International Journal of Sustainable Built Environment (2017) 6, 285–298

H O S T E D BY
Gulf Organisation for Research and Development

International Journal of Sustainable Built Environment

ScienceDirect
www.sciencedirect.com

Original Article/Research

Flood hazard assessment under climate change scenarios in the

Yang River Basin, Thailand
Sangam Shrestha ⇑, Worapong Lohpaisankrit
Water Engineering and Management, School of Engineering and Technology, Asian Institute of Technology, P.O. Box 4, Klong Luang, Pathumthani
12120, Thailand

Received 27 May 2015; accepted 26 September 2016

Abstract

Climate change is expected to increase both the magnitude and frequency of extreme precipitation events, which may lead to more
intense and frequent river flooding. This study aims to assess the flood hazard potential under climate change scenarios in Yang River
Basin of Thailand. A physically-based distributed hydrological model, Block-wise use of TOPMODEL using Muskingum-Cunge flow
routing (BTOPMC) and hydraulic model, HEC-RAS was used to simulate the floods under future climate scenarios. Future climate sce-
narios were constructed from the bias corrected outputs of three General Circulation Models (GCMs) for 2020s, 2050s and 2080s.
Results show that basin will get warmer and wetter in future. Both the minimum and maximum temperature of the basin is projected
to increase in future. Similarly average annual rainfall is also projected to increase in future, higher in near future and lower in far future.
The extreme runoff pattern and synthetic inflow hydrographs for 25, 50 and 100 year return flood were derived from an extreme flood of
2007 which were then fed into HEC-RAS model to generate the flood inundation maps in the basin. The intensity of annual floods is
expected increase for both RCP 4.5 and 8.5 scenarios. Compared to the baseline period, an additional 60 km2 area of basin is projected
to be flooded with the return period of 100 years. The results of this study will be helpful to formulate adaptation strategies to offset the
negative impacts of flooding on different land use activities in Yang River Basin.
Ó 2017 The Gulf Organisation for Research and Development. Production and hosting by Elsevier B.V. This is an open access article under the CC
BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

Keywords: Climate change; Flood hazard; Thailand; Hydrological modeling; Hydraulic modeling

1. Introduction flood management strategies have been conducted on

its main basin, Chi River Basin, in recent years
Yang River Basin is one of the most flood prone basins (Chaleeraktrakoon and Khwanket, 2013; Artlert et al.,
in Northeast Thailand (Kuntiyawichai et al., 2011a,b). 2013; Kuntiyawichai et al., 2011a,b). These studies
Several studies on climate change impact assessment and reported that climate change is consistent and it has strong
implications on the basin scale hydrological cycle. Other
studies done globally indicate the altercated meteorological
⇑ Corresponding author. Fax: +66 2 524 6425.
variables have great potential to change the frequency and
E-mail addresses: sangamshrestha@gmail.com, sangam@ait.asia
intensity of extreme events specially floods (Dobler et al.,
(S. Shrestha).
Peer review under responsibility of The Gulf Organisation for Research 2012; Viviroli et al., 2011). The increase in temperature
and Development. accelerates the evapotranspiration process which further

https://doi.org/10.1016/j.ijsbe.2016.09.006
2212-6090/Ó 2017 The Gulf Organisation for Research and Development. Production and hosting by Elsevier B.V.
This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
286 S. Shrestha, W. Lohpaisankrit / International Journal of Sustainable Built Environment 6 (2017) 285–298

influences the precipitation amount and ultimately con- resolution makes the data unreliable for basin scale impact
tributes in modification of seasonal runoff. The present assessment studies and is necessary to be bias corrected
intra-annular variability in the amount of runoff is (Muerth et al., 2013). A few studies have been conducted
expected to shift under climate change scenarios at many so far on analysis of different downscaling techniques with
regions of the world including Thailand (Dobler et al., emphasis on extreme events. A comparison study of six
2010). downscaling technique with three RCMs suggests both sta-
In addition to the projected changes in the hydrological tistical and dynamic downscaling tends to have similar
regime, the climate change will also have implications on bias. However, the choice of method of downscaling
the extreme events. Studies have demonstrated that flood depends on variables to be downscaled (Schmidli et al.,
intensity is highly sensitive to temperature in many parts 2007). Leander and Buishand (2007) satisfactorily used
of the world (Prudhomme et al., 2013; Menzel et al., the power law transformation method for RCM outputs
2002; Panagoulia and Dimou, 1997). Several other studies at Western Europe for estimation of extreme events.
also have argued that climate has been a contributing fac- Many studies have adopted various climate change sce-
tor to flood risk by raising the precipitation amount rela- narios to evaluate these effects. The scenarios presented in
tive to the average annual rainfall (Fleming et al., 2012; the Special Report on Emission Scenarios (SRES) in the
Hirabayashi et al., 2008). Therefore, basin scale assessment Intergovernmental Panel on Climate Change (IPCC)
of climate change impacts on flood plays a key role in for- Fourth Assessment Report (AR4) (IPCC, 2007) have been
mulation and evaluation of adaptation and mitigation widely applied to investigate hydrological responses to cli-
strategies for flood risk management. mate change (Praskievicz and Chang, 2011; Moradkhani
Literature suggests that climate change impact assess- et al.; 2010; Ficklin et al., 2009; Tu, 2009; Yoshimura
ment on extreme events has been less investigated and pos- et al., 2009). The Fifth Assessment Report (AR5) of the
sesses higher uncertainty (Dobler et al., 2012). In addition, IPCC published in 2014 includes new scenarios based on
whatsoever the research has been conducted, primary focus various technical developments. These new scenarios,
is on the basin of developed nations (Bauwens et al., 2011; called Representative Concentration Pathways (RCPs),
Prudhomme et al., 2010; Steele-Dunne et al., 2008). Also are a set of greenhouse gas concentration and emission
focusing on Asian countries, many studies on floods pathways designed to support research on the impacts of
induced by climate change have been conducted on several and potential policy responses to climate change (Riahi
basins in China (Li et al., 2013; Zheng et al., 2012; Yang et al., 2011; Van Vuuren et al., 2011; Moss et al., 2010).
et al., 2012). This implies less focus on basins of developing The RCPs are also considered to include impacts caused
countries lying on the tropical regions which are evidently by landuse and land cover (LULC) change.
more susceptible to floods where the region has already The present study is conducted to assess the climate
high precipitation and the hydrologic cycle is highly inter- change impact on flood hazard potential in Yang River
linked and sensitive to its components (Kite, 2001). Basin with the following objectives: (i) to develop
Although considerable studies on floods have been con- rainfall-runoff model of the Yang River Basin, (ii) to design
ducted in Northeast of Thailand yet only few of studies synthetic hydrographs with return periods of 25, 50 and
were on the impact of climate change on extreme events 100 years with regard to future climate conditions, and
(Jothityangkoon et al., 2013; Hunukumbura and (iii) to simulate flood hazard potential representing return
Thailand, 2012). Despite several flood events in Yang River periods of 25, 50 and 100 years under future climate change
Basin most of the studies focused on the management prac- scenarios. This study assumes that land use activities and
tices and socio-economic impacts of floods population remains the same in future. Although many
(Keokhumcheng et al., 2012; Dutta, 2011; Hungspreug GCMs are available, only 3 GCMs and two RCPs were
et al., 2000). Shrestha (2014) studied the climate change selected to construct the future climate scenarios to address
impact on flood hazard potential in Yang River Basin. the uncertainty in climate change projections.
However, the study used the climate change projections
from Special Report on Emission Scenarios (SRES) and 2. Materials and methods
only one Regional Climate Model, which poses greater
uncertainty in flood hazard assessment. Hence the basin 2.1. Study area and data description
scale study of climate change impact on flood hazard using
future climate data from multiple climate models and new The Yang River Basin, a sub-basin of Chi River Basin,
emission scenarios is important in Thailand. has a drainage area of approximately 4145 km2 which
Another important factor that has decisively influenced receives an average annual rainfall of 1390 mm (Fig. 1).
the climate change impact studies is the use of Global Cir- The annual relative humidity and temperature are approx.
culation Models (GCMs) and Regional Climate Models 71% and 26.7 °C respectively in the basin. The basin is
(RCMs) dataset for the future climate projection without influenced by two prominent wind systems, the northeast
bias correction (Cloke et al., 2013). Although RCMs and southwest monsoons which are responsible for the
perform nested dynamic downscaling to the outputs of rainfall patterns and temperature variations. In the north-
the General Circulation Models (GCMs), yet the spatial east monsoons, the dry cold wind picks up some moisture
 S. Shrestha, W. Lohpaisankrit / International Journal of Sustainable Built Environment 6 (2017) 285–298 287

Figure 1. Location map of study area with hydro-meteorological stations in Yang River Basin, Thailand.

from the northeast, it takes place from mid-October to ary. Those geometric data were used to generate DEM of
early February. The southwest monsoons begin around 20-meter resolution. A land cover/land-use map for this
mid-May and fade down by mid-September. In addition study area was obtained from Thailand Land Development
to monsoons, the Yang River Basin also faces tropical Department (LDD) and classified based on the Interna-
storms. The tropical depressions mainly come from the tional Geosphere Biosphere Programme (IGBP). Soil clas-
South China Sea. Consequently, the high moisture travel- sification map is acquired from Food and Agriculture
ing over the water surface causes the heavy rain during Organization (FAO) digital soil map of the world. A brief
the rainy season (Artlert et al., 2013). description of hydrological and meteorological data used in
Topographically, the basin is characterized by the Phu this study is summarized in Table 1.
Phan mountain range at a relatively high elevation of A three step modeling approach i.e. (i) correcting the
around 600 masl, with the Yang River as the major river biasness of the large-scale atmospheric data, (ii) hydrolog-
that flows through Kalasin province, and meets Chi River ical and hydraulic modeling for flood inundation and flood
at Yasothon province. The landuse in this basin consists of hazard analysis and (iii) change analysis was adopted in
agriculture (70%), forest (25%), urban (2%), water bodies this study. Fig. 2 gives an overview of the methodology
(1.2%) and others (1.8%). adopted in this study. Firstly, BTOPMC model and
The DEM, a digital representation of ground surface HEC-RAS model were calibrated based on the observed
topography, was constructed from geometric data acquired climate variables; while in the second step, GCMs outputs
from Thailand Land Development Department (LDD) and were bias-corrected. Thereafter, the future precipitation
Royal Irrigation Department (RID). The geometric data data were used as a forcing to the hydrological and hydrau-
consist of point elevation, 2-meter interval contours, lic model to generate flood inundation areas in the basin in
stream elevation, road elevation and Yang River bound- future.
288 S. Shrestha, W. Lohpaisankrit / International Journal of Sustainable Built Environment 6 (2017) 285–298

Table 1
Brief description of hydro-meteorological stations and data used in this study.
Code Lat Long Elevation (masl) Duration Source
Meteorological stations
A. Selaphum 49092 16°010 5500 103°560 1700 145 1974–2009 RID
Kuchinarai 388002 16°300 0000 104°030 0000 155 1976–2006 TMD
Somde 3888006 16°400 00 103°450 0000 187 1990–2009 TMD
Kuchinarai Self- help Settlement 3888009 16°390 00 103°540 00 183 1976–2011 TMD
Namon Agriculture Office 3888011 16°340 00 103°480 00 177 1990–2009 TMD
Hydrological stations
Thung Kaho Luang E 18 16°010 59 103°540 3800 137 1974–2010 RID
Kuchinarai E 54 16°260 2900 104°020 07 146 1990–2009 RID
Pon Thong E 70 16°170 15 104°000 33 141 1990–2009 RID
Mueang E 2A 15°470 06 104°080 36 127 2007–2008 RID
Maha Chana E 20A 15°310 59 104°150 2400 122 1976–2010 RID
RID: Royal Irrigation Department, TMD: Thai Meteorological Department.

Figure 2. Schematic representation of the methodology used in flood hazard map creation.

2.2. GCM, RCP scenarios and bias correction and strategies to reduce greenhouse gas emissions. The
RCP8.5 is a business as usual scenario and characterized
GCMs with sufficient historical and future datasets of by increasing greenhouse gas emissions and high green-
the representative concentration pathways (RCPs) family house gas concentration levels.
were selected. The selected number allows sufficient assess- Delta change method was used to correct biasness in
ment of uncertainty without exacerbating the computa- outputs of GCMs. This method is preferable when RCM
tional demands of the study. The GCM ensemble herein output is not available for rapid assessment of multiple cli-
includes 3 out of the 60 GCMs in the World Climate mate change scenario or when multi GCM experiments are
Research Programme’s (WCRP’s) Coupled Model Inter- required, as in this case. The delta method assumes the
comparison Project phase5 (CMIP5) dataset, BCC- future model biases for both mean and variability will be
CSM1.1, CCSM4 and MIROC5 (Table 2). The simulations same as that of present day simulations. Changes in the cli-
cover the period from 1975 to 2005 (historical run) and mate are calculated from model results and applied to the
from 2010 to 2099 with IPCC Representative Concentra- baseline data (Rasmussen et al., 2012; Lettenmaier et al.,
tion Pathways 4.5 (RCP4.5) and 8.5 (RCP8.5). The 1999; Gellens and Roulin, 1998).
RCP4.5 is a stabilization scenario where total radioactive Outputs from 3 GCMs for the 1971–2000 (baseline),
forcing is stabilized before 2100 by employing technologies 2010–2039 (2020s), 2040–2069 (2050s) and 2070–2099
 S. Shrestha, W. Lohpaisankrit / International Journal of Sustainable Built Environment 6 (2017) 285–298 289

Table 2
A list of the climate models used in this study with a brief indication of their origin, resolution and the number of realizations available for each climate
change scenario.
Model’s name Resolution (°) Grids (Long-Lat) Scenarios Research center
Beijing Climate Center Climate System Model 2.81  2.81 128  64 RCP4.5 BCC, China Meteorological Administration, China
version 1.1 (BCC-CSM1.1) RCP8.5
The Community Climate System Model version 1.25  0.90 288  192 RCP4.5 National Center for Atmospheric Research, USA
4 (CCSM4) RCP8.5
Model for Interdisciplinary Research on 1.40  1.40 256  128 RCP4.5 Atmosphere and Ocean Research Institute, Japan
Climate-Earth System, version 5 (MIROC5) RCP8.5

(2080s) were used to generate climate change scenarios. precipitation for any sub basin, i, and soil layer, k, and
The 2020s, 2050s and 2080s correspond to near-, mid and time, t is calculated based on Eq. (1).
far future, respectively. Climate scenarios information
was then transferred to the hydrological model and fre- Reði; tÞ ¼ Roði; tÞ  SrmaxðiÞ  Epði; tÞ ð1Þ
quency analysis was performed on the simulated hydrolog-
ical scenarios. The outputs of the hydrological modeling where Re represents the effective rainfall, Ro indicates total
were used as input for flood inundation analysis. precipitation, Srmax is the maximum storage capacity in
the root zone and Ep denotes evapotranspiration.
BTOPMC integrates Shuttleworth–Wallace model to cal-
2.3. Hydrological modeling culate the potential evapotranspiration. Fig. 3 gives the
details of the runoff generation for each grid cells in
The hydrological simulations were carried out using BTOPMC model. Further details are available in Ao
Block-wise use of TOPMODEL with Muskingum–Cunge et al. (2006) and Ishidaira et al. (2005). BTOPMC model
method (BTOPMC) model. It is an extension of the TOP- has been successfully applied in many basins worldwide
MODEL concepts (Beven et al., 1995), which is developed with satisfactory performance (Manandhar et al., 2013;
in order to overcome the limitations of using the TOPMO- Phan et al., 2010; Bao et al., 2010; Wan et al., 2009;
DEL for large river basins. For large river basins, spatial Shrestha et al., 2007). In addition, it has also been applied
heterogeneity and timing of flow to outlet are the impor- successfully to Mekong River basin which validates the
tant factors. For representing spatial variability in model’s ability to represent the hydrology of the basin
BTOPMC, a catchment is composed of grid cells, which (Kiem et al., 2008; Hapuarachchi et al., 2008).
can be divided into sub-catchments, where each sub- The model was calibrated by adjusting saturated transi-
catchment is considered as a block or a unit. The effective tivity, decay coefficient and rooting depth for the period of

P = gross rainfall

Pa = net rainfall on land surface

ET0 = interception evaporation

ET = actual evapotranspiration

Sr max= maximum root zone capacity

Srz = root zone storage

SD = local saturation deficit

Suz = unsaturated zone storage

qv = groundwater recharge

qb = base flow

Figure 3. Soil water balance and runoff generation in BTOPMC for each grid cell.
290 S. Shrestha, W. Lohpaisankrit / International Journal of Sustainable Built Environment 6 (2017) 285–298

Table 3
Hazard index for depth of flooding.
Depth of flooding (m) Degree of flood hazard Description HI
D > 3.50 Very high ‘‘Extreme danger: flood zone with deep fast flowing water” 4
1.00–3.5 High ‘‘Danger: flood zone with deep fast flowing water” 3
0.60–1.00 Moderate ‘‘Danger: flood zone with deep fast flowing water” 2
D < 0.60 Low ‘‘Flood zone with shallow flowing water or deep standing water” 1

2002–2006. The calibrated parameters were used as input analysis provided the probable maximum runoff with 25,
for the validation period of 1997–2001 in order to check 50 and 100 year return period. The dimensionless hydro-
for the best goodness-of-fit with the observed discharge graphs are used to derive synthetic inflow hydrographs
at Kuchinarai, Kalasin station (E54). for the above mentioned return periods.

2.4. Frequency analysis and designed hydrographs

2.5. Hydraulic modeling
In order to analyze and design the frequency of probable
maximum stream flow, annual maximum runoffs of 1980– Flood inundation areas were simulated using HEC-
2009 were used and a distributed hydrological model was RAS, a 1-D hydraulic model innovated by the US Army
used to simulate the annual maximum runoffs for the Corps of Engineers, under unsteady flow conditions (U.S.
future periods of 2020s, 2050s and 2080s. A Gumbel distri- Corps of Engineers, 2002). The topography of channel
bution method was applied to illustrate the extreme flood and floodplain information was derived from HEC-
frequency. For the flood inundation modeling under cli- GeoRAS software which is used as an extension in Arc-
mate change conditions, several return periods such as GIS, for processing geospatial data for use with HEC-
25, 50 and 100 years of stream flow are considered. RAS.
Designed synthetic hydrographs were developed by Discharge and river water level were set as upstream and
applying two methods, the dimensionless hydrograph downstream boundaries. The upstream boundary in this
method and the flood frequency analysis. The dimension- study is at Kuchinarai station (E.54)—a stream-flow gaug-
less hydrograph was derived from the extreme runoff pat- ing station. In order to take backwater effect into the Yang
tern in 2007. For flood frequency analysis, Gumbel’s River model, the water level at the junction between Yang
distribution was selected to analyze the historical runoff River and Chi River was also calculated by HEC-RAS
data recorded from 1980 to 2009. Moreover, the Gumbel model. Based on the availability of spatial extent of water
probability distribution was also used to predict future run- and flood level in the basin, A. Pon Thong (E.70) station
off simulated by the distributed hydrological model by was selected to setup the model. The model was calibrated
using the bias corrected future climate data. The frequency by adjusting manning’s roughness coefficient (n) in order to
give the best goodness-of-fit for the period from 1 July to
30 November 2005. A period from 1 July to 30 November
Table 4
2007 was used for model validation.
Relative change in annual precipitation in future periods compared to
baseline period of 1971–2000 in the Yang River Basin. The BTOPMC model performance was evaluated by
computing Nash-Sutcliffe Efficiency (NSE) (Nash and
Periods Scenarios Annual precipitation (mm)
Sutcliffe, 1970), Coefficient of Determination (R2) Volume
Max Avg. Min
ratio (Vr). Similarly, HEC-RAS model performance was
2020s RCP4.5 104.00 41.61 25.39 evaluated by computing Coefficient of Determination
RCP8.5 74.53 32.44 21.47
(R2) Volume Bias (VB) and Relative Peak Error (RPE).
2050s RCP4.5 109.05 49.89 37.80 The Nash–Sutcliffe Efficiency (NSE) is defined as:
RCP8.5 105.47 42.60 38.79
Pn
2080s RCP4.5 37.03 25.13 8.58 ðX obs;i  X sim;i Þ2
NSE ¼ 1  Pi¼1 n 2
ð2Þ
RCP8.5 67.23 37.91 37.70
i¼1 ðX obs;i  X sim;i Þ

Table 5
Percentage changes in annual average precipitation in future periods compared to baseline period (1971–2000) in the Yang River Basin.
GCMs Percentage change as compared to baseline
RCP4.5 RCP8.5
2020s 2050s 2080s 2020s 2050s 2080s
BCCCSM1.1 47.00 48.50 17.85 39.19 47.69 24.68
CCSM4 26.46 30.14 6.90 21.9 25.11 36.81
MIROC5 18.00 24.64 24.26 15.66 26.92 24.58
 S. Shrestha, W. Lohpaisankrit / International Journal of Sustainable Built Environment 6 (2017) 285–298 291

where X obs and X sim are the observed and simulated vari- The Volume Bias (VB) is expressed as follows:
ables, respectively, X obs is the average observed discharge  Pn 
i¼1 ðX obs;i  X sim;i Þ
and n is the number of time step. VB ¼ Pn ð4Þ
The Coefficient of Determination (R2) is calculated as: i¼1 X obs;i

0 12 where X obs and X sim are the observed and simulated climate
Pn
i¼1 ðX obs;i  X obs ÞðX sim;i  X sim Þ
B C variables at time i, respectively, and n is the number of time
R2 ¼ @qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiA
Pn 2 Pn 2 step.
i¼1 ðX obs;i  X obs Þ i¼1 ðX sim;i  X sim Þ The Volume Ratio (Vr) indicates the relation between
ð3Þ simulated discharges and observed discharges and calcu-
lated as below:
where X obs and X sim are the observed and simulated climate Pn
variables at time i, respectively, X obs is the average observed X sim;i
V r ¼ Pi¼1
n ð5Þ
climate variables and n is the number of time step. i¼1 X obs;i

Figure 4. Changes in annual average Tmax (a, b & c) and Tmin (d, e & f) for the 2020s, 2050s and 2080s relative to the baseline period (1975–2005) under
RCP4.5 and RCP8.5 in the Yang River Basin. The line in the shaded box shows the median value, and the bottom and top of the box show the 25th and
75th percentiles, and the whiskers show the minimum and maximum values of temperatures.
292 S. Shrestha, W. Lohpaisankrit / International Journal of Sustainable Built Environment 6 (2017) 285–298

where X obs and X sim are the observed and simulated climate where X obs and X sim are the observed and simulated climate
variables at time i, respectively, and n is the number of time variables at time i, respectively, and n is the number of time
step. step.
The Relative Peak Error (RPE) is applied to measure
the relative difference between observed and simulated 2.6. Flood frequency analysis
peaks. It is calculated as:
The objective of flood frequency analysis was to relate
X obs  X sim
RPE ¼  100% ð6Þ the magnitude of floods to their frequency of occurrence
X obs using probability distribution. In order to commence this
objective, the calculation of the statistical parameters of
Table 6 the proposed distribution was made by the method of
Performance of BTOPMC in Yang River basin.
moments from the given data. In this study the annual
Period NSE (%) R2 Vr maximum runoffs for four different periods (1990s, 2020s,
Calibration (2002–2006) 62.80 0.63 0.97 2050s and 2080s) were calculated for different return period
Validation (1997–2001) 66.45 0.68 1.12 floods.

Simulated Discharge Observed Discharge Precipitation

450 0

400
20
350

Precipitation (mm)
300 40
Discharge (m³/s)

250
60
200

150 80

100
(a) 100
50

0 120
1/1/2002 1/1/2003 1/1/2004 12/31/2004 12/31/2005

Simulated Discharge Observed Discharge Precipitation

450 0

400
20
350
Precipitation (mm)
Discharge (m³/s)

300 40

250
60
200

150 80

100
(b) 100
50

0 120
1/1/1997 1/1/1998 1/1/1999 1/1/2000 12/31/2000 12/31/2001

Figure 5. Comparison of observed and simulated discharge at Kuchinarai, Kalasin (E54) (a) calibration period (1 Jan 2002 to 31 Dec 2006) and (b)
validation period (1 Jan 1997 to 31 Dec 2001).
 S. Shrestha, W. Lohpaisankrit / International Journal of Sustainable Built Environment 6 (2017) 285–298 293

2.7. Flood hazard mapping Considering the median values of ensembles, the highest
increases in temperature were predicted for the RCP8.5,
Flood hazard mapping is to determine areas with a while the lowest were predicted for RCP 4.5. The differ-
probability of a flooding event for a defined return period ences between the predicted changes in temperature for
(Han, 2011). With the results of hydraulic calculations, the the different emissions scenario will also increase in the
flood outline can be calculated. The degree of flood hazard future. This sharp difference is because of the significant
depends on several hydrological factors such as flood increase in differences among the different emission scenar-
velocity and inundated depths. Since this study applied ios themselves. The median values of the projections justify
the 1-D hydraulic model, the hazard index was assigned this finding. During 2080s, for Tmax, the median value as
with corresponding to different inundated depths. The projected under the RCP8.5 scenario is 1.1 °C higher than
degree of flood hazard was classified into four hazard cat- as projected under the RCP4.5 scenario shown in the fol-
egories based on inundation depth classes corresponding to lowing box plot (Fig. 4). Similarly, during the same period,
three critical inundated depths 0.6, 1.0 and 3.5 m as sug- change in Tmin is 2.2 °C higher for RCP8.5 than RCP4.5
gested by Tu and Tingsanchali (2010) which is described as projected by median values.
in Table 3.
3.1.1. Performance of the BTOPMC
During the calibration, BTOPMC uses observed rainfall
3. Results and discussion data of 2002–2006 as input into a systematic search for
model parameters which produce the best goodness-of-fit
3.1. Changes in mean temperature and rainfall between the simulated discharge and observed discharge
at Kuchinarai, Kalasin (E54) station. Thereafter, 5 years
Table 4 shows the absolute changes in annual rainfall from 1997 to 2001 were used for validation purposes where
from ensemble of three GCMs and Table 5 shows the per- both calibration and validation periods cover extreme
centage change in rainfall for the 2020s, 2050s and 2080s in flood events in the study area. The performance of the
comparison to baseline period (1971–2000). The annual model during calibration and validation period is summa-
rainfall is projected to increase in near and mid future rized in Table 6 and Fig. 5. The low flow periods are under-
and comparatively higher under RCP 4.5 whereas the rain- estimated during calibration whereas overestimated during
fall is projected to increase in the lower magnitude in far validation for the first three years followed by underestima-
future under both scenarios (Table 4). Similarly, percent- tion. In addition, it can also be noticed that the model is
age change in annual rainfall differs among the GCMs, able to simulate the low peaks very well however it is
higher change is projected by BCCCSM1.1 and lower by unable to fetch the high peak flow during heavy rainfall
MIROC5 under both RCP scenarios (Table 5). events during validation period. Nonetheless, the model
As shown in Fig. 4, the minimum and maximum annual evaluation by NSE, R2 and Vr indicates that the model per-
temperature was projected to increase in all future periods. forms reasonably well in the basin.

Table 7
Probable peak discharge estimated for RCP 8.5 and RCP 4.5 of three GCMs at Kuchinarai station.
Time period Scenario GCM 25 years Relative change (%) 50 years Relative change (%) 100 years Relative change (%)
Baseline 552 612 670
2020s RCP 8.5 BCCCSM1.1 702 27 809 32 915 37
CCSM4 692 25 790 29 870 30
MIROC5 653 18 750 23 800 19
RCP 4.5 BCCCSM1.1 684 24 780 28 900 34
CCSM4 663 20 760 24 880 31
MIROC5 651 18 745 22 850 27
2050s RCP 8.5 BCCCSM1.1 721 31 852 39 980 46
CCSM4 712 29 832 36 954 42
MIROC5 702 27 802 31 930 39
RCP 4.5 BCCCSM1.1 691 25 780 28 900 34
CCSM4 678 23 760 24 856 28
MIROC5 653 18 742 21 843 26
2080s RCP 8.5 BCCCSM1.1 750 36 901 47 1040 55
CCSM4 727 32 880 44 990 48
MIROC5 700 27 847 38 950 42
RCP 4.5 BCCCSM1.1 708 28 856 40 985 47
CCSM4 690 25 840 37 960 43
MIROC5 650 18 813 33 912 36
294 S. Shrestha, W. Lohpaisankrit / International Journal of Sustainable Built Environment 6 (2017) 285–298

3.1.2. Performance of hydrological analysis and design 3.1.3. Performance of HEC-RAS model
3.1.2.1. Flood frequency analysis. Table 7 illustrates the The calibration and validation statistics suggests the
probable peak discharge at Kuchinarai station under model simulates the observed flood in good agreement
RCP 4.5 and RCP 8.5 scenarios. Maximum change in dis- (Table 8). Fig. 6 demonstrates the simulated water level
charge is observed for 100 years return period for all the compared to the observed water level at Pon Thing station
three future periods under both scenarios. The highest rel- (E.70). The validation results suggest the maximum
ative change was observed under RCP 8.5 in 2080s whereas observed water surface level at 141 masl on 7 October,
the lowest relative change was observed under RCP 4.5 in 2007. The simulated water level is 139 masl on the same
2020s. day illustrating the model’s ability to fetch the time to peak
at the same time.
3.1.2.2. Designed hydrographs. Synthetic hydrograph gener-
ated for different return periods at Kuchinarai station for
baseline period suggests maximum flood is expected for 3.2. Projected changes in floods
100 year return period followed by 50 and 25 years. In
addition, it is also observed that the time to peak for all 3.2.1. Flood inundated areas for baseline period (1980–
the return period floods follows same trend as that of the 2009)
observed hydrograph. Moreover, maximum variation in Table 9 illustrates the areas of inundation under differ-
the magnitude of the designed hydrographs is observed at ent degrees of flood hazard. Based on the assigned degree
the peak. of flood hazard, although all class of flood hazards shows
an increase trend, yet the spatial coverage of high flood
depth is observed to be maximum under present climate.
In case of moderate and high flood depth events highest
Table 8
Performance of HEC-RAS model in Yang River. spatial coverage is observed for 25 year return period. In
addition, very high flood depth is noted to have maximum
Period R2 RPE VB
coverage of 38.69 km2 which includes 16.15% of total area
Calibration (2005) 0.96 0.71 0.01
for 100-year return period. It is also noteworthy that the
Validation (2007) 0.92 0.94 0.01
trend in the area of inundation for degree of flood hazard

142
Observed Simulated
141
140
Water Level (masl)

139
138
137
(a)
136
135
134
133
132
131
30-Jun 20-Jul 9-Aug 29-Aug 18-Sep 8-Oct 28-Oct 17-Nov
Year (2005)

142
Observed Simulated
141
140
Water Level (masl)

139
138
137
(b)
136
135
134
133
132
131
30-Jun 20-Jul 9-Aug 29-Aug 18-Sep 8-Oct 28-Oct 17-Nov
Year (2007)

Figure 6. Comparison of observed (blue) and simulated (red) water level at Pon Thong station (E.70) during (a) calibration (2005) and (b) validation
(2007) of HEC-RAS model.
 S. Shrestha, W. Lohpaisankrit / International Journal of Sustainable Built Environment 6 (2017) 285–298 295

Table 9 reduces from 25- to 100-years return period for low, mod-
Flood inundation areas in different return period floods for baseline erate and high flood depth. However, for very high flood
period.
depth, area of inundation is observed to follow increasing
Degree of flood hazard Simulated flood area in km2 (in %) trend for the corresponding return periods. For the total
25-year 50-year 100-year area of inundation, flood of 100-year return period has
Low 25.44 (12.38) 28.01 (12.51) 28.79 (12.02) the maximum area of inundation about 239.51 km2 imply-
Moderate 19.25 (9.37) 17.32 (7.73) 17.73 (7.40) ing the severity of extreme floods in the region under pre-
High 141.66 (68.95) 151.63 (67.71) 154.31 (64.42) sent climate.
Very high 19.10 (9.30) 26.99 (12.05) 38.69 (16.15)
Total 205.44 223.94 239.51
The different landuse under flood inundation was also
analyzed and found that cropland areas are significantly

Table 10
Flood inundation depth generated by HEC-RAS – landuse relationship for the baseline period in 25, 50 and 100 year return period floods.
Return Low Moderate High Very high Total area
period Land use D < 0.6 0.6 < D < 1.0 1.0 < D < 3.5 D > 3.5
(km2) (km2) (km2) (km2) (km2)
25-year Forest 0.88 0.56 9.81 4.06 15.30
Croplands 21.69 16.52 114.47 6.78 159.46
Grasslands 1.67 1.45 16.04 8.23 27.38
Urban 1.19 0.72 1.35 0.03 3.30
Water Bodies 1.64 2.21 38.96 11.48 54.29
Total 25.44 19.25 141.66 19.10 205.44
50-year Forest 1.05 0.64 7.85 6.55 16.08
Croplands 24.28 14.74 125.86 10.68 175.56
Grasslands 1.41 1.16 15.92 9.72 28.21
Urban 1.27 0.78 2.00 0.03 4.08
Water Bodies 0.97 1.28 35.17 17.41 54.84
Total 28.01 17.32 151.63 26.99 223.94
100-year Forest 1.26 0.63 6.66 8.30 16.84
Croplands 24.47 15.17 130.20 18.63 188.47
Grasslands 1.27 1.02 14.87 11.71 28.88
Urban 1.79 0.90 2.58 0.05 5.32
Water Bodies 0.73 0.77 29.89 23.84 55.23
Total 28.79 17.73 154.31 38.69 239.51

Table 11
Flood hazard areas estimated from various return periods under ensembles of RCP 8.5 and RCP 4.5 in Yang River Basin.
Time Return Scenarios Low Moderate High Very high Total area
period period D < 0.6 0.6 < D < 1.0 1.0 < D < 3.5 D > 3.5
(km2) (km2) (km2) (km2) (km2)
2020s 25 RCP 8.5 30 18 150 45 243
RCP 4.5 16 18 142 43 218
50 RCP 8.5 38 18 147 74 277
RCP 4.5 20 14 151 112 297
100 RCP 8.5 25 18 143 140 325
RCP 4.5 17 13 117 130 277
2050s 25 RCP 8.5 27 17 141 106 291
RCP 4.5 20 18 132 102 272
50 RCP 8.5 30 15 156 123 324
RCP 4.5 21 13 146 106 286
100 RCP 8.5 32 16 103 142 293
RCP 4.5 22 14 92 91 219
2080s 25 RCP 8.5 32 18 142 93 284
RCP 4.5 20 17 141 96 274
50 RCP 8.5 36 17 141 95 289
RCP 4.5 26 11 132 121 290
100 RCP 8.5 40 11 131 122 304
RCP 4.5 30 10 118 143 300
296 S. Shrestha, W. Lohpaisankrit / International Journal of Sustainable Built Environment 6 (2017) 285–298

affected. The area increases from 159.46 to 188.47 km2 for (2020s) for both climate scenarios which signifies the severe
25 to 100 year return period (Table 10). It is also evident threat of flood in future under climate change. The output
that, only change in inundation area for high and very high of this study not only indicates the severity of flood in the
flood depth influences the total area for croplands indicat- region but also indicates potential damage in food produc-
ing agricultural vulnerability. This implies that farming tion and negative effects on the livelihoods of local people.
lands acts as retention areas for flood water in the region. Therefore, proper land use planning and risk-based design
It can also be noticed that forests experiences high and very of hydraulic structures must be integral part of mitigation
high flood depth of all return periods compared to low and plan when addressing vulnerabilities to future flood dam-
moderate flood depth. In addition, the spatial extent of ages in the basin.
high flood depth declines with increased return period.
Acknowledgements
3.2.2. Changes in future flood inundation area
The simulation was carried out for the present and
The authors would like to appreciate Dr. Jun Magome
future scenarios of extreme rainfall events. Future flood
and the BTOPMC Development Team of Yamanashi
inundation areas were simulated for 2020s, 2050s and
University, Japan for providing the model to conduct this
2080s with respect to RCP 4.5 and RCP 8.5 scenarios
study.
(Table 11). It is noticed that under RCP 8.5 scenarios,
the spatial coverage of flooded area increases noticeably
for all return periods and three future periods relative to References
the baseline period. The total flooded areas ranges from
243 to 291 km2 for 25 year return period, 277 to 324 km2 Ao, T., Ishidaira, H., Takeuchi, K., Kiem, A.S., Yoshitari, J., Fukami, K.,
Magome, J., 2006. Relating BTOPMC model parameters to physical
for 50 year return period and 293 to 325 km2 for 100 year
features of MOPEX basins. J. Hydrol. 320, 84–102.
return period. Under RCP 4.5 scenarios the total areas Artlert, K., Chaleeraktrakoon, C., Nguyen, V.T.V., 2013. Modeling and
under floods are 218–274 km2 for 25 year return period, analysis of rainfall processes in the context of climate change for
286–290 km2 under 50 year return period and 219 to Mekong, Chi and Mun River Basins (Thailand). J. Hydro-Environ.
300 km2 under 100 year return period. Res. 7 (1), 2–17.
Bao, H.J., Wang, L.L., Li, Z.J., Zhao, L.N., Zhang, G.P., 2010.
Similar studies done in Mekong delta region of Vietnam
Hydrological daily rainfall-runoff simulation with BTOPMC model
suggest under climate change high and very high floods are and comparison with Xin’anjiang model. Water. Sci. Eng. 3 (2), 121–
expected to be more intense (Dinh et al., 2012). Moreover, 131.
studies on streamflow under climate change on Mekong Bauwens, A., Sohier, C., Degré, A., 2011. Hydrological response to
river basin suggests the frequency of peak discharge is climate change in the Lesse and the Vesdre catchments: contribution of
a physically based model (Wallonia, Belgium). Hydrol. Earth Syst. Sci.
expected to change from 75 year return period to 25 year
15 (6), 1745–1756.
for future climate by 2045s (Lauri et al., 2012) which Beven, K.J., Lamb, R., Quinn, P.F., Romanowicz, R., Freer, J., 1995.
implies the results of the present research are in line with TOPMODEL. In: Singh, V.P. (Ed.), Computer Models of Watershed
other studies done in the region. Hydrology. Water Resources Publications, pp. 627–668.
Chaleeraktrakoon, C., Khwanket, U., 2013. A statistical downscaling
model for extreme daily rainfalls at a single site. In: World Environ-
4. Conclusions
mental and Water Resources Congress 2013: Showcasing the Future –
Proceedings of the 2013 Congress, pp. 1058–1067.
In this study, climate change impacts on flood hazard Cloke, H.L., Wetterhall, F., He, Y., Freer, J.E., Pappenberger, F., 2013.
potential was examined using three GCMs and hydrologi- Modelling climate impact on floods with ensemble climate projections.
cal and hydraulic models in the Yang River Basin of Thai- Quart. J. Roy. Met. Soc. 139, 282–297.
Dinh, Q., Balica, S., Popescu, I., Jonoski, A., 2012. Climate change impact
land. The hydrological and hydraulic models were
on flood hazard, vulnerability and risk of the Lonng Xuyen
calibrated and validated satisfactorily against runoff data Quadrangle in the Mekong Delta. Int. J. River Basin Manage. 10
and future flows were generated. There was a considerable (1), 103–120.
uncertainty in the results; however, several inferences can Dobler, C., Bürger, G., Stötter, J., 2012. Assessment of climate change
be made based on this study. Generally, temperature and impacts on flood hazard potential in the Alpine Lech watershed. J.
Hydrol. 460–461, 29–39.
rainfall is expected to increase in future. Different GCMs
Dobler, C., Stötter, J., Schöberl, F., 2010. Assessment of climate change
projected the increase in temperature and rainfall in differ- impacts on the hydrology of the Lech Valley in northern Alps. J. Water
ent range. Area inundated by flood under future climate Clim. Change 1 (3), 207–218.
suggests 100 year return period floods are most severe. It Dutta, D., 2011. An integrated tool for assessment of flood vulnerability
showed that approximately 300–325 km2 area will be of coastal cities to sea-level rise and potential socio-economic impacts:
a case study in Bangkok, Thailand. Hydrol. Sci. J. 56 (5), 805–823.
flooded under future climate change scenarios of which sig-
Ficklin, D.L., Luo, Y., Luedeling, E., Zhang, M., 2009. Climate change
nificant areas are under croplands. In addition, for a partic- sensitivity assessment of a highly agricultural watershed using SWAT.
ular return period of flood, a shift from very high under J. Hydrol. 374 (1–2), 16–29.
baseline period to moderate flood is expected for future cli- Fleming, K.M., Tregoning, P., Kuhn, M., Purcell, A., McQueen, H., 2012.
mate under RCP 8.5 scenario. Furthermore, 100 year The effect of melting land-based ice masses on sea-level around the
Australian coastline. Aust. J. Earth Sci. 59 (4), 457–467.
return period flood is expected to be 25 years in near future
 S. Shrestha, W. Lohpaisankrit / International Journal of Sustainable Built Environment 6 (2017) 285–298 297

Gellens, D., Roulin, E., 1998. Streamflow response of Belgian catchments Moradkhani, H., Meier, M., 2010. Long-lead water supply forecast using
to IPCC climate change scenarios. J. Hydrol. 210, 242–258. large-scale climate predictors and independent component analysis. J.
Han, D., 2011. Flood Risk Assessment and Management. Bentham Hydrol. Eng. 15, 744–762.
Publishers. Moss, R.H., Edmonds, J.A., Hibbard, K.A., Manning, M.R., Rose, S.K.,
Hapuarachchi, H.A.P., Zhou, M.C., Kiem, A.S., Geogievsky, M.V., van Vuuren, D.P., Carter, T.R., Emori, S., Kainuma, M., Kram, T.,
Magome, J., Ishidaira, H., 2008. Investigation of the Mekong River Meehl, G.A., Mitchell, J.F., Nakicenovic, N., Riahi, K., Smith, S.J.,
basin hydrology for 1980–2000 using YHyM. Hydrol. Process. 22, Stouffer, R.J., Thomson, A.M., Weyant, J.P., Wilbanks, T.J., 2010.
1246–1256. The next generation of scenarios for climate change research and
Hirabayashi, Y., Kanae, S., Emori, S., Oki, T., Kimoto, M., 2008. Global assessment. Nature 463 (7282), 747–756.
projections of changing risks of floods and droughts in a changing Muerth, M.J., St-Denis, B.G., Ricard, S., Velázquez, J.A., Schmid, J.,
climate. Hydrol. Sci. J. 53 (4), 754–772. Minville, M., Caya, D., Chaumount, D., Ludwig, R., Turcotte, R.,
Hungspreug, S., Khao-uppatum, W., Thanopanuwat, S., 2000. Flood 2013. On the need for bias correction in regional climate scenarios to
management in Chao Phraya river basin. In: Chao, The. (Ed.), Phraya assess climate change impacts on river runoff. Hydrol. Earth Syst. Sci.
Delta: Historical Development, Dynamics and Challenges of Thai- 17, 1189–1204.
land’s Rice Bowl: Proceedings of the International Conference, 12–15 Nash, J.E., Sutcliffe, J.V., 1970. River flow forecasting through conceptual
December 2000, Kasetsart University, Bangkok. Kasetsart University, models 1: Discussion of principles. J. Hydrol. 10, 282–290.
Bangkok, Thailand. Panagoulia, D., Dimou, G., 1997. Sensitivity of flood events to global
Hunukumbura, P.B., Tachikawa, Y., 2012. River discharge projection climate change. J. Hydrol. 191 (1–4), 208–222.
under climate change in the Chao Phraya River Basin, Thailand, using Phan, T.T.H., Sunada, K., Oishi, S., Sakamoto, Y., 2010. River discharge
the MRI-GCM3. 1S Dataset. J. Meteorol. Soc. Jpn. 90(A), 137–150. in the Kone River Basin (Central Vietnam) under climate change by
IPCC, 2007. Climate Change 2007: The Physical Science Basis. Contri- applying the BTOPMC distributes hydrological model. J. Water Clim.
bution of Working Group I to the Fourth Assessment Report of the Change 1 (4), 269–279.
Intergovernmental Panel on Climate Change [Solomon, S., Qin, D., Praskievicz, S., Chang, H., 2011. Impacts of climate change and urban
Manning, M., Marquis, M., Averyt, K., Tignor, M., Miller, H.L., development on water resources in the Tualatin River basin, Oregon.
Chen, Z. (Eds.)]. Cambridge University Press, Cambridge, U.K. and Annals Assoc. Am. Geogr. 101 (2), 249–271.
New York, N.Y., U.S.A. Prudhomme, C., Crooks, S., Kay, A.L., Reynard, N., 2013. Climate
Ishidaira, H., Takeuchi, K., Magome, J., Hapuarachchi, P., Zhou, M., change and river flooding: part 1 classifying the sensitivity of British
2005. Application of distributed hydrological model YHyM to large catchments. Clim. Change 119, 933–948.
river basins in Southeast Asia. In: Proceedings of International Prudhomme, C., Wilby, R.L., Crooks, S., Kay, A.L., Reynard, N.S., 2010.
Symposium on Southeast Asian Water Environment, vol. 3. Scenario-neutral approach to climate change impact studies: applica-
Jothityangkoon, C., Hirunteeyakul, C., Boonrawad, K., Sivapalan, M., tion to flood risk. J. Hydrol. 390 (3–4), 198–209.
2013. Assessing the impact of climate and land use changes on extreme Rasmussen, J., Sonnenborg, T.O., Stisen, S., Seaby, L.P., Christensen, B.
floods in a large catchment. J. Hydrol. 490, 88–105. S.B., Hinsby, K., 2012. Climate change effects on irrigation demands
Keokhumcheng, Y., Tingsanchali, T., Clemente, R.S., 2012. Flood risk and minimum stream discharge: impact of bias-correction method.
assessment in the region surrounding the Bangkok Suvarnabhumi Hydrol. Earth Syst. Sci. 16, 4675–4691.
Airport. Water Int. 37 (3), 201–217. Riahi, K., Krey, V., Rao, S., Chirkov, V., Fischer, G., Kolp, P.,
Kiem, A.S., Ishidaira, H., Hapuarachchi, H.P., Zhou, M.C., Hirabayashi, Kindermann, G., Nakicenovic, N., Rafai, P., 2011. RCP-8.5: exploring
Y., Takeuchi, K., 2008. Future hydroclimatology of the Mekong River the consequence of high emission trajectories. Clim. Change 109, 33.
basin simulated using the high-resolution Japan Meteorological Schmidli, J., Goodess, C.M., Frei, C., Haylock, M.R., Hundecha, Y.,
Agency (JMA) AGCM. Hydrol. Process. 22 (9), 1382–1394. Ribalaygua, J., Schmith, T., 2007. Statistical and dynamical down-
Kite, G., 2001. Modeling the Mekong: hydrological simulation for scaling of precipitation: an evaluation and comparison of scenarios for
environmental impact studies. J. Hydrol. 253 (1–4), 1–13. the European Alps. J. Geophys. Res. https://doi.org/10.1029/
Kuntiyawichai, K., Schultz, B., Uhlenbrook, S., Suryadi, F.X., Griensven, 2005JD007026.
A.V., 2011a. Comparison of flood management option for the Yang Shrestha, S., 2014. Climate Change Impacts and Adaptation in Water
River Basin, Thailand. Irrig. Drain. 60, 526–543. Resources and Water Use Sectors: Case Studies from Southeast Asia.
Kuntiyawichai, K., Schultz, B., Uhlenbrook, S., Suryadi, F.X., Corzo, G. Springer International Publishing Switzerland.
A., 2011b. Comprehensive flood mitigation and management in the Hi Shrestha, S., Bastola, S., Babel, M.S., Dulal, K.N., Magome, J.,
River Basin, Thailand. Lowland Tech. Int. 13 (1), 10–18. Hapuarachchi, H.A.P., Kazama, F., Ishidaira, H., Takeuchi, K.,
Lauri, H., de Moel, H., Ward, P.J., Räsänen, T.A., Keskinen, M., 2007. The assessment of spatial and temporal transferability of a
Kummu, M., 2012. Future changes in Mekong River hydrology: physically based distributed hydrological model parameters in different
impact of climate change and reservoir operation on discharge. physiographic regions of Nepal. J. Hydrol. 347 (1–2), 153–172.
Hydrol. Earth Syst. Sci. 16, 4603–4619. Steele-Dunne, S., Lynch, P., McGrath, R., Semmler, T., Wang, S.,
Leander, R., Buishand, T.A., 2007. Resampling of regional climate model Hanafin, J., Nolan, P., 2008. The impacts of climate change on
output for the simulation of extreme river flows. J. Hydrol. 332, 487– hydrology in Ireland. J. Hydrol. 356 (1–2), 28–45.
496. Tu, J., 2009. Combined impact of climate and land use changes on
Lettenmaier, D.P., Wood, A.W., Palmer, R.N., Wood, E.F., Stakhiv, E. streamflow and water quality in eastern Massachusetts, USA. J.
Z., 1999. Water resources implications of global warming: a U.S. Hydrol. 379, 268–283.
regional perspective. Clim. Change 43, 537–579. Tu, V., Tingsanchali, T., 2010. Flood hazard and risk assessment of
Li, Y., Guo, Y., Yu, G., 2013. An analysis of extreme flood events during Hoang Long River basin, Vietnam. In: Proceedings of the Interna-
the past 400 years at Taihu Lake, China. J. Hydrol. 500, 217–225. tional MIKE by DHI Conference.
Manandhar, S., Pandey, V.P., Ishidaira, H., Kazama, F., 2013. Pertur- U.S. Corps of Engineers, 2002. HEC-RAS River Analysis System,
bation study of climate change impacts in a snow-fed river basin. Hydraulic Reference Manual. Hydraulic Engineering Center Report
Hydrol. Process. 27 (24), 3461–3474. CPD-69, Davis, CA.
Menzel, L., Niehoff, D., Bürger, G., Bronstert, A., 2002. Climate change van Vuuren, D.P., Edmonds, J., Kainuma, M.L.T., Riahi, K., Thomson,
impacts on river flooding: a modelling study of three meso-scale A., Matsui, T., Hurtt, G., Lamarque, J.-F., Meinshausen, M., Smith,
catchments. In: Beniston, M. (Ed.), Climatic Change: Implications for S., Grainer, C., Rose, S., Hibbard, K.A., Nakicenovic, N., Krey, V.,
the Hydrological Cycle and for Water Management, 10 Springer, Kram, T., 2011. Representative concentration pathways: An overview.
Netherlands, pp. 249–269. Clim. Change 109, 213.
298 S. Shrestha, W. Lohpaisankrit / International Journal of Sustainable Built Environment 6 (2017) 285–298

Viviroli, D., Archer, D.R., Buytaert, W., Fowler, H.J., Greenwood, G.B., Yang, C., Yu, Z., Hao, Z., Zhang, J., Zhu, J., 2012. Impact of climate
Hamlet, A.F., Huang, Y., Koboltschnig, G., Litaor, M.I., Lopez- change on flood and drought events in Huaihe River Basin, China.
Moreno, J.I., Lorentz, S., Schädler, B., Schreier, H., Schwaiger, K., Hydrol. Res. 43 (1–2), 14–22.
Vuille, M., Woods, R., 2011. Climate change and mountain water Yoshimura, C., Zhou, M., Kiem, A.S., Fukami, K., Hapuarachchi, H.A.,
resources: overview and recommendations for research, management Ishidaira, H., Takeuchi, K., 2009. 2020s scenario analysis of nutrient
and policy. Hydrol. Earth Syst. Sci. 15 (2), 471–504. load in the Mekong River Basin using a distributed hydrological
Wan, Y.A., Yin, F., Liu, Z.Z., Cui, W.C., Ao, T.Q., 2009. Study on the model. Sci .Total Environ. 407 (20), 5356–5366.
effect of DEM spatial resolution and sampling algorithms on runoff Zheng, P., Li, Z., Bai, Z., Wan, L., Li, X., 2012. Influence of climate
simulation by BTOPMC. IAHS-AISH Publ. 335, 130–136. change to drought and flood. Disaster Adv. 5 (4), 1331–1334.