Projected Climate Change over Indus River
Basin using CORDEX-SA models
Ashwini Kulkarni
Indian Institute of Tropical Meteorology, Pune, India
[email protected]
17-18 Feb 2016
The impacts of a changing climate, and the adaptation
strategies required to deal with them, will occur on
more regional and national scales than provided by
Global climate models
This is where Regional Climate Downscaling (RCD) has
an important role to play by providing projections
with much greater detail and more accurate
representation of localised extreme events
Regional Climate Models (RCM) and Empirical
Statistical Downscaling (ESD), applied over a
limited area and driven by GCMs can provide
information on much smaller scales
supporting more detailed impact and
adaptation assessment and planning, which is
vital in many vulnerable regions of the world.
PRECIS Runs at IITM
(56.77o – 103o E ; 1.5o – 38o N)
PRECIS runs with HadCM3 LBCs
Baseline : 1961-1990
A2, B2 scenarios : 2071-2100
Resolution : 50 km x 50 km
PRECIS runs with ECHAM4 LBCs :
1960-2100
Resolution : 50 km x 50 km
PRECIS runs with LBCs from 3 members
of QUMP simulations (A1B scenario) :
1961 – 2098
Resolution : 50 km x 50 km
Quantifying Uncertainties in Model Predictions (QUMP)
• 17 versions of the fully coupled model HadCM3, one with the standard parameter
setting and 16 versions in which 29 of the atmosphere component parameters are
simultaneously perturbed : Perturbed Physics Ensembles(PPE)
•LBCs taken from the 17 member Perturbed Physics Ensembles based on Hadley Centre
Coupled Model are used in QUMP project
•The LBCs for three QUMP simulations viz. Q0, Q1 and Q14 for scenario A1B were made
available by Hadley Centre, UK.
•These three QUMP runs were carried out at IITM, Pune for the period 1961-2098 over
the domain 56.77o – 103o E ; 1.5o – 38o N and are utilized to generate an ensemble of
future climate change scenarios.
• The baseline period is taken as 1970s(1961-1990) and the impacts are studied for
three time slices viz. 2020s (2011-2040), 2050s (2041-2070) and 2080s (2071-2098).
IITM, Pune
Coordinated Regional Downscaling Experiment (CORDEX)
CORDEX is a project of the WCRP Group on Regional Climate (WGRC)
CORDEX is providing global coordination of Regional Climate Downscaling for
improved regional climate change adaptation and impact assessment.
CORDEX GOALS
• To better understand relevant regional/local climate phenomena, their
variability and changes, through downscaling
• To evaluate and improve regional climate downscaling models and
techniques
• To produce coordinated sets of regional downscaled projections
worldwide
• To foster communication and knowledge exchange with users of regional
climate information
WCRP CORDEX South Asia
Co-ordinated Regional Downscaling Experiment – CORDEX South Asia
(CORDEX South Asia – led by CCCR, IITM)
South Asia
Participating Modeling Groups
•
LMDZ model (~ 35 km ) CCCR (IITM), IPSL
•
RegCM model (~ 50 km)
•
WRF model (~ 50 km) - CCCR (IITM), BCCR and TERI
•
MRI NHRCM model (~ 50 km) MRI, Japan
•
RCA model (~ 50 km) Rossby Centre, Sweden
•
REMO model (~ 50 km) Max Planck Inst, Hamburg
•
CCAM model ( ~ 50 km) CSIRO, Australia
CCCR (IITM)
Regional Climate Information for Application Studies
Update on CORDEX South Asia
CCCR is leading CORDEX
(Coordinated Regional Climate
Downscaling Experiment)
over South Asia Region
High Resolution (50 km)
Dynamical Downscaling of
CMIP5 Climate Projections
based on RCP Scenarios
during 1950-2100 using
multiple RCMs
R. Krishnan and Team members
Historical (1886-2005): Includes natural and anthropogenic (GHG, aerosols, land
cover etc) climate forcing during the historical period (1886 – 2005) ~ 120 years
Historical Natural (1886 – 2005): Includes only natural climate forcing during the
historical period (1886– 2005) ~ 120 years
RCP 4.5 scenario (2006-2100) ~ 95 years:
Future projection run which includes both
natural and anthropogenic forcing based on
the IPCC AR5 RCP 4.5 climate scenario . The
evolution of GHG and anthropogenic aerosols
in RCP 4.5 scenario produces a global
radiative forcing of + 4.5 W m-2 by 2100
CORDEX South Asia data (50km) is available on the
CCCR-IITM Climate Data Portal (non-ESGF):
http://cccr.tropmet.res.in/cordex/files/downloads.jsp
•
Thanks to: S. Ingle M. Mujumdar
http://cccr.tropmet.res.in/globaldata/
CORDEX South Asia (WAS-44) Datasets Published on
Earth System Grid Federation (ESGF)
Downscaling
RCMs
Driving AOGCMs
Updated: 04/Jan/2016
Published:
• Control(1951-2005):7 RCP2.6(2006-2100):3 RCP4.5(2006-2100):7 RCP8.5(2006-2100):7
Finished:
• Control(1951-2005):20 RCP2.6(2006-2100):3 RCP4.5(2006-2100):19 RCP8.5(2006-2100):17
LMDz (Laboratoire de
Me´te´orologie Dynamique)
model
Horizontal resolution :
high-resolution (~ 35 km)
zooming over South Asia.
Vertical resolution : 19
levels
Historical run (1886-2005) : Uses both natural (eg., Volcanoes, Solar
forcing) and anthropogenic (eg., GHG, aerosols evolution estimated from
transport models, land use and land cover changes, etc) forcing. The SST
and Sea Ice for this period are based on the corresponding CMIP5 coupled
run.
Historical Natural (1886-2005) : Uses only natural forcing. The SST and
Sea Ice for this period are based on the corresponding CMIP5 coupled run.
Future Climate Change RCP 4.5 (2006-2100): Uses both natural and
anthropogenic forcing based on IPCC AR5 RCP4.5 (Representative
Concentration Pathway) climate scenario in which the Net Radiative Forcing
at the end of 2100 is 4.5 Wm-2
The upper Indus basin consists of mountainous terrain, and includes
parts of the Hindu Kush, Karakorum, and Himalayan mountain ranges;
the lower Indus basin comprisesthe southern plains. The separation is
mainly based on altitude.
Indus river is one of the most important source of water to South
Asian region
The water resources in this basin are threatened by
Burgeoning population
economic development
tremendous increase in financially well-off people who tend
to use higher qualities of water
urbanisation
large portion of water used by agriculture sector
pollution
decay of water infrastructure
mismanagement of water resources
The Indus River is extremely sensitive to climate change due to the high portion of its
flow derived from glaciers. Temperature controls the rate of glacier melt, which in
turn, provides more water in dry, warm years and less water in cool years.
River catchments with a large portion of glacial melt water experience less
variability in water flows. With climate warming, many glaciers will no longer exist
to moderate the flow of these rivers. Thus communities which depend on glacier
water will face more severe water shortages, variability and potentially greater
flooding too.
The Himalayan glaciers provide the Indus with 70-80% of its water, the highest
proportion of any river in Asia.
Although the Indus system is currently robust enough to cope with shortages of
10-13% in river flows, when the rivers flow drops to 15-20% below the average,
irrigation shortages occur. Climate change will surely exacerbate the problems of
irregular and low flow.
Ganges and Indus are from the first top 10 rivers at risk due to climate change,
water extraction, agricultural pollution, water infrastructure, 6 proposed large
dams(WWF)
Simulation of seasonal (JJAS) rainfall compared with APHRODITE
APPHRODITE
MODEL
BIAS
COSMO_CLIM_4.8
Pattern very well simulated
Wet bias over NW region, slight dry bias
Over remaining parts
GFDL-ESM2_IITM-RegCM4
Wet bias over southern parts
Good simulation
Simulation of seasonal (JJAS) rainfall compared with APPHRODITE
Wet bias over southern parts
Zoomed model : better simulation
Negligible bias
Ensemble simulation JJAS Rainfall
Simulation JJAS Temperature
Cold bias : colder towards northern
Hilly regions
Warm bias over the basin
Simulation JJAS Temperature
Pattern well captured, warm bias in
southern planes
Southern parts
Well simulated
ow cold bias
Projected change in JJAS Rainfall
Largest number of Consecutive dry days/year
Projected to increase over southern parts
Number of dry periods with more than 5 dry days : simulation(1986-2005),
Change(2041-2060), (2081-2100)
COSMO_CLM_4.8
GFDL
Projected to decrease especially
Over southern parts
LMDZ
One-day Extreme Rainfall
Model overestimates
Projects decrease in southern
Parts
Increase in northern
One-day Extreme Rainfall
Overestimates in southern parts
Projects to decrease in southern
Increase in northern parts
One-day Extreme Rainfall
Captures reasonably well
Projects to increase in
northern parts
Projected change in Max and Min temp
GFDL
Projest general
Increase of 1-2o
In max and 2-3o
In min temp
Projected change in Max and Min temp
ICHEC
1-2o warming in both
Projected change in Max and Min temp
LMDZ
Min temp projected to warm
more than max temp
SUMMARY
CORDEX-SA makes available the multi-model simulations which may help to examine and reduce
the uncertainties in model simulations. The analysis shows that
•
Zoomed model (LMDZ) better simulates the climate over basin
•
The climate over Indus basin, which has been changing rapidly, will continue to do so in the
future, with severe consequences for populations locally and downstream.
•
•
Temperatures across the basin will increase by about 1–2°C (in northern parts at some
places more than 3°C) by 2050s which may initiate more glacier melt.
•
Precipitation may increase by 1-2 mm/day in 2050s, with more increase in northern parts
•
Maximum temperature may rise by 1-2o C while minimum by 1-3o C
•
Largest number of consecutive dry days may increase in future
•
Models simulate substantial wet bias in one day extreme rainfall which may increase in
future in southern parts
•
Changes in temperature and precipitation will have serious and far-reaching consequences
for climate-dependent sectors, such as agriculture, water resources and health.
A WORD OF CAUTION …..
While the scenarios presented are indicative of the range of
rainfall and temperature changes that could occur in the
future,
•
it must be noted that the quantitative estimates still have
large uncertainties associated with them, due to the
differences between global climate model results and the
limited number of high resolution regional model simulations
available
• For quantification of uncertainties and for achieving more
confidence in the projections more simulations from high
resolution regional models are required.
THANK YOU