Comparison of statistical downscaling procedures for climate change - - PowerPoint PPT Presentation

▶
comparison of statistical downscaling procedures for
SMART_READER_LITE
LIVE PREVIEW

Comparison of statistical downscaling procedures for climate change - - PowerPoint PPT Presentation

Sep 26, 2023 418 likes •652 views

Comparison of statistical downscaling procedures for climate change impact assessment of water resources Henrik Madsen, Maria Sunyer, Keiko Yamagata DHI, Denmark HydroPredict 2010, 20-23 September 2010, Prague, Czech Republic Downscaling

slide-1
SLIDE 1

Comparison of statistical downscaling procedures for climate change impact assessment of water resources

Henrik Madsen, Maria Sunyer, Keiko Yamagata DHI, Denmark

HydroPredict 2010, 20-23 September 2010, Prague, Czech Republic

slide-2
SLIDE 2

Downscaling

Global climate model projections Downscaling

  • Dynamical
  • Statistical

Local-scale impact assessment

slide-3
SLIDE 3

Regional climate model (RCM)

  • Driven by GCM boundary

conditions

  • Higher resolution (10-50 km)
  • Resolves sub-GCM grid scale

forcings in a physically-based way

  • Further statistical

downscaling needed Dynamical downscaling

1 2 3 4 5 6 7 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Mean [mm/day] Observed HIRHAM-ECHAM5

slide-4
SLIDE 4

Statistical downscaling

Statistics

Future CF

Statistics

RCM Control

Statistics

Observed

  • Define relationship between large-scale model (GCM or RCM)

and local climate

  • Methods based on Change Factor Methodology:
  • Mean correction
  • Mean and variance correction
  • Weather Generators

Statistics

RCM Future

slide-5
SLIDE 5

Statistical downscaling methods

  • Mean correction

(delta change)

  • Mean and variance

correction

  • Weather generators
  • Neyman-Scott

Rectangular Pulses

  • Markov Chain
  • LARS WG
  • Mean
  • Variance
  • Proportion of dry days
  • …

Future Statistics

FutsyncTS

WG

Model fitting:

  • Mean
  • Variance
  • Skewness
  • Dry-day prob.
  • Autocorrelation
  • Transition prob.

Change factors

Model fitting:

  • Mean

Model fitting:

  • Mean
  • Variance
slide-6
SLIDE 6

Case study – North Sealand

Asses climate change impacts

  • n hydrology

North Sealand (3000 km2)

Global Climate Model

[meter] Above 100 90 - 100 80 - 90 70 - 80 60 - 70 50 - 60 40 - 50 30 - 40 20 - 30 10 - 20 0 - 10
  • 10 -
  • 20 - -10
  • 30 - -20
  • 40 - -30
Below -40 Undefined Value 620000 640000 660000 680000 700000 720000 [meter] 6100000 6110000 6120000 6130000 6140000 6150000 6160000 6170000 6180000 6190000 6200000 6210000 6220000 [meter]

Regional Climate Model

Dynamical downscaling Statistical downscaling

slide-7
SLIDE 7

ENSEMBLES DATA

ARPEGE ECHAM5 HIRHAM ALADIN HIRHAM REMO Driving GCM: RCM:

OBSERVED DATA SET (1990-2008)

  • Precipitation
  • Temperature
  • Pot. evap.

Future Time Series CHANGE FACTORS STATISTICAL DOWNSCALING

Weather generator Mean & Variance Correction Mean Correction

Methodology

MIKE SHE Hydrological Model (Impact Assessment)

Scenario: A1B

slide-8
SLIDE 8

ENSEMBLES DATA

ARPEGE ECHAM5 HIRHAM ALADIN HIRHAM REMO Driving GCM: RCM:

OBSERVED DATA SET (1990-2008)

  • Precipitation
  • Temperature
  • Pot. evap.

Future Time Series CHANGE FACTORS STATISTICAL DOWNSCALING

Weather generator Mean & Variance Correction Mean Correction

Methodology

MIKE SHE Hydrological Model (Impact Assessment)

Scenario: A1B

slide-9
SLIDE 9

RCM compared to observations - precipitation

ALADIN-ARPEGE HIRHAM-ECHAM5 REMO-ECHAM5 HIRHAM-ARPEGE Observed

Mean St.dev. Skewness

  • Prop. dry days
slide-10
SLIDE 10

ENSEMBLES DATA

ARPEGE ECHAM5 HIRHAM ALADIN HIRHAM REMO Driving GCM: RCM:

OBSERVED DATA SET (1990-2008)

  • Precipitation
  • Temperature
  • Pot. evap.

Future Time Series CHANGE FACTORS STATISTICAL DOWNSCALING

Weather generator Mean & Variance Correction Mean Correction

Methodology

Mike SHE Hydrological Model (Impact Assessment)

Scenario: A1B

slide-11
SLIDE 11

Change factors precipitation (2070-2100)

0,50 0,60 0,70 0,80 0,90 1,00 1,10 1,20 1,30 1,40 1,50 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec CF Mean 0,50 0,60 0,70 0,80 0,90 1,00 1,10 1,20 1,30 1,40 1,50 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec CF StDev 0,50 1,00 1,50 2,00 2,50 3,00 3,50 4,00 4,50 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec CF Skew

  • 0,1
  • 0,1

0,0 0,1 0,1 0,2 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Absolute change Pdry

ARPEGE ECHAM5

Mean St.dev. Skewness

  • Prop. dry days
slide-12
SLIDE 12

ENSEMBLES DATA

ARPEGE ECHAM5 HIRHAM ALADIN HIRHAM REMO Driving GCM: RCM:

OBSERVED DATA SET (1990-2008)

  • Precipitation
  • Temperature
  • Pot. evap.

Future Time Series CHANGE FACTORS STATISTICAL DOWNSCALING

Weather generator Mean & Variance Correction Mean Correction

Methodology

MIKE SHE Hydrological Model (Impact Assessment)

Scenario: A1B

slide-13
SLIDE 13

Statistical downscaling (2070-2100)

0,5 1 1,5 2 2,5 3 3,5

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

Mean [mm/d] 1 2 3 4 5 6 7 8 9

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

Standard Deviation [mm/d] 2 4 6 8 10 12 14 16

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

Skewness [-] 0,4 0,5 0,6 0,7 0,8 0,9

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

Probability Dry days [-]

HIRHAM-ECHAM5 Observed

SNSRP -WG

Mean and Var Corr.

Mean Corr. Mean St.dev. Skewness

  • Prop. dry days
slide-14
SLIDE 14

Statistical downscaling (2070-2100)

Mean St.dev. Skewness

  • Prop. dry days

Mean correction Mean and variance correction Markov chain WG LARS WG NSRP WG

HIRHAM- ECHAM5

slide-15
SLIDE 15

Statistical downscaling - extreme events

30 50 70 90 110 5 10 20 50 100 200

Return Period Precpitation [mm]

Obs Mean correction Mean and variance correction SNSRP CI-95%

SNSRP > Mean and Variance Correction

>

Mean Correction

HIRHAM- ECHAM5

slide-16
SLIDE 16

Statistical downscaling - extreme events

Mean correction Mean and variance correction Markov chain WG LARS WG NSRP WG

HIRHAM- ECHAM5

slide-17
SLIDE 17

ENSEMBLES DATA

ARPEGE ECHAM5 HIRHAM ALADIN HIRHAM REMO Driving GCM: RCM:

OBSERVED DATA SET (1990-2008)

  • Precipitation
  • Temperature
  • Pot. evap.

Future Time Series CHANGE FACTORS STATISTICAL DOWNSCALING

Weather generator Mean & Variance Correction Mean Correction

Methodology

MIKE SHE Hydrological Model (Impact Assessment)

Scenario: A1B

slide-18
SLIDE 18

MIKE SHE model of NE Sealand

Precipitation Temperature and Pot. Evap.

[meter] Above 100 90 - 100 80 - 90 70 - 80 60 - 70 50 - 60 40 - 50 30 - 40 20 - 30 10 - 20 0 - 10
  • 10 -
  • 20 - -10
  • 30 - -20
  • 40 - -30
Below -40 Undefined Value

620000 640000 660000 680000 700000 720000 [meter] 6100000 6110000 6120000 6130000 6140000 6150000 6160000 6170000 6180000 6190000 6200000 6210000 6220000 [meter]

Downscaling

Precipitation:

  • Mean correction
  • Mean and variance correction
  • SNSRP weather generator

Temperature and pot. evap.

  • Mean correction
slide-19
SLIDE 19

MIKE SHE Results – water balance (2070-2100)

50 100 150 200 250 300 350 Obs HIRHAM-ECHAM HIRHAM-ARPEGE REMO-ECHAM ALADIN-ARPEGE CM Baseflow to river CMV Baseflow to river SNSRP Baseflow to river CM GWRecharge CMV GWRecharge SNSRP GWRecharge 0,8 1 1,2 Obs HIRHAM-ECHAM HIRHAM-ARPEGE REMO-ECHAM ALADIN-ARPEGE Prec 1

Baseflow & Recharge Precipitation

slide-20
SLIDE 20

MIKE SHE Results – extremes (2070-2100)

5 10 15 20 25 30 35 40 10 20 50 100 Discharge [m3/s] Return Period [years]

0,25 0,27 0,29 0,31 0,33 0,35 0,37 0,39 0,41 0,43

100 50 20 10 Discharge [m3/s] Return Period [years]

Annual Maximum Discharge Annual Minimum Discharge

HIRHAM- ECHAM5

slide-21
SLIDE 21

Concluding remarks

  • Statistical downscaling required for climate change impact

assessments

  • Scaling of GCM/RCM to the appropriate spatial and temporal

scales

  • Statistical adjustments of GCM/RCM
  • Choice of statistical downscaling procedure depends on

application

  • Water balance studies -> Mean correction
  • Extreme event analysis -> Stochastic weather generators
  • Assessment of uncertainties important
  • Scenario uncertainty
  • GCM/RCM uncertainty
  • Statistical downscaling uncertainty
  • Impact model uncertainty
  • Probabilistic projections needed for climate change impact

assessments and decisions on adaptation.

slide-22
SLIDE 22

Thank you for your attention

Henrik Madsen hem@dhigroup.com

HydroPredict 2010, 20-23 September 2010, Prague, Czech Republic