The prediction of seasonal and inter-annual climate - in the eastern - - PowerPoint PPT Presentation

▶
the prediction of
SMART_READER_LITE
LIVE PREVIEW

The prediction of seasonal and inter-annual climate - in the eastern - - PowerPoint PPT Presentation

Nov 16, 2023 225 likes •497 views

The prediction of seasonal and inter-annual climate - in the eastern seaboard of Thailand - Department of Geohydraulics and Engineering Hydrology, University of Kassel Introduction : background / overview Short-prediction : seasonal prediction

slide-1
SLIDE 1

The prediction of seasonal and inter-annual climate

  • in the eastern seaboard of Thailand -

Department of Geohydraulics and Engineering Hydrology, University of Kassel

slide-2
SLIDE 2

Introduction Short-prediction Summary Long-prediction Impacts

2

Introduction : background / overview Short-prediction : seasonal prediction Long-prediction : inter-annual prediction Impacts : climate change impacts Summary : results summary

slide-3
SLIDE 3

1.Water crisis during drought (Rayong, East of TH, 2005)

Jan Dec

no more water supply

Sample impacts : Climate change / traditional management

2.Water crisis during Flood (North and Central of TH, 2011) no supply

Jan Dec

Spillway

  • verflow

flood

Introduction Short-prediction Summary

3

Long-prediction Impacts

pattern of climate fluctuation changes, season shift and extreme weather

slide-4
SLIDE 4

Objectives : prediction of climate

To investigate of the climate change impact toward water resources

Development

  • Climate prediction tool for local area

by employing statistical/stochastical framework

Prediction

  • Local weather toward the impact of

climate change on water/climate

short and long term

Ultimate goal

Water resources planning Introduction Short-prediction Summary

4

Long-prediction Impacts

slide-5
SLIDE 5

& \ & \ & \ & \ & \ & \ & \ & \ & \ & \ & \ & \ & \ Bang Phra Res. Nong Khoa Res.

A.K A.Sattahip A.Ban Khai A.Ban Bung A.Si Racha A.Bo Thong A.Nong Yai A.Ban Chang A.Wang Chan A.Phan Thong A.Bang Lamung A.Pluak Daeng i Chang K.A.K A.Phanat Ni Khom

" !

2

" !

1

" !

3

" !

4

" !

4

" !

9

" !

6

" !

7

" !

5

W ater Supply Area

Dog Krai Res.

Indus trial Area

Khlong Yai Res.

Nong Pla Lai Res.

Irrigation Area

N

Thailand

Rayong Chonburi

Sea (Thai Gulf) Thai Gulf

Coastal basin with 3 reservoirs

1560 km2 Khlong Yai basin

Study area / Pilot area : economic-heart of TH

drought crisis in 2005

Introduction Short-prediction Summary

5

Long-prediction Impacts

slide-6
SLIDE 6

Introduction Short-prediction Summary Long-prediction Impacts

6

  • seasonal prediction
  • up to 1 year
  • using Teleconnection
  • development of AR models
slide-7
SLIDE 7

Teleconnection : connection between local & ocean climate

7 ±

75 150 225 300 375 37.5 Kilometers

Autocorrelation coeff (within -/+11 months)

Thailand borderline
  • 1.0
  • 0.9
  • 0.8
  • 0.7
  • 0.6
  • 0.5
  • 0.4
  • 0.3
  • 0.2
  • 0.1
0.0 +0.1 +0.2 +0.3 +0.4 +0.5 +0.6 +0.7 +0.8 +0.9 +1.0

Mean min. temp. vs NINA 1.2

Correlation coef.

  • 0.5

+0.7

Equator

Nino 4 Nino 3 Nino1+2 Nino 3.4 SWIO WTIO SETIO SOI Thailand

Correlation of

minimum temp

and

Ocean state index ENSO Niño 1+2 with 3 month lag

Introduction Short-prediction Summary Long-prediction Impacts

7

Nino1+2

slide-8
SLIDE 8

29 30 31 32 33 34 35 36 37 temperature (deg C)

  • bs Tmax

prediction : HiRes+SSTs (with teleconnection) prediction : HiRes (no teleconnection)

short-prediction enhancement : by teleconnection

8

Introduction Short-prediction Summary Long-prediction Impacts

8

reduce prediction residuals

prd seasonal developing in reduce RMSE annual dry premonsoon monsoon1 monsoon2 Tmax 13% 3% 54% 18% 35% Tmin 3% 2% 4% 17% 0% PCP 5% 5% 5% 6% 11%

slide-9
SLIDE 9

short-term climate prediction

  • p order of autoregressive terms
  • d order of integrated term (non-seasonal differences; linear, quardatic, etc.)
  • q order of moving average (forecast errors)

9

Introduction Short-prediction Summary Long-prediction Impacts

9

slide-10
SLIDE 10
  • 1.0
  • 0.8
  • 0.6
  • 0.4
  • 0.2

0.0 0.2 0.4 0.6 0.8 1.0 HiRes GCMs GCMs+HiRes SSTs ECHO-G + SST GCMs+SSTs GCMs+HiRes+SSTs HiRes+SSTs ARIMAex-GCMs ARIMAex-HiRes ARIMAex-SST GCMs GCMs+teleconnection ARIMA

average Nash-Sutcliffe coefficience Tmax Tmin PCP

Model performance : short-term

1

Introduction Short-prediction Summary Long-prediction Impacts

10

(Nash,1970)

slide-11
SLIDE 11

Introduction Short-prediction Summary Long-prediction Impacts

11

  • climate downscaling
  • GCMs
  • conventional tools
  • development of new models
slide-12
SLIDE 12

Conventional downscaling tools

statistical downscaling model (Wilby, 1999)

Introduction Short-prediction Summary Long-prediction Impacts

12

stochastic downscaling model (Racsko, 1991; Semenov, 2002 )

SDSM LARS-WG

slide-13
SLIDE 13

Developing of downscaling model

y = x1𝛾1 + ⋯ + x𝑞𝛾𝑞 + 𝜁

  • y dependent climate variable vector (local climate)
  • x independent GCM predictor vector
  • 𝛾 regression coefficients of GCM predictor

Introduction Short-prediction Summary Long-prediction Impacts

13

slide-14
SLIDE 14

multi-domain GCMs

Introduction Short-prediction Summary Long-prediction Impacts

14

experiments domain resolution model number sub- model MLR predictor

  • 1. Single-domain MLR

single domain 2.5°x2.5° ECHO-G 1 21

  • 2. Hi-Res MLR

single domain 0.5°x0.5° High-Resolution grid 5 5x5

  • 3. Multi-domain MLR

multi domain 2.5°x2.5° ECHO-G,BCCR, ECHAM5,GISS,PCM 1 1340

  • 4. Multi-domain

+HiRes MLR multi domain 2.5°x2.5°+ 0.5°x0.5° ECHO-G,BCCR, ECHAM5,GISS,PCM,Hi-Res 5 1340+5x5

MLR experiments 24 stations : precipitation 4 stations : max & min temperature : humidity 2 stations : solar radiation

slide-15
SLIDE 15

multi-season : optimal seasonal schemes

  • ptimal number of predictors

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec no variation 2 season 3 season 4 season 2) pre-monsoon 3) monsoon_1 4) monsoon_2 1) Dry single period 1) Dry 1) 2) Wet 1) Winter 1) 2) Summer 3) Rainy

Introduction Short-prediction Summary Long-prediction Impacts

15

seasonal schemes

season 1 season 2 season 3

max_temp[season 1] = 0.345*giss_model_e_r.r1.ta.20000 + 0.1549*giss_model_e_r.r1.ta.15000 + 0.0959*mpi_echam5.r2.rsutcs + 0.254*ncar_pcm1.r1.ta.40000 - 149.23 max_temp[season 2] = 1.4e8*ncar_pcm1.r1.tro3.92500 + 0.609*mpi_echam5.r2.rsutcs - 12.46 max_temp[season 3] = 0.438*ncar_pcm1.r1.ta.30000 -0.109*ncar_pcm1.r1.ta.25000
  • 0.217*mpi_echam5.r1.ta.1000 +0.765

R2 number of predictors

no season

AIC R2 AIC R2 AIC R2 AIC

max_temp = -1.2e8*ncar_pcm1.r1.tro3.92500 + 1.95e8*ncar_pcm1.r1.tro3.70000
  • 0.012*giss_model_e_r.r1.rsus + ... - 18.98
slide-16
SLIDE 16

Model performance : Long-term

Introduction Short-prediction Summary Long-prediction Impacts

16

(Nash,1970)

  • 0.69
  • 0.27
  • 0.15

0.00 0.44 0.13 0.27

  • 0.31
  • 0.13

0.37

  • 0.40

0.71 0.59 0.73 0.87 0.79 0.88 0.68 0.53 0.76 0.41 0.30 0.37 0.59 0.57 0.62 0.22 0.38 0.55

  • 1.0
  • 0.8
  • 0.6
  • 0.4
  • 0.2

0.0 0.2 0.4 0.6 0.8 1.0 SDSM LARS-WG MLR MLR MLR MLR MLR AR ARIMA ARIMAex-GCM ECHO-G ECHO-GECHO-G HiRes CMIP3 CMIP3 +HiRes

  • CMIP3

+HiRes daily ensemble daily ensemble monthly ensemble monthly ensemble monthly ensemble conventional single-domain MLR multi-domain MLR autoregressive Nash–Sutcliffe model efficiency coefficient Tmax Tmin PCP

slide-17
SLIDE 17

Stochastic generation of daily climate

Introduction Short-prediction Summary Long-prediction Impacts

17 time series of daily precipitation 1971-2006

time series of downscaled monthly precipitation amount 2006-2096

time series of daily max/min temperature 1971-2006

daily precipitation time series of downscaled monthly precipitation prob. 2006-2096 stochastic generation of daily precipitation

  • ccurrence

daily wet/dry state stochastic generation of daily max/min temperature precipitation amount on wet days precipitation parameters estimation temperature parameters estimation temperature on wet and dry days monthly precipitation probability monthly precipitation amount stochastic generation of daily precipitation amount maximum temperature minimum temperature

precipitation generation temperature generation

slide-18
SLIDE 18

Introduction Short-prediction Summary Long-prediction Impacts

18

  • impact studies
  • application of predicted results
  • hydrological impacts
slide-19
SLIDE 19

climate change

Introduction Short-prediction Summary Long-prediction Impacts

19

2 4 6 8 10 12 14 16 18 monthly preicpitation (mm/day) Year

  • bs PCP

sim PCP 20c3m sim PCP A1B sim PCP A2 sim PCP B1 12 per. Mov. Avg. (sim PCP A1B) 12 per. Mov. Avg. (sim PCP B1)

  • bs PCP
20c3m A1B A2 B1 sim PCP

18 23 28 33 38 43 temperarure (deg C) Year

  • bs Temp

sim A1B sim A2 sim B1 sim 20c3m 12-month moving avg. linear trend

A1B A2 A1B A2 B1 Tmax B1 Tmin 20c3m
slide-20
SLIDE 20

2 4 6 8 10 12 14 16 18 20 22 24 monthly runoff (cms)

30 rlz sim sb4 avg 30 rlz sim sb4

  • bs Z4
  • bs climate sim sb4
calibration verification

200 400 600 800 1000 1200 1400 1600 1800 2000 amount of water (mm/year) year Soil+Surface ET PERC PCP.obs.sim ET.obs.sim PERC.obs.sim

20c3m SRES evapotranspiration precipitation percolation

Hydrological study

1971

Meteorological data

HMD, SLR

2 4 8 6 3 9 10 1 7 5 11 12

1971

Introduction Short-prediction Summary Long-prediction Impacts

20

(Arnold, 1998)

2096 2096

runoff

hydrological components

slide-21
SLIDE 21

Impact on Hydrology

Introduction Short-prediction Summary Long-prediction Impacts

21

57.3% 61.8% 69.3% 67.7% 69.2% 72.0% 68.8% 69.1% 24.5% 22.7% 19.3% 20.2% 18.7% 17.9% 19.6% 19.0% 18.2% 15.5% 11.4% 12.1% 12.1% 10.1% 11.6% 11.9%

0% 10% 20% 30% 40% 50% 60% 70% 80%

  • bs.sim

20c3m A1B A2 B1 A1B A2 B1 1980-1999 2000-2049 2000-2096

ratio to amount of precipitation

ET PERC Soil+Surface

average hydrological component along 21st century +12 to 15 % ET (evaportranspiration)

  • 5 to 7 % PERC (groundwater recharge)
  • 6 to 8 % Soil+Surface (surface water)
slide-22
SLIDE 22

Impact on Water Resources

Introduction Short-prediction Summary Long-prediction Impacts

22

5 10 15 20 25 30 35 40 45 monthly water flow-in (cms) year A1B.mean 20c3m / A2.mean B1.mean

  • bs.sim

rlz.max rlz.min

  • est. flow-in

20c3m SRES

rlz.max rlz.min
  • est. flow-in
  • bs.sim
20c3m SRES A1B/A2/B1 2 4 8 6 3 9 10 1 7 5 11 12

A1B

Forecaste up to the end 21st century

A2 B1 A1B A2 B1

slide-23
SLIDE 23

Introduction Short-prediction Summary Long-prediction Impacts

23

  • ARIMA is most optimal for short-term
  • Teleconnection improve seasonal prediction
  • MLR performs best with mix of multi-domain
  • Less & extrem surface water supply in 21st century
slide-24
SLIDE 24

Operation : forecast short term

Dec 2011 Jan 2011 Precipitation (current)

Equator

Nino 4 Nino 3 Nino1+2 Nino 3.4 SWIO WTIO SETIO SOI Thailand

Ocean state indices Jan 2012 Feb 2012 Prediction (future)

Introduction Short-prediction Summary Long-prediction Impacts

24

slide-25
SLIDE 25

Planning : planing long term

Introduction Short-prediction Summary Long-prediction Impacts

25

1971

Meteorological data

HMD, SLR

2 4 8 6 3 9 10 1 7 5 11 12

1971

(Arnold, 1998)

2096 2096

slide-26
SLIDE 26

 Water Resources System Research Unit, Chulalongkorn

University, Thailand (WRSRU_CU)

 Royal Irrigation Department, Thailand (RID)  Thai meteorological department, Thailand (TMD)

Questions & Answers

26