Statically Downscaling using different Representative Concentration Pass ways
of Emission Scenario; in the Case Wolikite, South West Ethiopia
Ethiopian Environment and Forest Research Institute, Ethiopia
Submission: July 24, 2020; Published: August 07, 2020
*Corresponding author: Moges Molla, Ethiopian Environment and Forest Research Institute, Hawassa Center, Hawassa, Ethiopia.
How to cite this article: Moges M. Statically Downscaling using different Representative Concentration Pass ways of Emission Scenario; in the Case
Wolikite, South West Ethiopia. Int J Environ Sci Nat Res. 2020; 25(3): 556162.DOI: 10.19080/IJESNR.2020.25.556162
Nowadays the sign of climate change and its impact is revealing on different natural and manmade systems, in one or other ways. This study mainly deals to develop future climate change scenario for Wolikite using statistically downscaling of large-scale climate variables. Projection of the future climate variables is done by using Global Circulation Model (GCM) which is considered as the most advanced tool for estimating the future climatic condition. The Statistical Downscaling Model (SDSM) is used to downscale present and future daily precipitation and temperature using observed station data. Three future emission scenarios, RCP2.6 (low emission), RCP4.5 (intermediate emission) and RCP8.5 (high emission) are considered for three 30 years periods for near term (2020-2039), 2020’s mid-term (2040-2059), 2050’s end of century (2080-2099), 2080’s. The average annual minimum temperature will be increased by 3.3°C, 5.3°C and 9.0°C for rcp2.6, rcp4.5 and rcp8.5 scenario respectively towards the end of this century. Similarly, annual maximum temperature will be 7.0°C, 6.0°C, and 8.0°C. Under RCP2.6 the mean temperature increases by approximately 2°C at the end of the century relative to the baseline period. For RCP 4.5, which represents the moderate scenario, the projected increase in temperature is around 2.9°C. The ensemble models are broadly consistent in indicating the shortening the main rain seasons that means monomial rainfall shrinks and ranging from over eight months of rain to only three months (JAS) under all RCPs for all time horizon the station which requires water harvesting, effective and efficient utilization of water resource.
General Circulation Models (GCMs), which are widely used nowadays to simulate future climate scenarios. There are different techniques for down scaling large scale GCM outputs to small scale resolutions to use in other models. All the available techniques and rationale of downscaling are categorized under two broad groups, namely dynamic downscaling and statistical downscaling [1-3]. The dynamic downscaling is performed by Regional Climate Models (RCMs) or Limited Area Models (LAMs) at 0.5o x 0.5o or an even higher resolution that parameterizes the atmospheric processes. The RCMs utilize large scale and lateral boundary conditions from GCMs to produce higher resolution outputs that demands high competition time. The statistical downscaling techniques involve developing quantitative relationships between large scale atmospheric variables (the predictors) and local surface variables (the predictands).
Pragmatic downscaling starts with the premise that the regional climate is the result of interplay of the overall atmospheric or oceanic circulation as well as of regional topography, land-sea
distribution and land use. As such empirical downscaling seeks to drive the local scale information from the larger scale through inference from the cross-scale relationship using some random or deterministic functions. In most cases, the regional climate is seen as random process conditioned up on a driving large scale climate regime. Therefore, the confidence that may be placed in downscaled climate change information is foremost dependent on the validity of the large –scale field from GCM. For instance, derived variables (not fundamental to the GCM physics but derived from the physics) such as precipitation are usually not robust information at the regional and local scale . Conversely, tropospheric quantities like temperature or geo-potential height are intrinsic parameters of the GCM physics and are more skillfully represented by GCM.
The general circulation models (GCMs) used to simulate and project future climate with forcing by greenhouse gases and aerosols, typically dived the atmosphere and ocean in to horizontal
grid with a resolution of 2o latitude by 4o longitude, with 10 to
20 layers in the vertical. In general, most GCMs simulate global
and continental scale processes in detail and provide a reasonable
accurate representation of the average planetary climate. Over the
past decade ,the sophistication of such model has increased and
their ability to simulate present and past global and continental
scale climate has substantially improved .Nevertheless, while GCM
demonstrates significant skill at the continental and hemispherical
scale and incorporate a large portion of the complexity of the
global system, they are inherently unable to represent local subgrid
scale features and dynamics, such as local topographical
feature and convective cloud process . Moreover, GCM were not
designed for climate change impact studies and do not provide a
direct estimation of the hydrological response to climate change.
For example, assessment of future river flow may require (subscenarios)
daily precipitation in the catchment even at station
scales. Therefore, there is a need to convert GCM out put into at
least a reliable daily rainfall series at the scale of the watershed
to which the hydrological impact is going to be investigated. The
method used to covert GCM output into local meteorological
variables required for reliable hydrological modeling are usually
referred to as downscaling techniques. There are two categories
of climate downscaling methods namely dynamic downscaling
and statistical downscaling [5,6].
According to  regardless of human action in the immediate
future, the effects of climate change will persist for centuries. Along
with the increase in mean global temperatures, precipitation,
humidity, and cloudiness are also expected to increase. Globally
averaged surface air temperatures are forecast to increase by
1.4ºC to 5.8ºC by the year 2100 [7,8]. The frequency and intensity
of extreme weather and climatic events will increase in many
regions. As such, projected increases in precipitation may not be
evenly distributed throughout the year. Rather, precipitation may
occur in the form of more frequent intense storm events, which
will result in high runoff levels and increased risks of flooding
. Higher temperatures will increase evapotranspiration.
 Reviewed the current state of climate change science,
reporting that mean global surface temperatures increased by
about 0.7ºC during the 20th century, with 0.4°C to 0.5ºC of this
change occurring since 1970. Historical precipitation trends are
much less clear, because spatial and temporal distributions are
characteristically much more variable than those for temperature.
Available data suggest recent increases of 0.5%-1.0% per decade
in mean annual precipitation on land in the mid- to high latitude
regions of the northern hemisphere, with slight decreases in the
subtropics [9,11-13]. The Atlantic hurricanes of 2004-2005 and a
general trend of increasing storm damage.
Climate data encompass both point-based data (for specific
climate stations) and gridded data (estimated from observations
made at climate stations), as well as historical time series, climate
normal (averages calculated over specified 30-year periods),
and scenarios of past and future climate developed from GCM
The IPCC  proposed that simulations of the impacts of
climate change should be based on a suite of GHG emissions
scenarios, each of which would represent a plausible future
“story” of human population change and economic growth. These
are known as the special report on emissions scenarios . The
various groups working on GCMs around the globe have all been
expected to carry out simulation experiments “forced” by some or
all of these SRES scenarios.
In preparation for the Fifth Assessment Report
(AR5), researchers developed a new approach for
creating and using scenarios in climate change research. This new
approach was motivated by the changing information needs of
policy makers. For example the increasing interest in exploring
different approaches to achieving specific climate change targets
(such as limiting change to 2°C), and growing interest in a “risk
management” approach that combines reductions in emissions
and adaptation to reduce climate change damages.
Scientific advances also dictated the need for new
scenarios. Since the Fourth Assessment Report (AR4) important
improvements in climate models have been made. As the climate
models became more sophisticated, more detailed input was
needed. Simultaneously, models that are used in the production of
scenarios have improved and more advanced input can therefore
be provided [15,16].
a) RCP 8.5 – High emissions: This RCP is consistent
with a future with no policy changes to reduce emissions. It
was developed by the International Institute for Applied System
Analysis in Austria and is characterized by increasing greenhouse
gas emissions that lead to high greenhouse gas concentrations
over time comparable SRES scenario A1 F1. This future is
consistent with: Three times today’s CO2 emissions by 2100,
Rapid increase in methane emissions, Increased use of croplands
and grassland which is driven by an increase in population, A
world population of 12 billion by 2100, Lower rate of technology
development, Heavy reliance on fossil fuels, High energy intensity
and no implementation of climate policies [15,16].
b) RCP 6 – Intermediate emissions: This RCP is developed
by the National Institute for Environmental Studies in Japan.
Radiative forcing is stabilized shortly after year 2100, which is
consistent with the application of a range of technologies and
strategies for reducing greenhouse gas emissions. Comparable
SRES scenario: B2. This future is consistent with:- Heavy reliance
on fossil fuels, Intermediate energy intensity, Increasing use
of croplands and declining use of grasslands, Stable methane
emissions, and CO2 emissions peak in 2060 at 75 per cent above
today’s levels, then decline to 25 per cent above today [15,16].
c) RCP 4.5 – Intermediate emissions: This RCP is
developed by the Pacific Northwest National Laboratory in the
US. Here radiative forcing is stabilized shortly after year 2100,
consistent with a future with relatively ambitious emissions
reductions comparable SRES scenario B1. This future is consistent
with: Lower energy intensity, Strong reforestation programmes,
Decreasing use of croplands and grasslands due to yield increases
and dietary changes, Stringent climate policies, Stable methane
emissions and CO2 emissions increase only slightly before decline
commences around 2040 [15,16].
d) RCP 2.6 – Low emissions: This RCP is developed by PBL
Netherlands Environmental Assessment Agency. Here radiative
forcing reaches 3.1W/m2 before it returns to 2.6W/m2 by 2100.
In order to reach such forcing levels, ambitious greenhouse gas
emissions reductions would be required over time. Comparable
SRES scenario: None. This future would require: Declining use of
oil, Low energy intensity, A world population of 9 billion by year
2100, Use of croplands increase due to bio‐energy production,
More intensive animal husbandry, Methane emissions reduced by
40 per cent, CO2 emissions stay at today’s level until 2020, then
decline and become negative in 2100 and CO2 concentrations
peak around 2050, followed by a modest decline to around 400
ppm by 2100 [15,16].
Figure 1 illustrates the general approach of downscaling;
firstly statistical downscaling is analogous to the “model output
statistics” and “perfect prog” approaches used for short-range
numerical weather prediction .Secondly, Regional Climate Model
(RCM) simulates sub-GCM grid scale climate features dynamically
using time –varying atmospheric conditions supplied by a GCM
bounding a specific domain. Both approaches will continue to
play a significant role in the assessment of potential climate
change impacts arising from future increase in greenhouse gas
concentration. The SDSM is the first tool of its type freely offered
to the broader climate change impacts community
Use of all available GCMs and emission scenario will result
in a better understanding of climate change. However, due to the
limited amount time available to complete the present study, this
research deals with the output from CanESM2 model for RCP
scenarios. Canadian Earth System Model CanESM2 combines
the CanCM4 model and the terrestrial carbon cycle based on the
Canadian Terrestrial Ecosystem Model (CTEM) which models
the land-atmosphere carbon exchange. The concentrations of
greenhouse gases and solar variability are based on the CMIP5
recommendations. In addition, the effects of volcanic eruptions
are included. CanESM2 is applied in this study because the model
is widely applied in many climate change impact studies and it
provides large scale daily predictor variables which can be used
for Statistical Downscaling Model (SDSM) [16,17].
SDSM which is designed to downscale climate information
from coarse resolution of GCMs to local or site level was applied
here to downscale the precipitation, maximum and minimum
temperatures for the study area. SDSM uses linear regression
techniques between predictor and predictand to produce multiple
realizations (ensembles) of synthetic daily weather sequences.
The predictor variables provide daily information about large
scale atmosphere condition, while the predict and described the
condition at the site level. The main reasons to apply the SDSM
model for the study are; it is widely applied in many regions of the
world over a range of different climatic condition, It can be runs on PC-based systems and has been tested on Windows 98/NT/2000/
XP, The availability of the software (i.e. new users can register
and download freely the software package at https://co-public.
lboor.ac.uk/cocwd/SDSM/), Compared to other downscaling
methods, the knowledge of atmospheric chemistry required by
the SDSM is less, The required time for simulating the surface
weather parameter is low and the ability of the model to permit
risk/uncertainty analyses by using the generated ensembles 
The SDSM predictor data files for the CanESM2 model are
downloaded from the Canadian Institute for Climate Studies. The
predictor variables of CanESM2 are provided on a grid box by
grid box basis of size 2.5° latitude x 3.75° longitude. To represent
the station data from the nearest grid box (BOX_014X_35Y) were
downloaded from CICS for wolikite meteorological stations. This
predictor is found in zip file format. When the zip file is opened
the following climatic parameters are found.
a) NCEP_1961-2005: this directory contains 44 years of
daily observed predictors’ data, derived from the NCEP reanalysis,
and normalized (with respect to the mean and standard deviation)
over the complete 1961-1990 period.
b) CanESM2_historical_1961_2005: this directory
contains 44 years of daily GCM predictor’s data, derived from the
CanEsm2 historical data experiment, and normalized over the
c) CanESM2_rcp2.6_2006_2100: this directory contains
94 year of daily GCM predictor data, derived from the RCP2.6
experiment, and normalized over the 1961-1990 period.
d) CanESM2_rcp45_2006_2100: this directory contains
94 year of daily GCM predictor data, derived from the RCP4.5
experiment, and normalized over the 1961-1990 period.
e) CanESM2_rcp8.5_2006_2100: this directory contains
94 year of daily GCM predictor data, derived from the RCP4.5
experiment, and normalized over the 1961-1990 period.
NCEP data which are re-analysis sets from the National
Center for Environmental Prediction was re-gridded to match with the grid system of CanESM2. These data are used for model
calibration. Both NCEP and CanESM2 data have daily predictors.
There exist 26 predictor’s variables in both NCEP and CanESM2
which used for analysis .
At recent decade, the problem of climate variability and
climate change, due to anthropogenic as well as natural processes,
has come with daily bad news . Drought, rain fall delay, fire
damage and heavy and unexpected rain fall are climate related
hazards that mainly faced and also total crop loss, reduced yield,
reduced seeding quality, delayed maturity and increased crop
pest/disease are the major climate impacts, .
Finally, this specific point location study using statistical
downscaling model is aimed to fill the gap of the climate
information of temperature and Precipitation, on the other hand,
remained fairly stable over the last 50 years when averaged over
the country. However, the spatial and temporal variability of
precipitation is high thus large-scale trends do not necessarily
reflect local conditions thus statistical down scaling provide the
future monthly precipitation and temperature climate information
Wolkite the administrative center of the Gurage Zone of
the Southern Nations, Nationalities and Peoples’ Region (SNNPR),
this town has a latitude and longitude of 8°17′N 37°47′E and an
elevation between 1910 and 1935 meters above sea level, is in the
part of the nine regions of Ethiopia.
The observed temperature and precipitation data of the
station was obtained from national Meteorology Agency. The
NMA provide statistical dataset of daily or monthly precipitation
and temperature. These data covering form the period of 1988
to 2018. After collecting the necessary data filling of missed data
and quality checking was be made. GCM have been developed to
simulate the present climate and have been used to predict future
climatic change but GCM are at high resolution and there need to
be downscale the results from such models to individual sites or
localities for impact studies using SDSM. Atmospheric large scale
variables (CanESM2 Predictors) was downloaded from IPCC’s
Fifth Assessment Report (AR5) CMIP5/ Coupled Model Intercomparison
Project, Phase 5 (CMIP5)/ a collaborative climate
modeling process coordinated by the World Climate Research
The second generation of Earth System Model CanESM2 is
the fourth generation coupled global climate model developed by
the Canadian Centre for Climate Modelling and Analysis (CCCma)
of Environment Canada (http://climate-scenarios.canada.
ca/?page=pred-canesm2). SDSM permits the spatial downscaling
of daily predictor-predictand relationships using multiple linear
regression techniques. The predictor variables provide daily
information concerning the large-scale state of the atmosphere,
whilst the predictand describes conditions at the site scale.
The first step before model calibration was quality control
using SDSM through identification of gross data errors, missing
data codes and outliers to get the appropriate quality data. The
screening Predictor variables will be done by trial and error
procedure for model calibration. Using the partial correlations
statistics, predictors which showed the strongest association
with the predictand will be selected. Assembly and calibration
of statistical downscaling model(s) the large-scale predictor
variables identified are used in the determination of multiple
linear regression relationships between these variables and the
local station data. Then SDSM manual procedure will be followed
to generate climate scenario for the basins.
Observed daily precipitation and maximum and minimum
temperatures data will be obtained from weather stations located
in or near the watershed. National Centre for Environmental
Prediction (NCEP) data will be generated for missing data filling
and GCM-derived predictors will be generated form global data
base. Climate data was downscaled using SDSM. The data was
analyzed and tested using trend analysis Man Kandell.
Screening of the potential predictors for the selected
predictand (i.e. observed precipitation, minimum and maximum
temperature) were used to select the appropriate downscaling
predictors for model calibration and the most crucial and decisive
part in statistical downscaling model. Identifying an appropriate
large-scale gridded predictor’s result in good correlation between
observed and downscaled climate variables during model
calibration and scenario generation. The recommended methods
for screening the potential predictors is starting the processes
by selecting seven or eight predictor at a time and analyze
their explained variance, then select those predictor which has
higher explained variance (The significance level which tests the
significance of predictor-predictand correlation was set to the
default P<0.05) and drop the rest.
For the selected predictor analyze or calculate their
correlation matrix with the observed predictand, this statistics
identify the amount of explanatory power of the predictor to
explain the predictand and finally the scatter plot is carried out in
order to identify the nature of the association (linear, non–linear, etc.), whether or not data transformation(s) may be needed, and
the importance of outliers. This procedure is repeated by holding
those predictors which passé the above criteria and add new
predictors from the reset of available predictors.
The model calibration operation takes a selected predictand
along with a set of predictors variables and computes the
parameters of multiple regression equations via an optimization
algorithm (either dual simplex of ordinary least squares). There
are options in SDSM model structure to perform calibration
process either monthly, seasonally, or annual time scale. Selecting
one of these model type decide how the regression parameters
are developed (for example if a model type monthly is selected,
then the model develops one regression equation for the whole
months and if annul model type is selected again one regression
equation is developed for the whole one year and so on). For this
study among the total period length of 1988-2003, 15 years of
daily data was used for model calibration and the rest 15 years
(2004-2018) daily data was used for model validation using a
monthly model type.
The Weather Generator operation generates ensembles (up
to a maximum of 100) of synthetic daily weather series given
observed (or NCEP re–analysis) atmospheric predictor variables.
The procedure enables the verification of calibrated models (using
independent data) and the synthesis of artificial time series for
present climate conditions.
The Scenario Generator operation produces ensembles of
synthetic daily weather series from the starting of the baseline
period to the end of the next century(1961-2100) for a given
daily atmospheric predictor variables supplied by a GCM (either
under present or future greenhouse gas forcing). This function is
identical to that of the Weather Generator operation in all respects
except that it may be necessary to specify a different convention
for model dates and source directory for predictor variables.
The structure and operations of SDSM can be best described
with respect to seven tasks as indicated in bold box in the following
Figure 2 .
The monthly minimum temperature downscaled for NCEP in
the baseline period for Wolkite meteorological station is shown
in Figure 3.
The result of downscaling minimum temperature indicates
that there is less agreement between observed and simulated
minimum temperature compared to nearest meteorological
station minimum temperature this is due to less quality of station
data. As shown in Figure 4: it was also found that, during the
month of May and Jun the model error is negligible. However, for
the rest of the months the model overestimates. The model error
in each month is less than the projected temperature change in
The monthly maximum temperature downscaled for NCEP in
the baseline period is shown in Figure 5.
The result of downscaling maximum temperature indicates
that there is a very good agreement between observed and
simulated maximum temperature. However as shown in Figure
5 the model underestimates maximum temperature during the
month Jan and Nov. however the model overestimates for the rest
of the months except Jan and Nov.
As shown in Figure 7 below the result of downscaling Wolkite
station precipitation indicates that there is an excellent agreement
between observed and simulated precipitation this is due to good
quality of data.
The downscaled precipitation underestimates during the
month of May, Jun, Jul, Aug and Oct and overestimates for the
Results indicate that the models replicate observed intermonthly
and inter-annual variability faithfully, achieving
maximum correlations of the order of 0.98 for temperature (Figure
7) and 0.79 for rainfall (Figure 8) leaving residuals whose variance
is much less than the variance of the raw data. The performance
of the SDSM is almost as very good over the verification period
as it is over the calibration period, indicating that the empirical
model has not been over-fit to the data. Moreover, calibration and
validation of maximum and minimum temperature result show
better correlation coefficient as compared to rainfall.
The Figure 9 shows that a negative anomaly will be observed
in 2020’s and 2080’s time horizon under RCP 2.6 for the months
JJAS and FMA respectively.
As indicated Figure 10 & 11 below most of dry months revealed
that a positive anomaly in the 2050’s time horizon. Under worst
scenario (RCP8.5) only April and July show negative anomaly for
time horizon of 2080’s this indicates that the observed minimum
temperature was cooler than the long- term mean value.
A positive anomaly indicates that the observed temperature
was warmer than the long- term mean under RCP2.6 (Figure 12)
below observed Tmax increase JJAS for all time horizon but for the rest months negative anomaly which indicates the observed Tmax
is cooler than long term mean.
In Figure 13 below under RCP4.5 observed Tmax has positive
anomaly for months JJAS for 2020’s, 2050’s and 2080’s. While, a
negative anomaly indicates that the observed temperature was
cooler than the long- term mean value for months JFMA under
intermediate emission scenarios and High emission scenarios
Figure 14 below.
In the Figure 15 below shows that under RCP 2.6 Positive
anomaly for most of months which means the observed
precipitation is above the long-term normal for those months, and
the negative anomaly were only for few months (JJA) but for June
the negative anomaly will be on 2050’s time horizon.
Under emission scenario RCP 4.5 (Figure 16) below the
negative anomaly was found only for June and July for all threetime
horizons, for months August and September negative
anomaly for 2020-time horizon. This is similar for emission
scenario under RCP8.5 but the difference on percent of above
normal and below normal (Figure 17) below.
After tedious the calibration and validation of SDSM model,
carried out, the daily future climate variables are projected for
the next century using the CanESM2 Global Circulation Model.
The projection generates 20 ensembles of daily climate variables,
which are equally plausible; hence, these ensembles were
averaged out in order to consider the characteristics of all those
20 ensembles. With the aid of statistical downscaling model the GCMs global predictors were used for development of future
climate scenarios and the analysis was made for 2020s, 2050s and
2080s under RCP2.6, RCP4.5 and RCP8.5 (different representative
concentration path way scenarios).
The downscaled minimum temperature shows an increasing
trend in all future time horizons for RCP2.6, RCP4.5 and RCP8.5
scenarios for the stations. The downscaled minimum temperature
in 2020s indicated that the minimum temperature will rise by
0.90C under RCP2.6 scenarios. For the same time horizon under
RCP4.5 scenario the minimum temperature will rise by 1.20C for
wolkite meteorological stations. Under RCP8.5 it will increased by
1.30C. In 2050s the increment will be 4.60C for the station. The
increment will be expected to be high in 2080s, by 3.30C, 5.30C and
90C under RCP2.6, RCP4.5 and RCP8.5 scenarios respectively.
Similar to projected average monthly minimum temperature,
maximum temperature also reflects increasing trend in future
climate periods. The projected maximum temperature in
2020s-time horizon indicated that maximum temperature
would rise by 0.40C, 0.80C and 1.50C under RCP2.6, RCP 4.5 and
RCP 8.5 respectively. In 2050s the increment will be 2.30C under
RCP2.6, 2.90C under RCP4.5 and 3.90C under RCP8.5 scenarios
for Jimma station. The highest maximum temperature rise will
be expected under RCP2.6 by 20C, RCP4.5 by 30C and RCP8.5 by
4.60C for Wolkite station. In 2080s the maximum temperature will
be increased by 70C, 60C and 80C under RCP2.6, RCP 4.5 and RCP
8.5 scenario respectively. This shows that the future period will
experience increasing trend in maximum temperature under all
three representative concentration pathway scenarios. However,
the increments will be less for RCP2.6 scenario relative to RCP8.5
The projected mean annual precipitation in 2020s the stations
were indicated that mean annual precipitation will decrease by
5%, 5.4% and 5% under RCP2.6, RCP4.5 and RCP8.5 respectively.
In 2050’s the minimum decrement will be expected under RCP2.6
by 2.3%, RCP4.5 by 2.8% and RCP8.5 by 3% for Wolkite station. In
2080’s time horizon the projected precipitation will be decreased
by 5.9% under RCP2.6 scenario and under RCP4.5 scenario it will
be decreased by 5.6%. For the worst scenario (RCP8.5) the future
precipitation will be expected to decrease by 5.9% for stations.
Overall, the three scenarios of CanESM2 projected a lessening
trend in the annual precipitation.
The results of the climate projection showed that Statistical
downscaling model is able to replicate the observed and simulated
maximum and minimum temperature well; however, precipitation
couldn’t able to replicate well this is due its conditional nature and
high variability in space and time.
The average annual minimum temperature will be increased
by 3.3°C, 5.3°C and 9.0°C for rcp2.6, rcp4.5 and rcp8.5rcp scenario
respectively towards the end of this century. Similarly, annual
maximum temperature will be 7.0°C, 6.0°C, and 8.0°C. This also
will increase with from lowest to highest emission scenarios.
The ensemble models are broadly consistent in indicating
the shortening the main rain seasons that means monomial
rainfall shrinks and ranging from over eight months of rain to
only three months (JAS) under all RCPs for all time horizon for
Wolikite station which requires water harvesting, effective and
efficient utilization of water resource and mitigation activities.
The results from localities study used for different impact studies
and development plan.