Transcript The Impact of Climate Change on a Humid, Equatorial

The Impact of Climate Change
on a Humid, Equatorial
Catchment in Uganda.
Lucinda Mileham, Dr Richard Taylor, Dr Martin Todd
Department of Geography – University College London
Changing Climate
•
Africa has experienced a mean continental 20th Century warming of 0.7 °C
•
The periods of greatest warming are 1910-1930 and post 1970, with the five warmest
years on record occurring since 1988.
•
decrease in runoff of 17 % in major river basins
•
Predicted temperature increase of between 2-6 °C by 2100
•
Accompanied by evaporative increases of 19-27 % by 2080
•
Changes in the seasonality and intensity of future precipitation
Why is it important in Uganda
• Dependence on rainfall-fed agriculture
• Reliance on localised (untreated) sources of water
• Groundwater is the only reliable source of potable water
• Current population of 25.8 million is estimated to more than double by 2025
Main Objective
To quantify the impacts of climate change on groundwater
recharge and surface runoff in Uganda, East Africa.
Specific objectives
• To evaluate the ability of a RCM to reproduce the current (19601990) climate at scales appropriate for hydrological modelling.
• To develop a soil-moisture balance model for groundwater
modelling operating at a scale which allows coupling with a
RCM.
• To quantify the impacts of climate change (2070 to 2100) on
catchment-scale terrestrial water resources in Uganda.
PRECIS
What is PRECIS? (Providing Regional Climates for Impacts Studies)
A simple-to-use PC-based RCM, has been developed by the Hadley Centre (UK)
specifically to address the need for countries to make regional-scale climate
predictions.
• Model resolution 25 or 50 km²
• Daily time-step
• Boundary conditions - A three member ensemble of central experiments (1960-1990)
- ECMWF ERA 40 reanalysis experiment 1957-2001
- A three member ensemble of SRES A2 scenario experiments (2070-2100)
- A single SRES B2 scenario experiment 2070-2100
• User defined emission scenarios, and specification of output levels
• Limited user manipulation – land use, inland water features and topography can be
changed but model parameterizations cannot be altered.
Study Area / Method
Catchment Scale
River Mitano catchment.
•
2, 098 km 2
•
High relief and incised drainage reflect a runoff-dominated regime
•
Land use - Agriculture (79 %), Grassland (17 %) wetland (3 %), small areas of
forest and plantation
Regional scale - PRECIS Mean Climatology
.
7.00
6.00
Precipitation (mm/day)
5.00
4.00
3.00
2.00
1.00
PRECIS-ERA
PRCIS-GCM
VasCLIMO
UDEL
CRU
GPCC V3 0.5 PRECIP 1960-1990
GPCP PRECIP 1979-now
CMAP (1979-now)
TRMM
0.00
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
• Bi-modal precipitation regime
• Good representation of the temporal distribution of precipitation
• Overestimation of precipitation between December-March
Oct
Nov
Dec
PRECIS
GPCC
CRU
GPCP
VasCLIMO
TRMM
Catchment-scale validation of precipitation
7
PRECIS
CRU
Station
VasCLIMO
GPCC
GCPC
CMAP
TRMM
Preciptiation (mm.day)
.
6
5
4
3
2
1
0
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
•
<10 % error in mean annual precipitation
•
Better representation of Sep-Nov rainy season
•
Magnitude remains poorly resolved in Jan-Mar
•
Poor representation of peak Precipitation in first rains
Sept
Oct
Nov
Dec
Future PRECIS Change (2070-2100) cont …
•
•
•
•
•
Increased precipitation July-December
Shift in timing of peak seasons
Mean annual precipitation increase of 14 %
3.5 °C increase in Temperature
53 % increase in PET
Soil moisture balance model (SMBM)
• Simulates changes in soil moisture (‘green water’)
and provides estimates of rainfall-fed, groundwater
recharge and runoff (infiltration excess)
• Can be run as a lumped parameter model OR as a
semi-distributed model by running for different soil,
slope and vegetation characteristics.
• Critically, can be run using gridded RCM or
downscaled GCM data.
SMBM calibration (1965 – 1979)
• Validated against discharge data
from the River Mitano gauging station
• Poor performance is due to lag
responses
• Stormflow is significantly better
represented than baseflow due to its
shorter lag response
• Should only be used to represent
baseflow and stormflow on annual and
longer timescales
• SMBM reproduces well the mean
annual recharge and runoff
Mean monthly delta factors
• Difference between 19601990 and 2070-2100
modelled precipitation.
• All seasons exhibit small
increases in precipitation
across Uganda
DJF
MAM
• Delta factors 0.9 to 1.8
• Mean monthly factor
• Applied to daily data
JJA
SON
Effect on Hydrology
GRIDDED STAT RECHARGE
.
• 14 % increase in Recharge
DELTA RECHARGE
3.0
2.5
Recharge (mm.d-1)
• 84 % increase in Runoff.
2.0
1.5
1.0
0.5
0.0
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
4.50
OCT
NOV
DEC
DELTA RUNOFF
GRIDDED STAT RUNOFF
.
4.00
3.50
Runoff (mm.d-1)
3.00
2.50
2.00
1.50
1.00
0.50
0.00
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
DEC
Changes in Daily Precipitation Distribution
Monthly Delta factors fail to account for changes in the daily precipitation distribution
• Reduction in small precipitation events (<10 mm)
• Increase in large precipitation events (>10 mm)
• Variable results for extreme precipitation events
•The distribution of precipitation is key for modelling of recharge and runoff
Impacts on hydrology
.
3.50
3.00
Recharge (mm.d-1)
The transformation method –
matches future precipitation
and historical precipitation
distributions.
TRANSFORMED RECHARGE
GRIDDED STAT RECHARGE
4.00
2.50
2.00
1.50
1.00
• 66 % increase in Recharge
• 123 % increase in Runoff
0.50
0.00
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
DEC
TRANSFORMED RUNOFF
GRIDDED STAT RUNOFF
4.50
.
4.00
3.50
3.00
Runoff (mm.d-1)
Fails to account for changes
in occurrence
NOV
2.50
2.00
1.50
1.00
0.50
0.00
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
DEC
Conclusion
•
PRECIS does a reasonable job at representing the climate of east Africa
•
Observational uncertainty is large
•
SMBM represents well the mean annual catchment recharge and runoff
•
Recharge and runoff are sensitive to the distribution of precipitation
•
Increase in the magnitude and intensity of precipitation
•
Large increases in evapotranspiration and temperature
•
Increases in surface runoff and groundwater recharge under future climates