Hydrologic simulations of global warming impact using dynamically downscaled data: case study of the Seyhan river basin in Turkey

Transcription

1 Hydrologic simulations of global warming impact using dynamically downscaled data: case study of the Seyhan river basin in Turkey 9 September 29 Yoichi FUJIHARA Japan International Research Center for Agricultural Sciences

2 Background: Mekong region Turkey Japan International Research Center for Agricultural Sciences (JIRCAS) Intensification of water use and diversification of farm operations through farmer participatory approach in rainfed agriculture of Indochina (Rainfed agriculture project) Research Institute for Humanity and Nature (RIHN) Impact of climate changes on agricultural production system in arid areas (ICCAP) Project leader: Prof. Tsugihiro WATANABE The study region is Mediterranean region, Turkey Seyhan River basin (21,7 km2) Research period: 22 April 27 March 2

3 Why in Turkey? future (27 299) present ( ), A O 6 I 2 18 r _ '1 ir!il' mtuilyear [ CGCM [ -5 5 O ECHAM5 6O [ -5 5 O Decrease HadCM3 Increase [ -5 5 O Decrease CCCMA Increase 3

4 Research organization Japan Climate Turkey Hydrology and Water Resources Vegetation RIHN Irrigation and Drainage Socio Economics Crop Production TÜBITAK Scientific and Technical Research Council of Turkey 43 collaborators 54 collaborators Advisers Coordinators 6 Researchers from the third countries and International organizations 4

5 Seyhan River basin in Turkey 1 E 2 E 3 E 4 E 5 E 6 E 7 E Istanbul Ankara 5 N 4 N 3 N 5 Adana 2 N 1 E 2 E 3 E 4 E 5 E 6 E 7 E Seyhan River

6 Outline of Seyhan River basin I (u)uoquaa t' OOZE I'Z LE 19E JSE PtLE M8E 19E MLE icc M8E M6E Lower Seyhan Irrigation Project (LSIP) 6 Cotton Corn Citrus

7 I I I I I Hydrological characteristics Rainfall: lower 7mm/y, middle 1mm/y, upper 4mm/y Runoff: 28mm/y(5.5 billion m3) 6 I I I I I I I I I ipitation (mm) I I (fl 5 4 ;:o 3 S Mean Monthly ( ) > C C) 'I, I Jan. Feb. Mar. Apr. May Jun. Jul. Aug. Sep. Oct. Nov. Dec. Rainfall I I I I I I F I I Jan. Feb. Mar. Apr. May Jun. Jul. Aug. Sep. Oct. Nov. Dec. Runoff 7

8 Reservoir volumes and water use [xi9j M 7. g6.o Observed -- Irrigation water -- Drinking water C.) 1 2.O ill Reservoir volume Seyhan:.88 billion m 3 Catalan: 1.6 billion m Year

9 Approach used in this study Present (199s) Future (27s): MRI CGCM2, SRES A2 Future (27s): CCSR MIROC, SRES A2 Dynamical downscaling method (25 km 1km scale) Present meteorological data Future meteorological data: (MRI) Future meteorological data: (CCSR) Assumed social scenario and Land use (Control, Scenario1, 2, ) Hydrologic models and Reservoir models Output for Scenario1 Output for Scenario1 Output for Scenario1 Output for Scenario2 Output for Scenario2 Output for Scenario2 9

10 Assumed social scenario and land use Present Actual conditions Irrigation area is 133, ha Future Basin conditions are the same as present Control run Scenario 1 Winter rain fed wheat is converted to pasture, due to little precipitation Lower Seyhan Irrigation Project (LSIP) IV is completed. Irrigation area is 174, ha Scenario 2a 2e Forest decreases, due to little precipitation LSIP IV is completed, and irrigated farmlands are newly developed Irrigation area is 174, ha + (2a) 2, ha, (2b) 4, ha, (2c) 6, ha, 1

11 Present and future land use 39N N 38N andcover class legend O:Water surface :Evergr Needleleaf Forest 4:Decid Broadleaf Forest 5:Mixed Forest 6:Shrub 7:Shrub/grass 8:Savanna 1 O:Crassland 1 :Maize 1 2:Dry cropand 1 3:Urban area 1 4:Crop/Natural 1 6:Barren 17:Cftras 37.8N- :1. Jr..?L!.' fl, tin.; 37N N 345E 35E 35.5E Present 1 36E 36.5E 37E Future 11

12 Scenario 1 39N N andcover class legend O:Water surface :Evergr Needleleaf Forest 4:Decid Broadleaf Forest 5:Mixed Forest 6:Shrub 7:Shrub/grass 8:Savanna 1 O:Crassland 11:Maize 1 2:Dry cropand 1 3:Urban area 1 4:Crop/Natural 16:Barren 1 7:Cftras - a.rr T 38N 37.8N- 37N N 345E 35E I I' 35.5E 36E 36.5E 37E 12

13 Scenario 2a 2e 39N N 38N andcover class legend O:Water surface :Evergr Needleleaf Forest 4:Decid Broadleaf Forest 5:Mixed Forest 6:Shrub 7:Shrub/grass 8:Savanna 1 O:Crassland 1 :Maize 1 2:Dry cropand 1 3:Urban area 1 4:Crop/Natural 1 6:Barren 17:Cftras 37.8N- 37N N 345E 35E 35.5E 1 36E 36.5E 37E 13

14 Water demand for each scenario Scenario Note Irrigation area (ha) Water demand (Gm3/y) Present, Future Scenario 1 Scenarios 2a 2e Present water use 133, 1.6 LSIP IV is completed LSIP IV is completed, and irrigated farmlands are newly developed 174, 2.8 (a) 174,+2, 2.32 (b) 174,+4, 2.56 (c) 174,+6, 2.8 (d) 174,+8,

15 GCMs River basins The GCMs outputs are inadequate to assess the impacts of climate change on river basins. The temporal and spatial resolutions of GCMs are too coarse compared with those of hydrologic models Pseudo global warming downscaling method has been recently developed GCM OUTPUT DOWNSCALING CATCHMENT 15

16 Dynamic downscaling method Present ( ) NCEP Re Analysis data RCM(Regional Climate Model) Present meteorological data sets Future (27s) NCEP Re Analysis data + RCM ΔT, ΔP,... (27 s global warming) by GCM under A2 Future meteorological data sets 16

17 Application of downscaling method Period Present: 199s (1 year), Future: 27s (1 year) GCMs MRI CGCM2 (MRI) CCSR/NIES/FRCGC MIROC (CCSR) SRES A2 RCM TERC RAMS (Tsukuba Univ.) Outputs Hourly, resolution: about 8.3km Prec, downward shortwave / long wave radiation, wind, temp, pressure, specific humidity Bias correction (1) DS by RCM, (2) use, (3) characteristics of bias, (4) turning of RCM Repeat this procedure (1) (4) many times Finally, statistical bias correction was done Precipitation (mm Present(1 yr) 34.5E 35.5E 36E

18 Temperature and Precipitation (present) 4 I I I I I I I I I I I I 15 I I I I I I I I I I I I p 3o Observed - -a- Downscaled (Raw) -- Downscaled (BiasCorrected) Observed Downscaled (Raw) Downscaled (BiasColTected) ci 2O I I I I I I I I I I I I I Jan. Feb. Mar. Apr. May Jun. Jul. Aug. Sep. Oct. Nov. Dec. 1 rlftft. a I Jan. Feb. Mar. Apr. May Jun. Jul. Aug. Sep. Oct. Nov. Dec. Temperature Rainfall 18

19 Application of hydrologic models Simulation region 34.25E 37.E and 36.5N 39.25N 5 min latitude longitude spatial resolution (33*33 grids). iii 35E 36E 3TE H 39N 39N Zamtj Qoks.i 38 38N Basin information DEM: Gtopo3 land use dataset: LANDSAT Soil (porosity, field capacity, and root zone dept): ECOCLIMAP 3T Cadan Dam Med4terraneaI S eseiwoir anlyaml'eservoir 35E 36E 37T o OO Elevation(m) I 3VN 19

20 SiBUC Simple Biosphere including Urban Canopy(Tanaka and Ikebuchi, 1994) land surface model The SiBUC simulates snow accumulation and melt, soil moisture dynamics and evapotranspiration, and surface runoff and base flow Mc Thr Mbr Thw Canopy Urban Canopy Tdu ) Mw Mug Urban Ground Tug Twb Tdw Water Body Mg Tg Ground Wi Surface Layer Tdg W2 Root Zone W3 Recharge Zone 2

21 River flow network The simulated surface runoff and base flow in each grid cell were routed using the flow routing network The flow direction in each grid cell was defined using a digital elevation model (DEM), and flow routing was performed using the kinematic wave method Reservoir model Seyhan:.88 billion m3 Catalan : 1.6 billion m3 Station: 1818 (1 1 %.alaictii L'ain Seyhan Dam 21 I I I- I o 6 ( Elevation(rn)

22 Hydrograph (present) ct 2 I I I I I I Observed Simulated (Raw) Simulated (BiasCorrected) E C C.) -t o o Observed Simulated (Raw) Simulated (BiasCorrected) 5. - I I I I I I I I I Monthly Runoff Annual Runoff 22

23 Reservoir models rc 'C Full reservoir level Operation rule curve Limited water level Jan. Feb. Mar. Apr. May Jun. Jul. Aug. Sep. Oct. Nov. Dec. Rule curve Water is stored in order to maintain the operation rule curve, and the water is released to meet the demand regardless of the rule curve (Fujihara et al., 28) C > 4 j 2 - d(h Observed Simulated Observed Simulated Observed and simulated reservoir volume and discharge (Seyhan)

24 Approach used in this study Present (199s) Future (27s): MRI CGCM2, SRES A2 Future (27s): CCSR MIROC, SRES A2 Dynamical downscaling method (25 km 1km scale) Present meteorological data Future meteorological data: (MRI) Future meteorological data: (CCSR) Assumed social scenario and Land use (Control, Scenario1, 2, ) Hydrologic models and Reservoir models Output for Scenario1 Output for Scenario1 Output for Scenario1 Output for Scenario2 Output for Scenario2 Output for Scenario2 24

25 Temperature and rainfall changes 3 I I I I I I I I 12 I I I I I I I I C a 2 a H 2 -e- Present -i*- Future (MRI) Future (CCSR) +2. C (MRI) +2.7 C (CCSR) C C a Li p 1 E4 8 iii 6 -I I 2 Present Future (MRI) I I Future (CCSR) 157 mm (MRI) 182 mm (CCSR) I I I I I I I I I I I Jan. Feb. Mar. Apr. May Jun. Jul. Aug. Sep. Oct. Nov. Dec. Temperature a I I Jan Feb. Mai. Api. May Jun. Jul Aug. Sep. Oct. No. Dec. Precipitation 25

26 Snow and river flow changes,4 o s ::,25 o O S PrNCEP) - Warm-up(MI rr - up(ccsf- Bain Total Snow (Gt) SiBUC 2 C Ionthly Inflow (n 1 4 N-) Present '' Future (MRI) Future (CCSR) 15 1oo C' S Jov Dec Ja Feb Mar Apr May Ju, 2 I I I I I I I I I I I I Snow water equivalent River flow 26

27 Water balance Precipitation (mm) Evaporation (mm) Flow (mm) Present MRI-future 476(-25%) 375(-9%) 19(-52%) CCSRfuture 452(-29%) 372(-1%) 89(-61%) 27

28 Reservoir models 'S C 1 Control - Scenario 1 Scenario 2a Scenario 2b - Scenario Scenario 2c Scenario 2d 2e 'S C 1 Control - Scenario 1 Scenario 2a Scenario 2b - Scenario Scenario 2c Scenario 2d 2e k Present MRI future 28

29 1.6 Reservoir reliability = Supply / Demand water o-- Control Scenario 1 Scenario 2a -o- Scenario 2b -o-- Scenario 2c Scenario 2c1 Scenario 2e 1.4 -o-- Control Scenario 1 Scenario 2a -o-- Scenario 2b -o-- Scenario 2c --- Scenario 2c1 Scenario 2e 1. S C Sc1jtS S U J.4.2 t) U J Present MRI future Scenario Note Irrigation area (ha) Water demand (Gm3/y) Present, Future Present water use 133, 1.6 Scenario 1 LSIP IV is completed. 174, 2.8 Scenarios 2a 2e 29 LSIP IV and developed (a) 174,+2, 2.32 irrigated farmlands. (b) 174,+4, 2.56

30 But, can we say scenario 1 is safety in future?? fruit soybean watermelon vegetables cotton citrus maize II maize Irrigation area planted area (thousand decare) year Year Sugestion1: Cropping pattern changes should be taken into account from now. Water use is increasing already regardless of climate change Total water use Annual planned release (Mm3)

31 Why water scarcity will not occur despite the impacts of climate change in scenario 1?? Reservoirs are quite large Seyhan:.88 billion m3 Catalan : 1.6 billion m3 Inflow volume Present: 5.5 billion m3 Future: 2.64 (MRI), 2.15 (CCSR) billion m3 Moreover, environmental flow from these reservoirs has not been introduced. They can store the water as possible as they want. There is very little flow at the down stream now especially in summer Sugestion2: Environmental flow from reservoirs should be taken into account. Some problems may be already taking place 31

32 Snow is key hydrologic variables in the basin Future: Impact on snow is large and inflow will decrease in April and May Present Snow melt sometimes lead to floods in early spring If snow amount is monitored in winter, it is possible to forecast the inflow in spring, and relatively easy to operate reservoirs for flood control and water resources. However, precipitation observation system has not been established in the basin Sugestion3: Meteorological observation system in mountain area should be developed immediately 32

33 Conclusions We developed an approach that uses dynamically downscaled data as inputs to hydrologic simulations and applied it to the Seyhan River Basin in Turkey. We assessed the impact of changes on the water resource systems of the basin. Suggestions to the basin Cropping pattern changes should be taken into account from now. Water use is increasing already regardless of climate change Environmental flow from reservoirs should be taken into account. Some problems may be already taking place Meteorological observation system in mountain area should be developed immediately CC studies are not only for future problems but also for present (existing) problems 33

34 Thank you for your attention!! 34

35 35

36 But, can we say scenario 2b (LSIP IV + 4, ha) is safety in future?? fruit soybean watermelon vegetables cotton citrus maize II maize Irrigation area planted area (thousand decare) year r Fruit 2L<B:762mm!y Cotton LJ2V: 542mm1y / From May Up to Oct Month Regional Technical 1 Workshop on Application 67 of Modelling Tools for Year Total intake Annual planned release (Mm3) Irrigation water

37 Reservoir volume Seyhan Dam E 'C C Control Scenario 2c Scenario 1 Scenario 2d Scenario 2a Scenario 2e Scenario 2b E 5 37 Climate Change 273 Impact 274 and 275 Vulnerability Assessment FH! ii

38 Reservoir reliability =Supply / Demand o-- Control Scenario 1 Scenario 2a -o- Scenario 2b -o-- Scenario 2c --- Scenario 2c1 Scenario 2e U j Regional Technical. Workshop on Application of Modelling Tools for 271 Climate 272 Change 273 Impact and Vulnerability Assessment