dak Lak and Dak Nong provinces 
Weather (1990–2010) Daily precipitation, minimum and maximum temperatures, relative humidity, wind speed, solar radiation The Central Highland Region Hydro-Meteorological Centre, Daily Climate Forecast System Reanalysis (CFSR) data (http://globalweather.tamu.edu/
Climate change data (2011–2069) SEASTART-AR4 GCMs A1B, A2 emission scenarios Southeast Asia START Regional Center (http://www.start.or.th/
Water discharge Monthly observed water discharge, period 1990–2010 The Central Highland Region Hydro-Meteorological Centre 

Topography, land use and soil data were collected from local government offices and from global organizations, such as ASTER Global DEM, Global Land Cover Characterization and Southeast Asia START Regional Center. Meteorological data for 1990–2010 was used. The model used meteorological data from four local stations namely Buon Ho, M'Drak, Buon Ma Thuot and Dak Nong and twelve global weather stations in Srepok watershed. For the calibration and validation model, the project uses hydrological data for Ban Don Station. The period of calibration is from 1990 to 2000 and that of validation is from 2001 to 2009. From 2010 four reservoirs have been operating. Thus, the study does not use the observed data for 2010. Furthermore, because of lack of operation data for these reservoirs, they are not mentioned in the project. All data collected were processed in accordance with SWAT model input format.

Climate change data

This study used climate change data from the Southeast Asia START Regional Center. This climate change data is based on dynamic downscaling of global change scenarios generated by ECHAM4 GMC and use of the PRECIS regional climate model (RCM) (Chinvanno 2009). Basically, there are two types of technique for obtaining high resolution regional climate change projections, namely statistical and dynamical (Hachigonta 2011). The dynamical downscaling method has been widely used in previous climate change studies (Jones et al. 2004; Hachigonta 2011). A RCM is a downscaling model that achieves better resolution by using climate scenarios of GCMs. Like GCMs, RCMs were developed based on physical climate processes. While GCMs represent climate on a large scale, such as hundreds of kilometers, RCMs can analyze data down to the resolution of 50 km or less, thus providing a higher accuracy of climate data in a small area like the Srepok watershed. The climate change data was downscaled by using the RCM PRECIS from the Hadley Centre, Met Office, UK (Jones et al. 2004). According to some publications from the Southeast Asia START Regional Center website, an illustration of downscaling is a future climate scenario developed by Jiamjai Kreasuwan; this scenario is based on dynamic downscaling using the Fifth-Generation NCAR/Penn State Mesoscale Model (MM5) to simulate regional-scale atmospheric circulation. In addition, Suppakorn Chinvanno's research uses PRECIS RCM-downscaled data of the ECHAM4 GCM model, which is based on A1B and A2 scenarios (IPCC 2007). The downscaled data were rescaled on a 0.22° × 0.22° latitude–longitude grid (∼20 km). The climate change data are precipitation, maximum and minimum temperature, wind speed and solar radiation with daily time step throughout the period 2011–2069.

METHODOLOGY

The study uses climate change data through future periods (2011–2069) based on the PRECIS-downscaled GCMs driven by the A1B and A2 emission scenarios. The A1B and A2 scenarios were selected in this study because the study focuses on midcentury change, in which period the two scenarios exhibit similar greenhouse gases emission forcing. The A1B (medium emission) scenario projects a future where technology is shared between developed and developing nations in order to reduce regional economic disparities, while A2 (high emission) scenario assumes global population growth peaks by mid-century and then declines, a rapid economic shift towards service and information economies, and the introduction of clean and resource-efficient technologies (IPCC 2007).

A SWAT project was set up, calibrated and validated for baseline period (1990–2010) scenario. In order to investigate the impacts of climate changes scenarios on the streamflow, the future prediction separated the A1B and A2 emission scenarios into two future periods (2011–2039 and 2040–2069). Then, the study used validated model parameters runs of the SWAT model for all scenarios. Finally, the model output of climate scenarios was compared to the baseline period scenario. Figure 2 shows the methodological approach applied to the Srepok watershed.
Figure 2

Framework of the research.

Figure 2

Framework of the research.

This study uses SWAT CUP – a computer program for calibration of SWAT models as a supporting tool to search for appropriate values for each parameter. SUFI2 (Sequential Uncertainty Fitting algorithm) is one of five uncertainty analysis procedures of SWAT CUP program described by a multivariate uniform distribution in a parameter hypercube; the output uncertainty is quantified by the 95% prediction uncertainty band (95PPU) calculated at the 2.5% and 97.5% levels of the cumulative distribution function of the output variables (Abbaspour 2015).

Brief description of the SWAT model

SWAT is a physically based semi-distributed hydrological model which was developed by the Blackland Research and Extension Center and the United States Department of Agriculture–Agricultural Research Service (USDA-ARS). This model can be applied at the river basin scale to simulate the impact of land management practices on water, sediment and agrochemical yields in large watersheds with varying soils, land use and agricultural conditions over extended periods of time (Arnold et al. 1998). While specialized processes can be simulated if sufficient input data are available, SWAT also runs with minimum data inputs, which is advantageous when working in areas with limited data. SWAT is a continuous time model able to simulate long-term impacts of land use, land management practices and buildup of pollutants (Neitsch et al. 2005). These qualities of the SWAT model aid in the quantification of long-term impacts of land use changes and variations in rainfall and air temperature on the hydrology of the Srepok basin.

SWAT provides several options when simulating hydrological processes, which the user can choose based on their data availability. For example, the surface runoff volume from hydrologic response units can be simulated with the Soil Conservation Service (SCS) curve number method (USDA-SCS 1972) or the Green–Ampt infiltration method (Green & Ampt 1911). SWAT simulates the hydrology of the watershed in two phases: (1) the land phase, which controls the amount of water, sediment, nutrient and pesticide loadings to the main channel in each sub-basin; and (2) the water or routing phase, which controls the movement of water, sediment, nutrient and pesticide loadings through the channel network of the watershed into the outlet.

RESULTS AND DISCUSSION

Climate change trend in Srepok watershed

The climate change data show that not only does the maximum daily average temperature increase but also the minimum daily average temperature increases in the period from 2011 to 2069 in both scenarios A1B and A2 compared to base scenario (Figures 3 and 4). Scenario A1B (medium emissions scenario) has the maximum daily average temperature as 29.0°C which increases by 0.3°C from 2011 to 2039. The minimum daily average temperature, which is 20.6°C, increases by 0.8°C. In the period between 2040 and 2069, the maximum and minimum daily average temperatures are nearly 30.3°C and 21.9°C, increases of 1.6°C and 2°C respectively. For scenario A2 (high emission scenario), it is not difficult to realize the increase of the daily average temperature in the two periods. From 2011 to 2039, the maximum and minimum daily average temperatures increase by 2.8°C and 2.3°C respectively. Between 2040 and 2069, the maximum and minimum daily average temperatures increase to nearly 32.7°C and 23.3°C, i.e. they grow dramatically by 4°C and 3.5°C, respectively. Evapotranspiration can be strongly affected by temperature. Consequently, more water will be lost by evaporation in scenario A2 than scenario A1B.
Figure 3

Change of daily temperature in (a) period 1990–2010 of base scenario; (b) period 2011–2039 and (c) period 2040–2069 of A1B scenario.

Figure 3

Change of daily temperature in (a) period 1990–2010 of base scenario; (b) period 2011–2039 and (c) period 2040–2069 of A1B scenario.

Figure 4

Change of daily temperature in (a) period 1990–2010 of base scenario; (b) period 2011–2039 and (c) period 2040–2069 of A2 scenario.

Figure 4

Change of daily temperature in (a) period 1990–2010 of base scenario; (b) period 2011–2039 and (c) period 2040–2069 of A2 scenario.

The annual precipitation of A1B and A2 scenarios changed in terms of the amount and spatial distribution in the whole watershed compared to the base scenario. One interesting thing about the annual precipitation of scenarios A1B and A2 is that they show opposite trends. While the annual precipitation of scenario A1B decreases by 13.3% and 4.7% from 2011 to 2039 and 2040 to 2069, respectively, the annual precipitation of scenario A2 increases by 5.9% in the period 2011–2039 and 2.3% in the period 2040–2069. Some previous studies have different annual precipitation trends and time period of the base scenario compared to this study in the Srepok watershed (Table 2). Thus, the increase or decrease of annual precipitation of climate change scenarios depend on the climate model and time period of the base scenario.

Table 2

The different annual rainfall trend of climate change scenarios in some studies

Reference Time period of base scenario Climate change scenario
 
A1B
 
A2
 
B1
 
2020s 2050s 2020s 2050s 2020s 2050s 
This study 1990–2010 −13.3% −4.7% +5.9% +2.3% +5.9% +2.3% 
Khoi (2013)  1980–2000 −4.4% −2.9% N/A N/A −3.9% −4.3% 
MONRE (2012)  1980–1999 +1% +3% +2% +4% +1% +1.5% 
Reference Time period of base scenario Climate change scenario
 
A1B
 
A2
 
B1
 
2020s 2050s 2020s 2050s 2020s 2050s 
This study 1990–2010 −13.3% −4.7% +5.9% +2.3% +5.9% +2.3% 
Khoi (2013)  1980–2000 −4.4% −2.9% N/A N/A −3.9% −4.3% 
MONRE (2012)  1980–1999 +1% +3% +2% +4% +1% +1.5% 

+: Increase; − Decrease.

Moreover, scenario A2 shows clearly that the precipitation decreases in the dry season and increases in the rainy season (Figure 5). The trend of monthly rainfall of scenario A1B is almost the same as in the base scenario, that is high rainfall from May to October and low rainfall from November to April. The trend of monthly rainfall of scenario A2 is also similar to the base scenario but the duration of the rainy season is shorter from May to September. There is a large difference in monthly rainfall between scenarios A1B and A2. The annual precipitation of scenario A2 is 27% higher than scenario A1B in the period 2011–2039 and 7% higher in the period 2040–2069. As a result, the total annual discharge of scenario A2 will be higher than scenario A1B. The monthly rainfall between A1B and A2 may be different because A1B, a medium emission scenario, has less effect on each season whereas A2, a high emission scenario, produces stronger effects leading to more change to the season.
Figure 5

The average monthly precipitation for the different climate scenarios.

Figure 5

The average monthly precipitation for the different climate scenarios.

Calibration of the SWAT model in Srepok watershed

First, the SWAT model is set up to simulate water discharge with the land cover in 2014 (base scenario). This model simulates water discharge from 1990 to 2010 using measured discharge at Ban Don Station on the Srepok River to calibrate the simulated streamflow. The period of calibration is from 1990 to 2000 (the period 1990–1992 was skipped to warm up the model) and validation is from 2001 to 2009.

It is difficult for researchers to find out the best calibration parameters, which fit the special characteristics of the study area. Fortunately, the processing was implemented quickly by automatic calibration with the SUFI2 algorithm in SWAT-CUP software. Five parameters were determined which influenced the water quantity most in the Srepok watershed: curve number (CN2), base flow alpha factor (ALPHA_BF), groundwater delay (GW_DELAY), threshold water depth in the shallow aquifer for flow (GWQMN) and available capacity of the soil layer (SOL_AWC); the optimal values are shown in Table 3.

Table 3

SWAT sensitive parameters and calibrated values

Parameter Description of parameter Calibrated value
 
Fitted value New minimum threshold value New maximum threshold value 
r_CN2. Initial SCS CN II value −0.59 −0.6 −0.4 
v_ALPHA_BF Base flow alpha factor 0.420 0.4 0.6 
v_GW_DELAY Groundwater delay 58.25 57 59 
v_GWQMN Threshold water depth in the shallow aquifer for flow 0.80 0.8 0.9 
r_SOL_AWC Available capacity of the soil layer 2.17 1.5 2.5 
Parameter Description of parameter Calibrated value
 
Fitted value New minimum threshold value New maximum threshold value 
r_CN2. Initial SCS CN II value −0.59 −0.6 −0.4 
v_ALPHA_BF Base flow alpha factor 0.420 0.4 0.6 
v_GW_DELAY Groundwater delay 58.25 57 59 
v_GWQMN Threshold water depth in the shallow aquifer for flow 0.80 0.8 0.9 
r_SOL_AWC Available capacity of the soil layer 2.17 1.5 2.5 

Using these calibration parameters, the Nash–Sutcliffe index and the coefficient of determination reached a satisfactory level with R2 = 0.81, NS = 0.76 in calibration and R2 = 0.82, NS = 0.59 in validation, respectively, as is also indicated by the monthly average simulated discharge series which follow the observed series closely (Figure 6).
Figure 6

Observed and simulated water discharge in the calibration period 1992–2000 and validation period 2001–2009 at Ban Don Station.

Figure 6

Observed and simulated water discharge in the calibration period 1992–2000 and validation period 2001–2009 at Ban Don Station.

Assessment impact of climate change on streamflow and hydrological components

Results from the simulations showing the monthly average water discharges corresponding to weather data in the period 1990–2010 for the base scenario and two periods 2011–2039 and 2040–2069 for scenarios A1B and A2 are listed in Table 4. From 1990 to 2010, the monthly average flow discharge reaches the highest value in September (881.13 m3/s) and the lowest in March (119.84 m3/s). High flow rates over 500 m3/s occur from July to November. In the period 2011–2039, water discharge in scenario A1B reaches its peak in September and a high level from September to November; in scenario A2, streamflow also peaks in September and reaches a high level from May to September. For the final prediction period, the streamflow of scenarios A1B and A2 have similar trends to those of the previous period, but streamflow reaches its peak in October for scenario A1B.

Table 4

Monthly average streamflow of base scenario and climate change scenarios

Month Average monthly streamflow (m3/s)
 
Base (1990–2010) A1B (2011–2039) A1B (2040–2069) A2 (2011–2039) A2 (2040–2069) 
Jan 241.29 327.46 351.96 299.28 295.80 
Feb 160.50 285.01 314.11 277.74 268.70 
Mar 119.84 253.19 278.87 251.06 239.28 
Apr 157.28 239.85 263.62 292.86 295.87 
May 358.69 305.21 334.03 559.14 545.73 
Jun 459.50 416.69 470.51 736.08 695.30 
Jul 576.61 478.47 527.24 646.53 642.10 
Aug 794.33 462.03 539.10 722.66 704.53 
Sep 881.13 629.98 694.69 772.09 773.18 
Oct 820.94 600.13 725.64 474.04 440.54 
Nov 680.12 622.10 672.70 444.84 357.36 
Dec 440.77 440.96 450.41 352.28 328.51 
Average 474.25 421.76 468.57 485.72 465.58 
Month Average monthly streamflow (m3/s)
 
Base (1990–2010) A1B (2011–2039) A1B (2040–2069) A2 (2011–2039) A2 (2040–2069) 
Jan 241.29 327.46 351.96 299.28 295.80 
Feb 160.50 285.01 314.11 277.74 268.70 
Mar 119.84 253.19 278.87 251.06 239.28 
Apr 157.28 239.85 263.62 292.86 295.87 
May 358.69 305.21 334.03 559.14 545.73 
Jun 459.50 416.69 470.51 736.08 695.30 
Jul 576.61 478.47 527.24 646.53 642.10 
Aug 794.33 462.03 539.10 722.66 704.53 
Sep 881.13 629.98 694.69 772.09 773.18 
Oct 820.94 600.13 725.64 474.04 440.54 
Nov 680.12 622.10 672.70 444.84 357.36 
Dec 440.77 440.96 450.41 352.28 328.51 
Average 474.25 421.76 468.57 485.72 465.58 

Under the impact of climate change scenarios, the annual discharge is estimated to change during the two future periods. The streamflow changes slightly among these scenarios as the volumes in the various climate change scenarios. For scenarios A1B, the decrease of annual discharge is 11.1% and 1.2% in the periods 2011–2039 and 2040–2069, respectively. Decreasing streamflow can be attributed to decreasing precipitation and increasing temperature. This decrease in streamflow is quite consistent with streamflow results in previous studies, such as Khoi (2013) and Van Ty et al. (2012). For scenario A2, in the period 2011–2039, the annual discharge increases by 2.4% but it decreases by 1.8% in the period 2040–2069. These inverse trends are because in the second period of scenario A2 the precipitation goes lower and temperature increases more than in the first period. More evaporation would be expected in the second period of scenario A2. In the first period of scenario A2, since streamflow response mainly depends on precipitation, the increase in precipitation and a slight increase in temperature resulted in streamflow increase compared to base scenario.

The assessment of hydrological components also plays an important role in watershed water planning not only at present but also in the future. The ratio of hydrological components does not change between the two future periods of climate change scenarios. Hence, they are evaluated for the long period from 2011 to 2069 for all climate change scenarios. The ratio streamflow: rainfall flow availability in the Srepok watershed is nearly 83% for 1990–2010. However, comparing the scenarios A1B and A2, the ratio water discharge: precipitation decreased minimally by factors of nearly 2% and 1%, respectively as shown in Figure 7. Regarding the contribution of total flow into watershed, simulation results show that groundwater is the highest contributor in the three scenarios (over 30%). Therefore, in the Srepok watershed, groundwater can provide a large portion of the total water available in the near future. These climate change scenarios showed opposite ratios for base flow and surface runoff in total flow. Thus, these results have demonstrated the fluctuating impact of climate change scenarios on water resources in the watershed.
Figure 7

Hydrological components ratios in Srepok watershed for the different scenarios.

Figure 7

Hydrological components ratios in Srepok watershed for the different scenarios.

CONCLUSION

The research used climate change data based on dynamic downscaling by ECHAM4 GMC and use of the PRECIS model. In the future, the study area would be warmer because of the increase in daily average maximum and minimum temperatures, whereas the precipitation decreases in the dry season and increases in the rainy season for all climate scenarios when compared with the base scenario.

The simulated streamflow was calibrated and validated with observed data from the Ban Don hydrological station for the period from 1990 to 2009, with the Nash–Sutcliffe index and the coefficient of determination attaining fair levels. The annual streamflow of the A1B scenario has a similar trend to the base scenario but decreasing quantities. Scenario A2 exhibited several changes in rainfall not only in terms of the quantity but also distributions in the dry and rainy season with a relatively high increase in average daily maximum and minimum temperature that may lead to a reduction of streamflow and an increase of annual precipitation in the period 2040–2069. From hydrological components results, the ratio streamflow: rainfall demonstrates that flow availability in Srepok watershed is still good and the amount of evapotranspiration is average. Regarding the contribution of the total flow in the watershed, groundwater plays a pivotal role and it is one of the most important water contributors in the Srepok watershed.

Therefore, in the near future, the Srepok watershed may experience shortfalls related to water discharge and subsequent imbalances in terms of water supply. In that case, it may affect the region's development, especially agricultural production. The results of this research will provide useful information that decision-makers need in order to promote water resources planning efforts in the Srepok watershed, Central Highland, Vietnam.

ACKNOWLEDGEMENTS

The authors acknowledge the SEARCA for funding this research. We would like to convey our special appreciation to the following people for their valuable contributions towards this study: Our students in the Geographic Information System Laboratory; Research Center for Climate Change (RCCC) – Nong Lam University (NLU) – Ho Chi Minh City, whose jokes always brought some smiles during tough times.

REFERENCES

REFERENCES
Abbaspour
K. C.
2015
CUP: SWAT Calibration and Uncertainty Programs: A User Manual. Available from: http://swat.tamu.edu/media/114860/usermanual_swatcup.pdf
.
Arnold
J. G.
Srinivasan
R.
Muttiah
R. S.
Williams
J. R.
1998
Large area hydrologic modeling and assessment part i: model development
.
J. Am. Water Resour. Assoc.
34
,
73
89
.
Boyer
C.
Chaumont
D.
Chartier
I.
Roy
A. G.
2010
Impact of climate change on the hydrology of St. Lawrence tributaries
.
J. Hydrol.
384
,
65
83
.
Chinvanno
S.
2009
Future climate projection for Thailand and surrounding countries: climate change scenario of 21st century
. In:
The First China-Thailand Joint Seminar on Climate Change
.
Southeast Asia START Regional Center
,
Bangkok
,
Thailand
, pp.
1
7
.
Cousino
L. K.
Becker
R. H.
Zmijewski
K. A.
2015
Modeling the effects of climate change on water, sediment, and nutrient yields from the Maumee River watershed
.
J. Hydrol. Stud.
4
,
762
775
.
Ha
P. T.
2011
Srepok river basin council and IWRM in the Srepok basin
. In:
The International Conference on Watershed Management
.
Chiang Mai, Thailand
.
Hachigonta
S. P.
2011
Assessing Maize Water Requirements in the Context of Climate Change Uncertains Over Southern Africa
.
University of Cape Town, South Africa
.
Hammouri
N.
Abdulla
F.
Abu
H. Q.
Freiwan
M.
2015
Assessing the impacts of climate change on water on water resources of Jordan
. In:
World Water Congress XV
,
Edinburgh, UK
.
IPCC
2007
The Physical Science Basin: Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change
.
IPCC
,
Cambridge
,
UK
and
New York
,
USA
.
Jones
R. G.
Noguer
M.
Hassell
D. C.
Hudson
D.
Wilson
S.
Jenkins
G.
Mitchell
J. F. B.
2004
Generating High Resolution Climate Change Scenarios Using PRECIS
.
Exeter
,
UK. Published by the United Nations Development Programme (UNDP)
.
Kawasaki
A.
Takamatsu
M.
He
J.
Rogers
P.
Herath
S.
2010
An integrated approach to evaluate potential impact of precipitation and land-use change on streamflow in Srepok River Basin
.
Theor. Appl. GIS
18
,
9
20
.
Khoi
D. N.
2013
Impacts of climate change on hydrology in the Srepok watershed, Vietnam
. In:
Climate and Land Surface Changes in Hydrology
(
Kunstmann
H.
Boegh
E.
Blyth
E.
Hannah
D. M.
Hisdal
H.
Su
B.
Yilmaz
K. K.
, eds).
International Association of Hydrological Sciences Series of Proceedings and Reports Publication No. 359
, pp.
111
117
.
Gothenburg, Sweden
.
Kim
U.
Kaluarachchi
J. J.
2009
Climate change impacts on water resources in the upper Blue Nile river basin, Ethiopia
.
J. Am. Water Resour. Assoc.
45
,
1361
1378
.
MONRE (Ministry of Natural Resources and Environment)
2009
Climate Change, Sea Level Rise Scenarios for Vietnam
.
Ha Noi
,
Vietnam
.
MONRE (Ministry of Natural Resources and Environment)
2010
Vietnam Second National Communication: Under the United Nations Framework Convention on Climate Change
.
MONRE (Ministry of Natural Resources and Environment)
2012
Climate Change, Sea Level Rise Scenarios for Vietnam
.
Ha Noi
,
Vietnam
.
National Research Council (USA)
1998
Climate attributes that influence society
. In:
Decade-to-Century-Scale Climate Variability and Change: A Science Strategy
.
National Academy Press
,
Washington, DC
, p.
160
.
Neitsch
S. L.
Arnold
J. G.
Kiniry
J. R.
Williams
J. R.
2005
Soil and Water Assessment Tool, Theoretical Documentation: Version 2005
.
Thai
T. H.
Thuc
T.
2011
Impacts of climate change on the flow in Hong-Thai Binh and Dong Nai river basins
.
VNU J. Sci. Earth Sci.
27
,
98
106
.
The Government of Vietnam
2006
A Strategic Framework for Water Resource Management in the SrePok River Basin. Available from: http://documentslide.com/documents/httpmouthtosourceorgriverssrepokfiles201009strategic-framework-for-wrm-in-sre-pok.html.
USDA-SCS
1972
National Engineering Handbook Section IV, Hydrology
.
US Government Printing Office, Washington, DC
.
Zhang
X.
Zhang
L.
Zhao
J.
Rustomji
P.
Hairsine
P.
2008
Responses of streamflow to changes in climate and land use/cover in the Loess Plateau, China
.
Water Resour. Res.
44
.
doi: 10.1029/2007WR006711
.