This study characterizes the hydrological regime of the Upper Ayeyarwaddy River Basin (UARB) of Myanmar under current and future climate change scenarios by using the Soil and Water Assessment Tool (SWAT). The model simulation results show that the annual precipitation, actual evapotranspiration and water yields are 1,578, 524 and 1,010 mm, respectively. These will increase by 13–28%, 11–24% and 42–198% under two representative concentration pathways (RCPs), RCP 4.5 and RCP 8.5, for the future. There is seasonal variability across the cool, hot and rainy seasons in the agro-ecological regions – mountains, hills and inland plains. As in other Asian regions, the model shows that the wet (rainy) season is becoming wetter and the dry (cool) season is becoming drier in the UARB too.

  • The use of distributed hydrological model to quantify the water availability.

  • The characterization of hydrological parameters of the Ayeyarwaddy River Basin.

  • The quantification of sub-basin water balances.

  • The quantification of agro-ecological water balances.

  • The impact of climate change on the water balances.

Graphical Abstract

Graphical Abstract
Graphical Abstract

Globally, the source for over half the world's extracted freshwater is from rivers (Taft & Kühle 2018). However, global river systems have undergone significant changes, particularly in their streamflow, which is mainly driven by anthropogenic activities like land-use change, deforestation, damming rivers, water diversions and abstractions, sand mining and, more recently, climate change impacts (Pandey et al. 2019; Sirisena et al. 2021).

Climate change is reconfiguring the water system and the direct impacts on water will be multiplied via the effects on other sectors in the water-energy-food-environment-livelihood nexus. The sixth assessment report of the Intergovernmental Panel on Climate Change in 2021 corroborated that human-induced climate change is already causing many weather and climate extremes worldwide and affecting the hydrological cycle and water availability. Therefore, steady, predictable seasonal water flows are unlikely to be maintained and year-to-year variations will continue to occur.

In Myanmar, the climate is projected to shift dramatically in the coming decades affecting the hydrological water cycle and flows in rivers (Kreft et al. 2015; Horton et al. 2016). Development and land-use changes in Myanmar including deforestation, construction of dams and increases in agricultural water use have also affected the quality and quantity of river flows (Kattelus et al. 2015; Taft & Evers 2016). The climate change will increase the severity, duration and frequency of droughts in the near future in the Lower Mekong river basin (Sam et al. 2019), which might be also the case for the rivers in Myanmar due to a similar hydroclimatic zone. In Mekong, there will be chances of water shortage in the dry seasons and soil degradation in the wet season due to climate change (Khoi & Thang 2017) and there will increase in annual river discharge and total suspended sediment load in the near future (Khoi et al. 2020).

Currently, 90% of the total available water in Myanmar is consumed by the agriculture sector, which employs 60% of the country's labour force and contributes to 22% of the country's gross domestic product (Ghimire et al. 2019). The 2014 Census indicated that 26% of all households countrywide lacked safe sanitation and 31% lacked access to improved drinking water (3.4 million households) (HARP-F & MIMU 2018). With the rapid increase in the country's population, planned reservoirs and projected growth in the economy and living standards, water consumption will most likely increase. To plan for future water demand, a detailed hydrological assessment in the context of climate change is imperative. Impacts of climate change are not spatially uniform and hence the hydrological cycle at the basin scale will vary from basin to basin and within large river basins. Such variations can cause significant alteration of the hydrological regime of a river basin (Sirisena et al. 2021).

A few studies have tried to assess the impact of climate change on the hydrology of the Upper Ayeyarwaddy and Chindwin River basins in Myanmar using various climate change scenarios (Sirisena et al. 2018, 2021; Ghimire et al. 2019, 2020; Oo et al. 2019, 2020). Sirisena et al. (2021) used the SWAT model to predict streamflow and sediment loads during mid-century (2046–2065) and end-century (2081–2100) periods under two scenarios with and without planned reservoirs in the Ayeyarwaddy River. The results showed that discharge at the basin outlet would increase by 8–17% and 9–45% under representative concentration pathway (RCP) 2.6 and RCP 8.5, respectively.

The study conducted by Sirisena et al. (2018) simulated future discharge in the UARB using the SWAT model and projected a future climate from 10 general circulation models (GCMs) from the 5th Coupled Model Intercomparison Project – Phase 5 (CMIP5) and predicted a decrease in annual streamflow up to 1.92, 7.29 and 11.92% under RCP 2.6, 4.5 and 8.5 scenarios, respectively (Oo et al. 2020). Another study by the same authors found precipitation to increase in rainy and cool seasons and decrease in hot seasons under all scenarios and temperature to steadily increase in all seasons and more in hot seasons (Oo et al. 2019).

Ghimire et al. (2019) used the bias-corrected daily precipitation data of eight GCMs to assess the impact on streamflow in the UARB using the Hydrologic Engineering Center Hydrologic Modelling System (HEC-HMS). The study used 1975–2014 as the baseline period, made comparisons with projected future periods until 2100 and discovered high variability in future low flows and higher means in future high flows than that of the baseline. The study suggests an increase in magnitude of mean flows and flooding events, but a decrease in variability.

While a handful of above studies have been conducted to assess the impact of climate change in Myanmar, a detailed assessment of spatial and temporal variation in hydrology within a river basin and subsequent changes in water balance components like precipitation, evapotranspiration, water yields and river discharges are yet to be made as the studies conducted to date have not described the water balance with sufficient spatial information. As the changes in water balance components and the overall water availability will differ from season to season, it is essential to capture the seasonal distribution to better plan for water management. This study, therefore, focused on the UARB to quantify and assess the past and projected future spatial (sub-basins and agro-ecological regions) and temporal (monthly and seasonal) water balances and the impact on river discharges.

Study area

The Ayeyarwaddy River, with a drainage area of 414,000 km2 (60% of the country), runs from the north to the south of the country into the Andaman Sea (Ghimire et al. 2020). One of the largest rivers in Southeast Asia, it is predominantly fed by monsoon precipitation and, to a lesser extent, by meltwater from snow and glaciers in the Himalaya Mountains (Taft & Kühle 2018). About 91% of the basin lies within Myanmar with some parts in India and China (Salmivaara et al. 2013). The 2,100 km long river is the most important commercial waterway in Myanmar and the area around its mouth near the Andaman sea forms one of the largest delta systems in Southeast Asia (Sirisena et al. 2018). The basin is often described as ungauged because of the scarcity of hydrometeorological data available for assessment (Chavoshian et al. 2007; Ghimire et al. 2020).

Figure 1 shows the river divided into the Chindwin and Upper and Lower Ayeyarwaddy River systems. This study focuses on the UARB which drains an area of around 337,400 km2 (of which 89% falls in Myanmar, 5% in India and 6% in China) to a discharge station called Magway located a little downstream from the confluence of the Chindwin and the Ayeyarwaddy. The study is part of a larger project which aims to assess the feasibility of extending aquaculture ponds in Myanmar. The study area was selected for two reasons, (i) the variation in topography in the UARB provides ideal conditions to develop new aquaculture ponds and (ii) most of the meteorological and hydrological data available for this study are concentrated in the UARB region. The study can be extended for the entire Ayeyarwaddy River basin to include the impact of sea water intrusion on freshwater availability in coastal regions, which was not included in this study but could be conducted by using robust groundwater modelling tools.
Figure 1

Map of the Upper Ayeyarwaddy River Basin. Source: Author's creation.

Figure 1

Map of the Upper Ayeyarwaddy River Basin. Source: Author's creation.

Close modal

The UARB covers six regions of Myanmar (Figure 1). Elevation of the study area varies from 29 metres above sea level (masl) in the south to 5,775 masl in the mountainous region in the north. The middle of the basin is comprised of plateaus (∼500 masl) and floodplains. Most of the study area belongs to the tropical monsoon climatic zone.

SWAT model and data

Spatial and temporal data for the SWAT model set-up were collected from the Hydro-Informatics Centre in Myanmar, International Water Management Institute (IWMI) archives and other sources (Tables 1 and 2). Spatial datasets included the Digital Elevation Model (DEM), Land-Use and Land-Cover map, Soil map and stream networks (Figure 2). Temporal data included precipitation data from 30 stations, maximum and minimum temperature data from 23 stations, relative humidity from 14 stations, wind speed from 14 stations and discharge data from 14 hydrological stations.
Table 1

Overview of spatial datasets

DataPropertiesSources
Digital Elevation Model 30 m resolution Shuttle Radar Topography Mission (SRTM) United States Geological Survey 
Land use/land cover 17 classes Chandrasekharan & Rajah (2017
Soil map 19 classes Food and Agriculture Organization of the United Nations (FAO) 
River Networks Shape of main river network and streams Ayeyarwaddy State of the Basin Assessment for main river network; DEM for streams 
DataPropertiesSources
Digital Elevation Model 30 m resolution Shuttle Radar Topography Mission (SRTM) United States Geological Survey 
Land use/land cover 17 classes Chandrasekharan & Rajah (2017
Soil map 19 classes Food and Agriculture Organization of the United Nations (FAO) 
River Networks Shape of main river network and streams Ayeyarwaddy State of the Basin Assessment for main river network; DEM for streams 
Table 2

Summary of climatic and hydrological data stations used for modelling

SNDataSources
(a) Climate data (daily time series) 
15 Precipitation stations  
12 Temperature stations 
9 Relative humidity stations 
9 Wind speed stations 
Solar radiation data was generated by the model using weather generator 
(b) Hydrological data (daily time series) 
4 river flow stations (Katha, Sagaing, Monywa and Magway) Hydro-Informatics Centre, Myanmar; four stations used for calibration and validation 
SNDataSources
(a) Climate data (daily time series) 
15 Precipitation stations  
12 Temperature stations 
9 Relative humidity stations 
9 Wind speed stations 
Solar radiation data was generated by the model using weather generator 
(b) Hydrological data (daily time series) 
4 river flow stations (Katha, Sagaing, Monywa and Magway) Hydro-Informatics Centre, Myanmar; four stations used for calibration and validation 
Figure 2

Spatial datasets used in the study. Source: Author's creation.

Figure 2

Spatial datasets used in the study. Source: Author's creation.

Close modal

Once the model was set up, a sensitivity analysis was conducted using the Sequential Uncertainty Fitting (SUFI-2) algorithm in SWAT Calibration and Uncertainty Program (SWAT-CUP) to determine the most sensitive parameters in the basin. For the final step, the model results were analysed in terms of water balances at daily scales in the entire study area and compared to model runs using climate change projection data to evaluate the possible impacts from modelled climate change under RCP 4.5 and RCP 8.5 scenarios in the 2030s (2021–2040) and 2050s (2041–2060). A more detailed description of the SWAT model, input and output data and the calibration and validation processes are described in the sections below.

Soil and water assessment tool (SWAT)

SWAT is a continuous physically based distributed river basin model used in simulating the quality and quantity of surface water and groundwater (Lévesque et al. 2008). The version of SWAT model used for this study was ArcSWAT2012.10.24. The model can predict the impacts of land use, land management practices and climate change on water availability. The major components of the model include climate, hydrology, soil temperature and properties, plant growth, nutrients, pesticides, bacteria and pathogens, land management, and channel and reservoir routing (Arnold et al. 2012; Bharati et al. 2019).

The model divides a basin into sub-basins, each of which is connected by a stream channel. It further divides sub-basins into Hydrological Response Units (HRUs), the smallest spatial unit of the model, based on a homogeneous combination of soil, land use and slope. The model simulates hydrology, vegetation growth and management practices at the HRU level. As plant growth and the movement of sediments, nutrients and other elements are significantly impacted by water balance, all the processes in SWAT are driven by it (Arnold et al. 2012).

The disaggregation of the basin into sub-basins helps the model reflect variations in evapotranspiration for various soils and crops as the model maintains a continuous water balance. Therefore, the model predicts runoff for each sub-basin, which is later combined to generate the total runoff for the entire basin. This also increases the accuracy of the model and subsequently provides a more detailed physical description of the water balance. More information on the model and its processes can be found in Arnold et al. (2012).

Spatial data

To delineate the basin into sub-basins and HRUs, the SWAT model requires three distinct types of spatial data, namely a DEM, a land-use and land-cover, and a soil map. Additionally, a map of the stream network can be used to better align the shape of the river network to the location and shape based on actual surveys. Table 1 summarizes the datasets and their sources.

The 30 m resolution Shuttle Radar Topography Mission (SRTM) DEM was provided as an input to the model. Some empty pixels and missing values in the DEM were filled with the SRTM DEM provided by CGIAR. The stream network for the entire study area was generated from the final DEM. The derived stream network was also compared with the map of major rivers in Myanmar provided in the Ayeyarwaddy State of the Basin Assessment and Google Earth to rectify incorrect locations for the main river network and some parts of the main tributaries. The basin was then divided into 20 sub-basins.

The land-use and land-cover map used in this study was obtained from the archive of IWMI which had 17 classes (Chandrasekharan & Rajah 2017). The land-use map disaggregated agricultural land into rain-fed land, flood plains, water managed and irrigated land. The map also accounted for the frequency of cropping (single, double for rain-fed, floodplains and water managed lands and single, double and triple for irrigated land). The main crop types were identified using the 2010 data from the Spatial Production Allocation Model. All this information was then added to the model through its management schedule capabilities, which enables users to provide information on irrigation operations and the planting and growing seasons. The land-use map shows that forest covers almost 70% of the total land area, followed by agricultural areas (22%). The land-use and land-cover map in Figure 2 illustrates that rain-fed and irrigated agricultural areas are concentrated in the southern part of the basin. Irrigated agricultural areas lie nearby the river network as clearly evident in sub-basins 13, 15, 16, 18 and 19.

The Digital Soil Map of the World was downloaded from the website of FAO.1 This map was used because the different properties of soil such as the hydrologic group, texture and the maximum rooting depth of the soil profile were readily available from the map and eliminated the need for estimations or measurements of these variables. The soil map in Figure 2 shows that acrisol, cambisol and gleysol soil types dominate the basin with orthic acrisols the most dominant.

Temporal data

The SWAT model requires five climatic variables: (i) precipitation, (ii) maximum and minimum temperature, (iii) relative humidity, (iv) wind speed and (v) solar radiation. The model also requires hydrological data to compare the simulation with observed discharge values to carry out the model calibration and validation. Table 2 provides information on the climatic and hydrological data used in this study along with their sources. Figure 3 shows the location of the stations and the four sub-watersheds used for calibration and validation. Since the majority (89%) of the basin is inside Myanmar, only data from Myanmar was used in this study.
Figure 3

Location of temporal data stations and calibration stations used in this study. Source: Author's creation.

Figure 3

Location of temporal data stations and calibration stations used in this study. Source: Author's creation.

Close modal

Where applicable, the collected data were compared to ensure the quality of the dataset. For stations where data was only available from a single source, the temporal variation in the data was compared with data from nearby stations.

In addition to these data, SWAT also requires data for a weather generator. The weather generator helps the model fill in missing data values in the observed input dataset. It includes various monthly weather statistics of the climatic parameters for all the stations in the basin. The statistics include monthly mean and standard deviation of precipitation and temperature (maximum and minimum), monthly skew coefficient for daily precipitation, probability of a wet-day followed by a dry-day and a wet-day in the month, average number of days of precipitation in the month, maximum half-hour precipitation in the entire period of records for a month, monthly averages of solar radiation, wind speed and dew point temperature.

The study area had only five stations with complete sets of the required climatic data which could be used to calculate monthly statistics for the weather generator. For other stations, the statistics were obtained from the global dataset generated by the SWAT team.2

Projected future climate data

Climate projections such as precipitation and minimum and maximum temperature are needed for the SWAT model to assess future hydrology. Regional Climate Models (RCMs) provide projections with finer resolution of data with better characteristics that reflect local climate (Kim et al. 2021) compared to GCMs. This study downloaded historical simulations and future projections of precipitation and minimum and maximum temperature from Coordinated Regional Climate Downscaling Experiment models (Giorgi & Mearns 1999) in southeast, east and west Asia domains from an Earth System Grid Federation node (Cinquini et al. 2014).

Data under two representative concentration pathways – RCP 4.5 and RCP 8.5 (van Vuuren et al. 2011) – for approximately 100 years of projection period were used. As RCM outputs also have biases and uncertainties, it is necessary to first apply statistical downscaling as a bias-correction process for the raw projections. The study used the empirical quantile mapping method devised by Wood et al. (2004) and Reichle & Koster (2004) for the bias-correction, which does not need the assumption on any parametric distribution. The process used the same observation datasets for precipitation and minimum and maximum temperature obtained from the Hydro-Informatics Centre. Climate data operator (Schulzweida et al. 2020), Python and R were used for pre-processing and standardizing climate outputs into a common coordinate reference system, merging temporal files and extraction of point-based time series. An R-package Qmap (Gudmundsson et al. 2012) was used to undertake station-based quantile mapping.

Bias-correcting parameters obtained from a training period (1986–2005) were applied to both RCM historical and projected data. An ensemble mean is calculated from the projections (Table 3) that shows similar climatic characteristics with the observed climate for both precipitation and temperature after downscaling. The ensemble is then used as input for the SWAT model for assessing future periods.

Table 3

Selected regional climate model outputs after downscaling for this study

SNGCMRCMInstitute
ICHEC-EC-EARTH RegCM4-3 RU-CORE 
MOHC-HadGEM2-ES 
MPI-M-MPI-ESM-MR 
CNRM-CERFACS-CNRM-CM5 RCA4 SMHI 
ICHEC-EC-EARTH 
MIROC-MIROC5 
MPI-M-MPI-ESM-LR 
NCC-NorESM1-M 
NOAA-GFDL-GFDL-ESM2M 
10 ICHEC-EC-EARTH HIRHAM5 DMI 
11 CNRM-CERFACS-CNRM-CM5 CCLM5-0-2 CLMcom 
12 MOHC-HadGEM2-ES CCLM5-0-4 
13 MPI-M-MPI-ESM-LR CCLM5-0-5 
SNGCMRCMInstitute
ICHEC-EC-EARTH RegCM4-3 RU-CORE 
MOHC-HadGEM2-ES 
MPI-M-MPI-ESM-MR 
CNRM-CERFACS-CNRM-CM5 RCA4 SMHI 
ICHEC-EC-EARTH 
MIROC-MIROC5 
MPI-M-MPI-ESM-LR 
NCC-NorESM1-M 
NOAA-GFDL-GFDL-ESM2M 
10 ICHEC-EC-EARTH HIRHAM5 DMI 
11 CNRM-CERFACS-CNRM-CM5 CCLM5-0-2 CLMcom 
12 MOHC-HadGEM2-ES CCLM5-0-4 
13 MPI-M-MPI-ESM-LR CCLM5-0-5 

Future water availability and water balance

Monthly, seasonal and annual water availability and water balance components for each sub-basin were calculated from SWAT results from 1988 to 2015. The impact of climate change has been analysed as a percentage change of a projected future with respect to the baseline. In general, Myanmar has five seasons: pre-monsoon (mid-April to mid-May), main monsoon (mid-May to mid-October), post-monsoon (mid-October to end-November), dry and cold (end-November to mid-March) and hot season (mid-March to mid-April) (Sirisena et al. 2018). However, for the seasonal analysis in this study, only three seasons were considered; the cool season (November to February), the hot season (March to May) and the rainy season (June to October). The outflow from the sub-basin is the total water availability at the river reach of that sub-basin. The precipitation, actual evapotranspiration (AET) and net water yields (NWYs) of the sub-basins are the water balance components of the relevant sub-basins. The subsequent states and changes in these components under RCP 4.5 and RCP 8.5 scenarios in the 2030s and 2050s were calculated.

Model calibration and validation

After setting up the model, the most sensitive parameters (Table 4) were identified by making use of the SWAT-CUP model developed by Karim C Abbaspour in Eawag: Swiss Federal Institute of Aquatic Science and Technology, which supported the model calibration and validation processes.

Table 4

List of parameters, their initial value and final value for the sub-basins

Sub-basinParameterDefault valueCalibrated value
All sub-basins CN2.mgt 72–92 58.32–74.52 
ESCO.hru 0.95 0.70–0.95 
TLAPS.sub −5.8 
CH_K2.rte 79.242 
BIOMIX.mgt 0.2 0.556 
CANMX.hru 35.15 
SURLAG.bsn 8.7 
SMFMN.bsn 4.5 3.81 
SMFMX.bsn 4.5 10 
Sub-basins of Katha REVAPMN.gw 750 350 
GWQMN.gw 1,000 550 
Sub-basins of Sagaing GWQMN.gw 1,000 2,000 
ALPHA_BF.gw 0.048 0.41 
Sub-basins of Katha, Moniya and Magway ALPHA_BF.gw 0.048 0.2 
Sub-basinParameterDefault valueCalibrated value
All sub-basins CN2.mgt 72–92 58.32–74.52 
ESCO.hru 0.95 0.70–0.95 
TLAPS.sub −5.8 
CH_K2.rte 79.242 
BIOMIX.mgt 0.2 0.556 
CANMX.hru 35.15 
SURLAG.bsn 8.7 
SMFMN.bsn 4.5 3.81 
SMFMX.bsn 4.5 10 
Sub-basins of Katha REVAPMN.gw 750 350 
GWQMN.gw 1,000 550 
Sub-basins of Sagaing GWQMN.gw 1,000 2,000 
ALPHA_BF.gw 0.048 0.41 
Sub-basins of Katha, Moniya and Magway ALPHA_BF.gw 0.048 0.2 

The model was calibrated for the period 1994–2000 and validated for the period 2002–2008. These periods were selected because their flow data was more complete and consistent. More important, these periods also represent the pristine condition of the river basin where human interventions in the form of dams and reservoirs had not yet resulted in considerable alterations in the basin hydrology.

In the model, reservoirs were added in sub-basins 1, 7, 9, 12, 13, 15, 17, 18, 19 and 20. The number of reservoirs, their capacity to retain water and the flows they regulated were calculated and introduced into the model. The model was fine-tuned using auto calibration in SWAT-CUP with the reasonable value range for all the sensitive parameters (Table 4). The calibration was fine-tuned further by manual calibration. The calibration and validation was evaluated by three criteria, Nash–Sutcliffe Efficiency (NSE), Coefficient of Determination (R2) and Percentage Bias (PBIS).

The model was calibrated and validated using the daily flow data from four stations (Katha, Sagaing, Monywa and Magway, as shown in Figure 3). In the calibration process, first, parameters identified in the sensitivity analysis for Magway as the most sensitive parameters were adjusted and changes were made to the least sensitive parameters. For some stations, additional sensitivity analysis was done to identify sensitive parameters. Table 4 lists the selected parameters for this study along with their initial values and ranges and fitted values and ranges for two stations. The comparison between observed and simulated daily flows at all four stations is presented in Figure 4.
Figure 4

Observed and simulated daily flows at (a) Katha, (b) Monywa, (c) Sagaing and (d) Magway. Source: Author's creation.

Figure 4

Observed and simulated daily flows at (a) Katha, (b) Monywa, (c) Sagaing and (d) Magway. Source: Author's creation.

Close modal

After the model was deemed to be performing satisfactorily in the calibration period, it was validated for a different period (i.e. 2002–2008) using the same sets of parameters obtained during calibration. Table 5 shows the daily statistics of model calibration and validation, respectively, and Figure 4 shows the observed and simulated hydrographs at the four gauging stations. Overall, the model performed satisfactorily. The NSE values are above 0.60 for daily simulations for all four stations. The model simulation improves in the lower part of the basin as can be seen in the figure and statistics obtained at Magway.

Table 5

Daily calibration and validation statistics at four hydrostations

StationEvaluation criteriaCalibrationValidation
Katha NSE 0.62 0.69 
R2 0.65 0.72 
PBIAS +1.07% +4.83% 
Monywa NSE 0.82 0.73 
R2 0.84 0.83 
PBIAS −7.88% −13.51% 
Sagaing NSE 0.68 0.76 
R2 0.73 0.76 
PBIAS +20.15% +3.01% 
Magway NSE 0.86 0.83 
R2 0.87 0.86 
PBIAS +0.56% +5.34% 
StationEvaluation criteriaCalibrationValidation
Katha NSE 0.62 0.69 
R2 0.65 0.72 
PBIAS +1.07% +4.83% 
Monywa NSE 0.82 0.73 
R2 0.84 0.83 
PBIAS −7.88% −13.51% 
Sagaing NSE 0.68 0.76 
R2 0.73 0.76 
PBIAS +20.15% +3.01% 
Magway NSE 0.86 0.83 
R2 0.87 0.86 
PBIAS +0.56% +5.34% 

Characterization of current hydrology

The simulated results from 1988 to 2015 generated by the SWAT model were used to characterize the current hydrological regime and water balance of the UARB. For this analysis, we accounted for four water components: precipitation, AET, NWY and balance closure. Where the AET comprises total evaporation and transpiration of the sub-basins, the NWY is the total routed flows (surface runoff, groundwater and lateral flow) of the sub-basins and the balance closure is a collective term pertaining to groundwater recharge, change in soil moisture storage in the vadose zone and other transmission losses and losses through deep water percolation in the system. Outflow of the sub-basin is the sum of the routed flow of the respective sub-basin and flows from all the upstream sub-basins.

Sub-basin and agro-ecological region water balances

The annual average water balance components in Figure 5 for the simulation period from 1988 to 2015 highlight the spatial heterogeneity in all components. Sub-basins in the north-western region (2, 3, 4, 5 and 6) have the highest precipitation (>2,000 mm), while the lowest precipitation (<800 mm) is seen in the southern region (sub-basins 12, 15, 18 and 20). This was expected because mountainous areas have higher precipitation compared to regions in the central south, some of which fall inside the central dry zone. As AET relies predominantly on precipitation, land use, land cover and temperature, lower AET values can be seen in the northern mountainous regions (sub-basins 1 and 2: <50 mm). The highest AET values are in the central-southern region with forest cover and provisions for irrigated agriculture (>700 mm in sub-basins 9, 10, 15, 16, 17 and 19).
Figure 5

Sub-basin annual average water balance (1988–2015) in the basin. Source: Author's creation.

Figure 5

Sub-basin annual average water balance (1988–2015) in the basin. Source: Author's creation.

Close modal

NWY is highest in the northern mountain areas (i.e. sub-basin 2: 2,537 mm) followed by sub-basins 3 (2,371 mm), 4 (2,224 mm) and 1 (1,605 mm). The lowest NWY is observed in sub-basins 20 (48 mm), 12 (52 mm) and 18 (99 mm), which are in the central dry zone. The results indicate that in 14 of the 20 sub-basins, precipitation contributes to storage (aquifer recharge) because of the positive balance closure component (highest in sub-basins 1 and 2). The negative balance closure component was mostly observed in the southern region (higher sub-basin numbers), which signals the contribution from groundwater aquifer storage to the base flow that replenishes during the rainy season.

The distribution proportion of the long-term average of the water balance components for three seasons of the sub-basins is shown in Figure 6. The AET is higher in the southern part of the basin (which belongs to the higher sub-basin number) and the NWY is higher in the northern part of the basin (which belongs to the lower sub-basin number). Results also show that the rainy season is the main hydrological driver. Note that in Myanmar, more than 84% of the precipitation occurs in the rainy season from June to October owing to the southwest monsoon. In the remaining months, the base flow is sustained by groundwater recharge during the monsoon period. The results further show that the cool season has the least average seasonal precipitation. The highest precipitations in the cool, hot and rainy seasons were 105, 493 and 2,658 mm in sub-basins 19, 4 and 3, respectively. Similarly, the lowest precipitation in the cool season (16 mm) was seen in sub-basin 3 and in the hot (86 mm) and rainy seasons (327 mm) in sub-basin 20. This is mainly because in the northern mountainous regions, the cool season has snowfall rather than rainfall and snow is not readily available to contribute to river discharge.
Figure 6

Sub-basin proportional distribution of the seasonal average of the water balance components (1988–2015) in the basin (blue gradients for precipitation; green for AET; orange for NWY; and yellow for balance closure). Source: Author's creation. Please refer to the online version of this paper to see this figure in colour: https://dx.doi.org/10.2166/wcc.2022.407.

Figure 6

Sub-basin proportional distribution of the seasonal average of the water balance components (1988–2015) in the basin (blue gradients for precipitation; green for AET; orange for NWY; and yellow for balance closure). Source: Author's creation. Please refer to the online version of this paper to see this figure in colour: https://dx.doi.org/10.2166/wcc.2022.407.

Close modal

The highest AET (>100 mm) in the cool season was observed in sub-basins 10, 15, 16 and 19, which are mostly forested or agriculture areas, and the lowest in sub-basins 1 and 2 (<5 mm). AET rates higher than 150 mm are observed in sub-basins 5, 10 and 19 in the hot season with only sub-basins 1, 2, 3, 18 and 20 showing AET values less than 100 mm. Sub-basin 19 has the highest AET (559 mm) in the rainy season, followed by sub-basins 10 (543 mm) and 17 (541 mm). Sub-basin 10 had one of the highest AET levels in all three seasons mainly because there is a tributary of the Chindwin River and it is mostly forested with some land used for rain-fed and irrigated agriculture.

NWY are highest in the rainy season in all sub-basins, which is as expected because the water from precipitation directly contributes to the water yields of the sub-basin. The highest seasonal NWYs are 266 mm (sub-basin 2), 312 mm (sub-basin 2) and 2,194 mm (sub-basin 3) in the cool, hot and rainy seasons, respectively, whereas the lowest seasonal NWY are 12 mm (sub-basin 20), 3 mm (sub-basin 12) and 30 mm (sub-basin 20) in the cool, hot and rainy seasons, respectively. Note that the overall water balances of the basin are −141, +60 and +231 mm in the cool, hot and rainy seasons, respectively. This shows that the contribution from groundwater aquifers to the base flow persists in the cool season and groundwater aquifers are replenished in the rainy season.

Table 6 shows the annual and seasonal simulated water balances in the three agro-ecological regions within the basin for the period from 1988 to 2015. These three agro-ecological regions are classified based on the ground elevation from the metres above sea level, which are mountains (>1,000 masl), hills (300–1,000 masl) and inland plains (<300 masl). In this analysis, the sub-basin level water balances were aggregated at the agro-ecological region level by area weightage average techniques. The highest annual average precipitation (1,784 mm) was observed in the mountains and the lowest (1,406 mm) in the inland plains. The simulated seasonal precipitation also follows a similar pattern except in the cool season. All three agro-ecological regions have almost equal precipitation (∼40 mm) in the cool season. However, this is an average for the entire basin and it varies spatially throughout the basin. The seasonal analysis results of the simulated precipitation show that nearly 85% of the annual precipitation in all agro-ecological regions occurred in the rainy season only.

Table 6

Agro-ecological region annual and seasonal water balances in the basin (1988–2015)

Agro-ecological regionWater balance (mm)
ComponentsAnnualCool seasonHot seasonRainy season
Mountains (>1,000 masl) Precipitation 1,784 40 (2%) 229 (13%) 1,515 (85%) 
AET 410 55 (14%) 83 (20%) 272 (66%) 
NWY 1,285 123 (10%) 70 (5%) 1,092 (85%) 
Hills (300–1,000 masl) Precipitation 1,574 40 (3%) 204 (13%) 1,330 (84%) 
AET 553 76 (14%) 105 (19%) 372 (67%) 
NWY 978 103 (11%) 41 (4%) 835 (85%) 
Inland plains (<300 masl) Precipitation 1,406 38 (3%) 190 (13%) 1,177 (84%) 
AET 598 89 (15%) 115 (19%) 393 (66%) 
NWY 800 95 (12%) 28 (3%) 678 (85%) 
Agro-ecological regionWater balance (mm)
ComponentsAnnualCool seasonHot seasonRainy season
Mountains (>1,000 masl) Precipitation 1,784 40 (2%) 229 (13%) 1,515 (85%) 
AET 410 55 (14%) 83 (20%) 272 (66%) 
NWY 1,285 123 (10%) 70 (5%) 1,092 (85%) 
Hills (300–1,000 masl) Precipitation 1,574 40 (3%) 204 (13%) 1,330 (84%) 
AET 553 76 (14%) 105 (19%) 372 (67%) 
NWY 978 103 (11%) 41 (4%) 835 (85%) 
Inland plains (<300 masl) Precipitation 1,406 38 (3%) 190 (13%) 1,177 (84%) 
AET 598 89 (15%) 115 (19%) 393 (66%) 
NWY 800 95 (12%) 28 (3%) 678 (85%) 

The agro-ecological region simulated annual AET was the highest (598 mm) in the inland plains and the lowest (410 mm) in the mountains. A similar pattern was seen in the seasonal simulated results of the AET (Table 6). Unlike precipitation, the AET was comparatively higher in the basin lowlands. This is likely due to the dominant role of temperature that triggers higher evaporation and transpiration from larger water bodies and agricultural lands in this area. The seasonal variation of the AET was dominated by the seasonal precipitation in the agro-ecological regions. The seasonal analysis results of the simulated AET showed that nearly 66% of the annual AET in all agro-ecological regions are observed in the rainy season.

Similarly, the agro-ecological region simulated annual NWYs were highest (1,285 mm) in the mountains and lowest (800 mm) in the inland plains. Similar patterns were seen in the seasonal simulated results of the NWY (Table 6). The simulated NWY seasonal results show that about 85% of the total annual NWY of the agro-ecological regions contributes directly to stream and river flows in the rainy season alone. In comparing the cool and hot seasons, the NWY was higher in the cool season, while precipitation was higher in the hot season in all agro-ecological regions. This shows that the base flow contribution to stream and river flows is higher in the cool season, which comes from groundwater aquifers that get recharged during the rainy season.

Mean monthly water balances

The mean monthly variation in water balance components for all four stations used in calibration and validation for the simulation period of 1988–2015 is illustrated in Figure 7. As expected, the highest precipitation is seen in the rainy season (June to October). AET follows a similar trend as more precipitation implies more water available for evaporation and transpiration. As expected, maximum water yield can also be seen in the rainy season. Figure 7(d) shows the monthly water balance for Magway. Of the 1,578 mm of average annual precipitation in the UARB over the study period, 84.4% of the total precipitation occurred in the rainy season (1,332 mm). This is followed by the hot season with 13.1% (207 mm) of the total precipitation in UARB while the cool season has only 2.5% (39 mm) of precipitation. Of the total 524 mm average annual AET, the highest AET occurred in the rainy season (348 mm, i.e. 66.4%), followed by the hot season (102 mm, i.e. 19.5%) and then the cool season (74 mm, i.e. 14.1%). About 85.0% (858 mm) of the 1,010 mm of total NWY occurred in the rainy season and 4.6% (46 mm) and 10.5% (106 mm) in the hot and cool seasons, respectively.
Figure 7

Mean monthly water balances for the period 1988–2015 in different regions based on the location of hydrostations. Source: Author's creation.

Figure 7

Mean monthly water balances for the period 1988–2015 in different regions based on the location of hydrostations. Source: Author's creation.

Close modal

In contrast to the observed seasonal precipitation pattern, the NWY was lowest in the hot season and slightly higher in the cool season. This shows that NWY for the cool season make more of a contribution to the groundwater stored in the rainy season and groundwater recharge begins in the hot season. This suggests that in the cool season, regardless of the lowest seasonal precipitation, the recharged groundwater aquifers from the rainy season contribute to the base flow, thereby making the NWY higher than in the hot season. The little precipitation (207 mm) in the hot season mostly contributes to the replenishment of soil moisture and groundwater aquifers, resulting in the lowest NWY. Seasonal AET patterns showed similar precipitation patterns because evaporation and transpiration processes depend on canopy coverage, temperature, water at the surface and water content of the topsoil. The actual evaporation in the cool season (74 mm) was higher than precipitation in the cool season (39 mm), which means evaporation from the subsoil and canopy play a dominant role in the AET of the cool season.

Future climate projections

Sub-basin projected changes of water balances

The analysis for future water balance components were considered for two scenarios: RCP 4.5 (intermediate scenario) and RCP 8.5 (worst-case scenario) and two periods: near-future (2016–2045 or the 2030s) and mid-future (2036–2065 or the 2050s). Figures 8,910 show the percentage change in the future precipitation, AET and NWY in the entire study area in near- and mid-future periods under RCP 4.5 and RCP 8.5 scenarios.
Figure 8

Percentage change in future annual precipitation in all sub-basins compared to historical precipitation (1988–2015). Source: Author's creation.

Figure 8

Percentage change in future annual precipitation in all sub-basins compared to historical precipitation (1988–2015). Source: Author's creation.

Close modal
Figure 9

Percentage change in future actual evapotranspiration in all sub-basins compared to historical precipitation (1988–2015). Source: Author's creation.

Figure 9

Percentage change in future actual evapotranspiration in all sub-basins compared to historical precipitation (1988–2015). Source: Author's creation.

Close modal
Figure 10

Percentage change in future net water yield in all sub-basins compared to historical precipitation (1988–2015). Source: Author's creation.

Figure 10

Percentage change in future net water yield in all sub-basins compared to historical precipitation (1988–2015). Source: Author's creation.

Close modal

Compared with the baseline (1988–2015), precipitation in the entire basin in the cool season is predicted to decrease in most sub-basins both in the near- and mid-future. The average changes in precipitation in the cool season will be +0.3% (−31 to +139%), −7.4% (−28 to +13%), −14.7% (−46 to +18%) and −11.0% (−42 to +19%) under scenarios RCP 4.5-2030s, RCP 4.5-2050s, RCP 8.5-2030s and RCP 8.5-2050s, respectively (Figure 8(a)).

However, as evident from Figure 8(a), the changes in precipitation have spatial heterogeneity, which shows some sub-basins in the north-eastern region (sub-basins 2, 3 and 4) are expected to see an increase under RCP 8.5 for both the 2030s and 2050s. Sub-basins 15, 18 and 20 in the southernmost regions are expected to see an increase in precipitation under RCP 4.5-2030s. For the hot season (Figure 8(b)), future precipitation is expected to increase in many sub-basins under all scenarios.

The average changes in the precipitation in the hot season will be +6.6% (−4 to 63%), +14.6% (−2 to +88%), +3.3% (−6 to +46%) and +11.4% (−1 to 77%) under scenarios RCP 4.5-2030s, RCP 4.5-2050s, RCP 8.5-2030s and RCP 8.5-2050s, respectively (Figure 8(b)). Precipitation is expected to increase more in the mid-future than in the near-future. The sub-basins at the mid-western (for both RCP 4.5 and RCP 8.5) and central regions (only for RCP 8.5) are expected to increase in the near-future. Similarly, precipitation in the rainy season is expected to increase in most sub-basins (Figure 8(c)). The average changes in precipitation in the rainy season will be +2.9% (−2 to 11%), +7.1% (0 to +27%), +8.3% (+0 to +25%) and +9.9% (0 to +26%) under scenarios RCP 4.5-2030 s, RCP 4.5-2050s, RCP 8.5-2030s and RCP 8.5-2050s, respectively (Figure 8(b)).

The results show that precipitation in the rainy season in the eastern sub-basins is likely to decrease in the near- and mid-future under both scenarios. Thus, water availability will decrease in almost all sub-basins in the cool season, in the western sub-basins in the hot season and in the eastern sub-basins in the rainy season.

The results suggest that dry periods (cool season and partly hot season) will be drier and wet periods will be wetter. To minimize the uncertainties in the projections due to the different projection scenarios, the ensemble of the 13 RCMs was analysed to see the average impact of climate change on water availability and water balances.

The results show that AET is expected to increase in all three seasons in almost all sub-basins (Figure 9). In the cool season, the average changes in AET will be +3, +7, +3 and +7% under scenarios RCP 4.5-2030s, RCP 4.5-2050s, RCP 8.5-2030s and RCP 8.5-2050s, respectively (Figure 9(a)). In the hot season, the average changes in AET will be +3, +6, +2 and +4% under scenarios RCP 4.5-2030s, RCP 4.5-2050s, RCP 8.5-2030s and RCP 8.5-2050s, respectively (Figure 9(b)). Similarly, Figure 9(c) shows that in the rainy season, the average changes in AET will be +3, +7, +6 and +7% under scenarios RCP 4.5-2030s, RCP 4.5-2050s, RCP 8.5-2030s and RCP 8.5-2050s, respectively.

The changes in AET were found to be directly related to changes in precipitation throughout the basin and that the AET is decreasing in sub-basins covered equally by forest and agricultural land in the central and south-eastern part of the basin.

Overall, the changes in NWY with respect to the baseline are also projected to increase (Figure 10). In the cool season, the average changes will be +3, +6, +7 and +10% under scenarios RCP 4.5-2030s, RCP 4.5-2050s, RCP 8.5-2030s and RCP 8.5-2050s, respectively (Figure 10(a)). In the hot season, the average changes in NWY will be +0, +12, +0 and +6% under scenarios RCP 4.5-2030s, RCP 4.5-2050s, RCP 8.5-2030s and RCP 8.5-2050s, respectively (Figure 10(b)).

Similarly, in the rainy season, the average changes in NWY will be +4, +8, +10 and +13% under scenarios RCP 4.5-2030s, RCP 4.5-2050s, RCP 8.5-2030s and RCP 8.5-2050s, respectively (Figure 10(c)). This highlights the spatial variation in sub-basins for all seasons with mostly central eastern and western regions showing a decrease in future NWY and other regions showing a general increase.

The results suggest that although the overall water balance components show one trend, sub-basin analyses do not yield the same trends for the entire basin. It is essential to understand the spatial and seasonal variation of the changes within the basin to better prepare for water management planning and interventions. For instance, of the available 358.49 billion cubic metres of water in the basin, 5% was available in the hot season followed by 15% in the cool season and 80% in the rainy season.

Results for the 2030s showed that the total available water in the cool season will increase by 1.1% under RCP 4.5 and increase by 6.2% under RCP 8.5. This suggests more water will be available in the cool season if carbon emissions follow a worst-case scenario, which is echoed in the results for the hot and rainy seasons for the 2030s (a change of −0.7 and −0.9% in the hot season under RCP 4.5 and RCP 8.5, respectively, and an increase of 3.5 and 9.3% by the 2030s in the rainy season under RCP 4.5 and RCP 8.5, respectively).

The results for the 2050s demonstrate an overall increase in future water availability for all seasons under both RCP scenarios. The results for the 2050s follow the trend observed in the 2030s for all seasons, where the results of RCP 8.5 show higher water availability compared to RCP 4.5. The hot season in the 2050s has higher water availability under RCP 4.5 (4.5%) than RCP 8.5 scenario (2.8%). By the 2050s, water availability in the cool season is expected to increase by 4.0 and 9.0% under RCP 4.5 and RCP 8.5 scenarios, respectively. Water availability in the rainy season is expected to increase by 7.5 and 13.9% under RCP 4.5 and RCP 8.5, respectively, by the 2050s. The higher increase in water availability with the higher increase in temperature is most likely because of accelerated snowmelt in the northern regions and increases in precipitation, particularly in the rainy season.

Agro-ecological regions: projected changes of water balances

The box plots in Figure 11 illustrate water balance changes in relation to the baseline under the climate change scenarios RCP 4.5 and RCP 8.5 in near-future 2030s and mid-future 2050s projection periods.
Figure 11

Box plots of annual and seasonal (cool, hot and rainy) projected change in precipitation versus actual evapotranspiration (top row) and net water yields (bottom row) in three agro-ecological regions (mountains, hills and inland plains). Source: Author's creation.

Figure 11

Box plots of annual and seasonal (cool, hot and rainy) projected change in precipitation versus actual evapotranspiration (top row) and net water yields (bottom row) in three agro-ecological regions (mountains, hills and inland plains). Source: Author's creation.

Close modal

Figure 11(a) shows the annual changing scenarios of precipitation and AET (both annual precipitation and AET decrease in the mountains and inland plains and both increase in the hills under RCP 4.5-2030s). Annual precipitation slightly decreases in the mountains and inland plains and increases in the hills, whereas annual AET increases in all three agro-ecological regions under RCP 4.5-2050s. Similarly, annual precipitation increases in the hills and inland plains and slightly decreases in the mountains, whereas the annual AET increases in all three agro-ecological regions under RCP 8.5-2030s. Both annual precipitation and AET increase in all three agro-ecological regions under RCP 8.5-2050s. Both annual precipitation and AET show an increasing trend in both climate change scenarios and both periods except for the mountains and inland plains under RCP 4.5 in the near-future scenario.

Figure 11(b)–11(d) show plots for the cool, hot and rainy seasons. In the cool season, precipitation decreases and AET increases in all agro-ecological regions under all the projection scenarios. The change patterns in the cool season can be observed distinctly between the medium and worst-case scenarios in all agro-ecological regions. Similar change patterns for precipitation and AET appear in the hot and rainy seasons as well. In the hot and rainy seasons, both precipitation and AET decrease in the mountains and inland plains under RCP 4.5-2030s. Both precipitation and AET increase with respect to the baseline in all agro-ecological regions under other projection scenarios. Compared with the hot season, the increase is very high in the rainy season. These projected results show that the dry season will be drier and the wet season wetter in basin agro-ecological regions.

Figure 11(e)–11(h) show Δ change in mm, percent change of precipitation and NWY on an annual and seasonal basis under the projected change scenarios RCP 4.5 and RCP 8.5 for the near- and the mid-future periods in the three agro-ecological regions. The annual and seasonal NWYs are decreasing in the mountains and inland plains under RCP 4.5-2030s scenario and also in the mountains and inland plains under RCP 4.5-2050s scenario except in the hot season. Overall, the scenarios for annual and seasonal NWY changes have a similar pattern for AET changes.

In both plots of AET and NWY, significant trends can be observed in the relationships of the changes in precipitation with changes in AET and NWY in the rainy season. These are linear and have a significant correlation (Figure 11(a), 11(d), 11(e) and 11(h)). In the hot season, the relationships are linear and have a significant correlation (Figure 11(c) and 11(g)) but not a very significant correlation in the cool season (Figure 11(b) and 11(f)). This suggests large uncertainty in the projections for water balances for the cool season compared with other seasons.

This study was conducted for the Upper Ayeyarwaddy River Basin of Myanmar to characterize hydrological parameters under current and future climate change scenarios using the distributed SWAT hydrological model.

The results show that the mountainous region in the north sees more precipitation compared with the central south, some of which falls inside the central dry zone. Precipitations in the north-west regions are over 2,000 mm and in the southern regions less than 800 mm. The lowest AET was observed in the northern mountainous regions (<50 mm) and the highest values in the central-southern region with forest cover or provisions for irrigated agriculture (>700 mm). The NWYs are the highest in the northern mountain regions (>2,000 mm) and lowest in the southern regions and the central dry zone (<100 mm). The results indicate that in 14 out of the 20 sub-basins, precipitation contributes to storage (aquifer recharge) owing to the positive water balance component. The negative water balance component was mostly observed in the southern region, which signals a contribution from groundwater aquifer storage to the base flow which is replenished during the rainy season. Similar trends and patterns are found in the monthly and seasonal analysis of the results of both the sub-basins and agro-ecological region accumulations.

The results also show that precipitation in the rainy season in the eastern sub-basins is likely to decrease in the near- and mid-future under both scenarios. The seasonal analysis of change in precipitation shows that water availability will decrease in almost all sub-basins in the cooler seasons, at the western sub-basins in the hot season and at the eastern sub-basins in the rainy season. In summary, the results suggest that dry periods (the cool season and partly hot season) will be getting drier and wet periods wetter.

The hydrological analysis of the study area can be further improved and refined for future studies by using (i) a high resolution spatially disaggregated information for soil classes and land-use classes, which incorporates the heterogeneity of land surfaces in detail; (ii) the additional periods of temporal climate and discharge data of the existing stations and also from additional stations, which improves the results for water balance and water availability; and (iii) the detailed information on the water use practices in the basin, including reservoirs operations and agricultural uses. Furthermore, the water balance results can be further downscaled by adding more sub-basins, which will make it easier to plan and develop water management practices at the local level.

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

Arnold
J. G.
,
Moriasi
D. N.
,
Gassman
P. W.
,
Abbaspour
K. C.
,
White
M. J.
,
Srinivasan
R.
,
Santhi
C.
,
Harmel
R. D.
,
Griensven
A. v.
,
Liew
M. W. V.
,
Kannan
N.
&
Jha
M. K.
2012
SWAT: model use, calibration, and validation
.
Transactions of the American Society of Agricultural and Biological Engineers
55
(
4
),
1491
1508
.
Bharati
L.
,
Bhattarai
U.
,
Khadka
A.
,
Gurung
P.
,
Neumann
L. E.
,
Penton
D. J.
,
Dhaubanjar
S.
&
Nepal
S.
2019
From the Mountains to the Plains: Impact of Climate Change on Water Resources in the Koshi River Basin
.
IWMI Working Paper 187
.
International Water Management Institute
,
Colombo
,
Sri Lanka
, p.
41
.
Chandrasekharan
K.
&
Rajah
A.
2017
Land Use and Land Cover Map of Salween and Ayeyarwady River Basins for the Year 2014 [Dateset]
.
International Water Management Institute
,
Colombo
,
Sri Lanka
.
Chavoshian
A.
,
Ishidaira
H.
,
Takeuchi
K.
&
Yoshitani
J.
2007
Hydrological Modeling of Large-scale Ungauged Basin Case Study: Ayeyarwady (Irrawaddy) Basin, Myanmar. Paper presented at the HRSD 2007 Conference in Conjunction with the 15th Regional Steering Committee Meeting for UNESCO-IHP Southeast Asia and the Pacific, Manila, Philippines, 22–23 November 2007
.
Cinquini
L.
,
Crichton
D.
,
Mattmann
C.
,
Harney
J.
,
Shipman
G.
,
Wang
F.
,
Ananthakrishnan
R.
,
Miller
N.
,
Denvil
S.
,
Morgan
M.
,
Pobre
Z.
,
Bell
G. M.
,
Doutriaux
C.
,
Drach
R.
,
Williams
D.
,
Kershaw
P.
,
Pascoe
S.
,
Gonzalez
E.
,
Fiore
S.
&
Schweitzer
R.
2014
The Earth System Grid Federation: an open infrastructure for access to distributed geospatial data
.
Future Generation Computer Systems
36
,
400
417
.
Ghimire
U.
,
Babel
M. S.
,
Shrestha
S.
&
Srinivasan
G.
2019
A multi-temporal analysis of streamflow using multiple CMIP5 GCMs in the Upper Ayerawaddy Basin, Myanmar
.
Climatic Change
155
,
59
79
.
Ghimire
U.
,
Agarwal
A.
,
Shrestha
N. K.
,
Daggupati
P.
,
Srinivasan
G.
&
Than
H. H.
2020
Applicability of lumped hydrological models in a data-constrained river basin of Asia
.
Journal of Hydrologic Engineering
25
(
8
),
05020018
.
Giorgi
F.
&
Mearns
L. O.
1999
Introduction to special section: regional climate modeling revisited
.
Journal of Geophysical Research
104
(
D6
),
6335
6352
.
Gudmundsson
L.
,
Bremnes
J. B.
,
Haugen
J. E.
&
Skaugen
T. E.
2012
Technical note: downscaling RCM precipitation to the station scale using quantile mapping – a comparison of methods
.
Hydrology and Earth System Sciences Discussions
9
,
6185
6201
.
HARP-F & MIMU
2018
Vulnerability in Myanmar: a secondary data review of needs, coverage and gaps
. In:
Humanitarian Assistance and Resilience Programme Facility & the Myanmar Information Management Unit
, p.
113
.
Horton
R.
,
De Mel
M.
,
Peters
D.
,
Lesk
C.
,
Bartlett
R.
,
Helsingen
H.
,
Bader
D.
,
Capizzi
P.
,
Martin
S.
&
Rosenzweig
C.
2016
Assessing Climate Risk in Myanmar
.
Center for Climate Systems Research at Columbia University, WWF-US and WWF-Myanmar
,
New York, NY
,
USA
, p.
88
.
Kattelus
M.
,
Rahaman
M. M.
&
Varis
O.
2015
Hydropower development in Myanmar and its implications on regional energy cooperation
.
International Journal of Sustainable Society
7
(
1
),
42
66
.
Khoi
D. N.
,
Nguyen
V. T.
,
Sam
T. T.
,
Ky Phung
N.
&
Thi Bay
N.
2020
Responses of river discharge and sediment load to climate change in the transboundary Mekong River Basin
.
Water and Environment Journal
34
(
S1
),
367
380
.
Kim
G.
,
Cha
D.-H.
,
Park
C.
,
Jin
C.-S.
,
Lee
D.-K.
,
Suh
M.-S.
,
Oh
S.-G.
,
Hong
S.-Y.
,
Ahn
J.-B.
,
Min
S.-K.
&
Kang
H.-S.
2021
Evaluation and projection of regional climate over East Asia in CORDEX-East Asia Phase I experiment
.
Asia-Pacific Journal of Atmospheric Sciences
57
,
119
134
.
Kreft
S.
,
Eckstein
D.
,
Dorsch
L.
&
Fischer
L.
2015
Global Climate Risk Index 2016: Who Suffers Most from Extreme Weather Events? Weather-Related Loss Events in 2014 and 1995 to 2014
.
Federal Ministry for Economic Cooperation and Development
,
Bonn
,
Germany
, p.
31
.
Oo
H. T.
,
Zin
W. W.
&
Kyi
C. C. T.
2019
Assessment of future climate change projections using multiple global climate models
.
Civil Engineering Journal
5
(
10
),
2152
2166
.
Oo
H. T.
,
Zin
W. W.
&
Kyi
C. C. T.
2020
Analysis of streamflow response to changing climate conditions using SWAT model
.
Civil Engineering Journal
6
(
2
),
194
209
.
Pandey
V. P.
,
Dhaubanjar
S.
,
Bharati
L.
&
Thapa
B. R.
2019
Hydrological response of Chamelia watershed in Mahakali Basin to climate change
.
Science of The Total Environment
650
,
365
383
.
Reichle
R. H.
&
Koster
R. D.
2004
Bias reduction in short records of satellite soil moisture
.
Geophysical Research Letters
31
,
L19501
.
Sam
T. T.
,
Khoi
D. N.
,
Thao
N. T. T.
,
Nhi
P. T. T.
,
Quan
N. T.
,
Hoan
N. X.
&
Nguyen
V. T.
2019
Impact of climate change on meteorological, hydrological and agricultural droughts in the Lower Mekong River Basin: a case study of the Srepok Basin, Vietnam
.
Water and Environment Journal
33
(
4
),
547
559
.
Schulzweida
U.
,
Kornblueh
L.
&
Quast
R.
2020
User Guide: Climate Data Operator, Version 1.9.9. MPI for Meteorology, pp. 205–209
.
Sirisena
T. A. J. G.
,
Maskey
S.
,
Ranasinghe
R.
&
Babel
M. S.
2018
Effects of different precipitation inputs on streamflow simulation in the Irrawaddy River Basin, Myanmar
.
Journal of Hydrology: Regional Studies
19
,
265
278
.
Sirisena
T. A. J. G.
,
Maskey
S.
,
Bamunawala
J.
&
Ranasinghe
R.
2021
Climate change and reservoir impacts on 21st-century streamflow and fluvial sediment loads in the Irrawaddy River, Myanmar
.
Frontiers in Earth Science
.
doi: 10.3389/FEART.2021.644527
.
Taft
L.
&
Evers
M.
2016
A review of current and possible future human-water dynamics in Myanmar's river basins
.
Hydrology and Earth System Sciences
20
,
4913
4928
.
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
.
Climate Change
109
,
5
31
.
Wood
A.
,
Leung
L.
,
Sridhar
V.
&
Lettenmaier
D. P.
2004
Hydrologic implications of dynamical and statistical approaches to downscaling climate model outputs
.
Climate Change
62
,
189
216
.
This is an Open Access article distributed under the terms of the Creative Commons Attribution Licence (CC BY-NC-ND 4.0), which permits copying and redistribution for non-commercial purposes with no derivatives, provided the original work is properly cited (http://creativecommons.org/licenses/by-nc-nd/4.0/).