Transcript Dia 1
Impact of climate change on
hydrological extremes in Belgium
prof. dr. ir. Patrick Willems
K.U.Leuven – Hydraulics Division
Hydrological climate change impact
research for Belgium
•
Waterbouwkundig Laboratorium
impacts on high and low flows
BelSPO: CCI-HYDR
climate change scenarios
for hydrological impact analysis
incl. extremes
•
VMM:
−
−
•
impacts on floods (non-navigable rivers 1st
category)
update urban drainage design guidelines
INBO: impact on nature
incl. comparison with KNMI’06 scenarios
•
MIRA-S & NARA 2009
•
BelSPO: SUDEM-CLI: interfacing
climatology –hydrology – ecology
•
EU-FP7: Theseus: correlation with storm
surge and wave scenarios Belgian Coast
Climate change scenarios
for Belgium
Based on simulation runs by global and regional climate
models & IPCC SRES greenhouse gas emission scenarios
till 2100
GCM
RCM
31 runs (A2,B2) & 26 runs (A1B) by 10 RCMs
17 runs by ECHAM5 (A1B)
IPCC AR4: 27 runs by 21 GCMs (A2, A1B, B1)
DMI 25 km
Greenhouse gas emission
scenarios
IPCC SRES, 2001 & 2007:
Climate change scenarios
for Belgium
• Validation regional climate model runs for control period
(monthly temperature, 1961-1990):
Climate change scenarios
for Belgium
• Increase in monthly temperature:
summer:
+2 to +7°C
winter:
+1.5 to +4°C
Climate change scenarios
for Belgium
• Change in monthly precipitation:
GCMs 1961-1990 :
RCMs 1961-1990 :
GCMs 2071-2100 :
RCMs 2071-2100 :
winters natter
zomers droger
Climate change scenarios
for Belgium
• Change in monthly precipitation:
: increase in winter
: decrease in summer
(lower no. of small rain storms)
winter:
up to +60%
summer:
a.l.a. -70%
no. wet days
summer:
a.l.a. -50%
Climate change scenarios
for Belgium
• Change in precipitation
– Intensity increase for most extreme storms
most extreme storm
in 2 years
Climate change scenarios
for Belgium
• Change in precipitation
– Intensity increase for most extreme storms
most extreme storm
in 10 years
Historical climate trends for
Belgium
• Historical trend analysis
– DJF rainfall extremes Uccle (10 min -> seasonal) 1898 –2005:
anomaly in extremes [%]
30
Global warming
impact
20
10
Multidecadal
climate oscillation
0
-10
-20
winter, 10-year window
winter, 15-year window
long-term average
approximate cyclic variations
cyclic variations plus climate change
climate change effect
-30
-40
-50
1900
1910
1920
1930
1940
1950
1960
year [-]
1970
1980
1990
2000
2010
Climate change scenarios
for Belgium
• Consistency check with historical trend analysis
– Example: Winter (DJF), daily rainfall extremes:
1.35
Regional climate model simulations
SHMI-MPI-A2
1.3
High = Wet
SHMI-MPI-B2
Perturbation factor
1.25
CNRM-DE6
DMI-ECC-A2
1.2
DMI-ECC-B2
CNRM-DE5
1.15
ICTP-A2
HS2 / HS3 / CNRM-DC9
Historical trend 30 years blocksize
CNRM-DE7 / SHMI-HC22
GKSS-A2
GKSS-sn-A2 / METNO-A2
SHMI-HC-A2
ICTP-B2
DMI25 / KNMI
METNO-B2
1.1
Historical trend 30 years blocksize:
part c.c. increase
1.05
1
Mean = Mild
ETH / HS1
SHMI-HC-B2
Current
Control period
(1960-1990)
0.95
1960
1970
1980
Low = Dry
Scenario period
(2070-2100)
1990
2000
2010
2020
2030
2040
2050
2060
2070
2080
2090
2100
Climate change scenarios
for Belgium
• Regional differences
– Rainfall change about 10% higher along the Belgian coast
More wet in winter:
Low = Dry
Mean = Mild
High = Wet
Climate change scenarios
for Belgium
• Regional differences
– Rainfall change about 10% higher along the Belgian coast
Less dry in summer:
Low = Dry
Mean = Mild
High = Wet
Statistical downscaling
Large Scale
150 – 300 km; seasonally – monthly
General
Circulation Models
(GCMs)
Dynamical
downscaling
± 50 km; weekly - daily
± 25 km; daily
river catchment; hourly
Hydrological scale
Regional
Climate Models
(RCMs)
Statistical
downscaling
Perturbation tool
• Day -> hour (river), 10-min (sewer system)
• Based on quantile perturbations:
– change in rain storm frequency and rain storm intensity
– dependent on return period rainfall intensity, season, weather type, …
• Time horizons till 2030, 2050, …, 2100
Wet day frequency
perturbation
Wet day intensity
perturbation
Combined perturbation
Daily
Hourly
10min
High = Wet
Time
series
Month i
Month i
Month i
Time
series
Mean = Mild
Low = Dry
Perturbation tool
• Preserves physical consistency (dependency) between
seasons and variables (precipitation, temperature and
ETo)
Winter
Day-Winter
Day-Summer
Summer
1.4
1.2
1.2
factor
Precip. change
Rainfall Perturbation
[-]
factor
Precip. change
Rainfall Perturbation
[-]
1.4
1
High
High
Mean
Mean
Low
Low
0.8
0.6
0.4
1
0.8
0.6
0.4
0.8
1
1.2
1.4
Eto Perturbation [-]
ETo change factor
1.6
1.8
0.8
1
1.2
1.4
Eto Perturbation [-]
ETo change
factor
1.6
1.8
Perturbation of rainfall series
• Change in rainfall IDF-relations:
– Based on Uccle 10 min rainfall series 1898-2005:
Return period:
100 years
10 years
1 year
1 month
Perturbation of rainfall series
• Change in rainfall IDF-relations:
– Daily rainfall results, climate model runs:
Return period:
100 years
10 years
1 year
1 month
Perturbation of rainfall series
• Change in rainfall IDF-relations:
– 10 min downscaling results, climate model runs:
Return period:
10 years
1 year
1 month
Climate change impact on urban
drainage
• Change in rainfall IDF-relations:
– High, mean and low climate scenarios:
100 years
50 years
10 years
2 years
1 year
2 months
1 month
Perturbation of rainfall series
• Change in rainfall IDF-relations:
– Shift in return period high climate scenario:
Perturbation of “design” storms
• Change in rainfall IDF-relations:
– Change in composite storms, example T = 2 years:
High
Mean
Low
Hydro-impact modelling
Rainfall, ETo
Rainfall-runoff
NAM, PDM: conceptual
Spatially distributed:
SCHEME (KMI/IRM),
MIKE-SHE
WetSpa (VUB)
Bridge over tributary
(culvert + weir)
River
hydrodynamics
Physicochemical river
water quality
MIKE11
InfoWorks-RS
+ quasi 2D overstromingen
TRIBUTARY
Left floodplain
Right floodplain
MAIN RIVER
MIKE11 EcoLab
Spills
Calculation nodes
numerical scheme
Hydrological impacts
• Impact of climate scenarios on hourly runoff peaks:
80
variatie piekafavoeren (%)) )
High
High
Mean
Mean
Low
Low
60
rainfallincrease
40
20
0
-20
-40
0.1
1
10
Terugkeerperiode (jaar)
100
EToincrease
Hydrological impacts
RUNOFF PEAKS
Low scenario, Runoff peaks
(-70%)
(-49%)
(-29%)
(-21%)
(-12%)
- (-50%)
- (-30%)
- (-22%)
- (-13%)
-0
• Impact of climate scenarios on hourly runoff peaks:
Low scenario, Runoff peaks
Mean scenario, Runoff peaks
RUNOFF PEAKS
(-70%)
- (-50%)
(-14%)
- (-13%)
(-49%)
- (-30%)
(-12%)
(-29%)
- (-22%)
(-11%)
- (-9%)
(-21%) - (-13%)
(-8%)
(-3%)
(-12%) - 0
(-2%) - 3%
High scenario, Runoff peaks
Mean scenario, Runoff peaks
0
(-14%)
(-13%)
1% - -22%
(-12%)
23 %- 24%
(-11%)
(-9%)
25 %-- 32%
(-8%)
(-3%)
33% - 37%
(-2%) - 3%
Climate 2100, Flanders
High scenario, Runoff peaks
Change in flood risks is highly uncertain
0 -43 - -34
-33 - -19
1% - 22%
Runoff peaks due to rainfall/ETo change decrease in low
scenario and increase in high scenario (up to 35%)
2100, Flanders
Major influence due to sea level rise (Scheldt tidal Climate
river)
Demer.shp
23 %- 24%
-74 - -71
25 %32%
-70 - -56
33%-55
- 37%
- -44
CLIMAR
Climate change scenarios for
Belgium
• Sea level rise:
+20cm to
+2m
about 20cm past 100 years
Impacts on floods
• Flood map current climate:
T = 100 year
Current climate
Impacts on floods
• Flood map after climate scenarios:
T = 100 year
High scenario
High = Wet
Hydrological impacts
Low scenario
Low scenario, Runoff peaks
LOW FLOW PEAKS
(-88%)
(-87%)
(-67%)
(-62%)
(-54%)
- (-68%)
• Impact of climate scenarios on low flows
extremes:
- (-63%)
Low scenario
Mean
scenario
Mean
scenario,
Runoff peaks
Low scenario, Runoff peaks
- (-55%)
- (-48%)
LOW FLOW PEAKS
(-88%)
(-56%) - (-55%)
(-87%)
(-68%)
(-54%) -- (-52%)
(-51%) -- (-47%)
(-67%)
(-63%)
(-46%) -- (-40%)
(-62%)
(-55%)
(-39%) -- (-30%)
(-54%)
(-48%)
Meanscenario
scenario
High
Meanscenario,
scenario,Runoff
Runoffpeaks
peaks
High
(-35%)- -(-55%)
(-32%)
(-56%)
(-31%)
(-24%)
(-54%) - (-52%)
(-23%)- -(-47%)
(-21%)
(-51%)
(-20%) - (-15%)
(-46%) - (-40%)
(-14%) - (-10%)
(-39%) - (-30%)
Climate 2100, Flanders
High scenario
High scenario, Runoff peaks
-43 - -34
Low flow risks increase significantly in
-33 - all
-19 scenarios
(-35%) - (-32%)
(-31%) - (-24%)
Demer.shp
(-23%)
- (-21%)
- -71
(-20%) -74
- (-15%)
-70
- -56
(-14%) - (-10%)
May increase problems rel. water quality, navigation, drinking
water production, irrigation, ecological state
river valley, ...
-55 - -44
Climate 2100, Flanders
Hydrological impacts
• Drier summer climate can have severe impacts
• Mean water availability in Flanders and Brussels is very
limited: 1480 m3/(person.year)
– International standards: <2000 “zeer weinig”, <1000 “ernstig
watertekort”
– Causes:
• High population density:
– high urbanisation, pavements: increased surface runoff, decreased
infiltration
– high drainage in agriculture
– groundwater abstractions for drinking water supply (region of Waregem:
groundwater table levels >100m lower than natural conditions)
• (in Scheldt basin): less than half of available water is due to local rainfall
• strong dependency on neighbouring regions (The Netherlands: Meuse and canal
Gent-Terneuzen)
Climate change impact on urban
drainage
• Change in overflow frequencies storage + infiltration
facilities:
– Reservoir model:
Climate change impact on urban
drainage
• Change in overflow frequencies storage + infiltration
facilities:
– Change in storage capacity needed:
constant
throughflow
[l/(s.ha)]:
50
40
30
25
20
15
10
5
2
1
Return period overflow [years]:
0.5
+35%
+25%
+18%
+18%
+16%
+17%
+17%
+13%
+13%
+9%
1
+18%
+16%
+13%
+14%
+18%
+20%
+20%
+17%
+14%
+10%
2
+17%
+17%
+12%
+12%
+13%
+11%
+19%
+17%
+14%
+10%
5
+29%
+25%
+30%
+31%
+27%
+22%
+20%
+18%
+15%
+10%
10
+24%
+31%
+29%
+25%
+26%
+22%
+24%
+24%
+30%
20
+33%
+27%
+25%
+22%
+21%
+13%
+17%
+20%
+18%
Climate change impact on hydrology
Climate scenarios for Belgium till 2100:
Impact on water systems:
– Winter season:
• Rainfall increase
• ETo increase
• Sea level rise
Unclear impact on inland river floods
Increase in coastal flood risks
– Summer season:
• Rainfall decrease
Increase in low flow / water scarcity problems
• ETo increase
• More intense convective summer storms
Increase in sewer floods
Climate change impact on hydrology
For specific conditions of river basins in Flanders/Belgium:
• Climate scenarios up to 2100:
– Evolution towards more droughts
– Impact on flood risk along inland rivers unclear (<-> coast)
– Increase in short-duration rainfall extremes
• But: climate scenarios not equal to predictions !
– Evolve along with our knowledge and understanding
– High uncertainties; part of the uncertainties not accounted for
• Adaptation measures ?
– Use climate scenarios in water management planning !
– Take high uncertainties into account !
“concept of risk = probability * consequences”
“no regret” measures, “climate proof” investments
design a flexible strategy (adaptable measures/investments)
Actions or adaptation measures
• See river basin and subbasin management plans
• New Sigmaplan
– Geactualiseerd Sigmaplan: gecontroleerde overstromingsgebieden
en natuurgebieden (1650 ha in 2030) en dijkverhogingen in steden
en industriegebieden
• Prediction of and warning for flood risks:
• Real-time prediction and warning system
• Real-time regulation of reservoirs: more efficient use of available storage capacity
by model-predictive control algorithms
Actions or adaptation measures
• Against drier summer climate:
– Maximum upstream holding of rain water in infiltration facilities,
ditches, low-lying spaces, ...
– Revision of regulations (vergunningen- en heffingenbeleid) and
water-pricing system (sturend waterprijzenbeleid: variabele tarieven
voor drinkwater, progressieve tarieven naargelang verbruik)
– Reduction of water consumption, water loss, reuse of water
(sensibilisering rationeel watergebruik)
– “Water-audit” for new or renovated buildings
– Innovation programmes for water efficient industrial production and
irrigation
Actions or adaptation measures
• Against increased temporal variability of rainfall: drier
summers + increased short-duration rainfall extremes:
– calls for more attention to (local) rain water management (also at
municipal level)
– additional (upstream and local) rain water storage and infiltration
needs
• More small scale solutions (at large scale)
• More important role to local water managers
Actions or adaptation measures
Accept increased frequencies of “water in the streets”
local scale measures can significantly reduce the damage (for
same flood frequency)
Actions or adaptation measures
More local upstream storage (local terrain depressions)
combined effect: reduces sewer flood frequencies & rain water
feeds groundwater table
Actions or adaptation measures
Better integration of water management and spatial planning /
urban design
Multiple functions to
open spaces (e.g.
parks) in the city:
FWO research project
together with
K.U.Leuven - ASRO
(PhD researchers
Isabelle Putseys
& Christian Nolf)
Actions or adaptation measures
Better integration of water management and ecological / nature
management
Role of wetlands:
SUDEM-CLI cluster
project for BelSPO
(cooperation with
U.Antwerp – ECOBE)
Actions or adaptation measures
Better integration of water management and agriculture
Avoid winter runoff (and erosion) and increase infiltration by
“adapted” agricultural management practices:
• groenbemesting
• ploegrichting
• bufferstroken naast rivier
Additional needs
• Improved cooperation between water managers and spatial
planners / urban designers, managemers of green zones in
the city (parks, playing gardens)
• Improved interlinking between:
– Guidelines for design of urban drainage systems (code van goede
praktijk voor het ontwerp van rioleringssystemen)
– “Watertoets”
– Building regulations (gewestelijke stedenbouwkundige verordening;
andere stedenbouwkundige voorschriften)
More info
Research project CCI-HYDR on
“Impact of climate change on hydrological extremes (peak and low
flows) along rivers (Scheldt and Meuse basins) and urban drainage
systems in Belgium”
(for Belgian Science Policy Office):
http://www.kuleuven.be/hydr/CCI-HYDR
Impact studies:
Instituut voor Natuur- en Bosonderzoek (INBO):
http://http://www.inbo.be
Vlaamse Overheid:
Waterbouwkundig Laboratorium:
http://www.watlab.be
Vlaamse Milieumaatschappij:
http://www.milieurapport.be
http://www.watertoets.be/publicaties