Climate change impact assessment on maize production in ,

Meng Wang, Wei Ye and Yinpeng Li

1 Backgrounds

 APN CAPaBLE project with focus on integrated system development for food security assessment  Bio-physical & Economic

 Uncertainties: e.g. GCMs, CO2 emission scenarios  Adaptation measures (cross multi-scales)

2 SimCLIM model

Greenhouse gas MAGICC emission scenarios

Data Global Climate Projection Scenario selections

Climate and GCM pattern import Local Climate toolbox average, variability, extremes IPCC CMIP (GCMs) (present and future) USER -Synthetic changes - GCM patterns “Plug-in” Models Biophysical Impacts on: Agriculture, Coastal, - Land data Human Health, Water - Other spatial data

Impact Model

3 Case Study: Jilin Province

4 Climate Scenario

 Baseline Climate CRU global climatology dataset, 1961-1990 (New, 2000)

 Climate change scenarios • Pattern scaling (Santer, 1990; Mitchell, 2003) • 20 GCMs change patterns (Covey et al., 2003) • 6 SRES emission scenarios (IPCC, 2000)

5 DSSAT model – to simulate maize growth

 CERES-Maize model (Jones, 1986)

• Site-based, daily time step • Input – weather, soil, cultivating strategies, cultivar parameters • Output – yield, phenological parameters (e.g. growing season, growing phase date), etc.

6 DSSAT – weather generator

 SIMMETEO (Geng & Auburn, 1986)

• Input – monthly Tmax, Tmin, Rs, Prec. • Random seed sensitive 9.5 Ensemble 1 (b) 8.5 Ensemble 2

) Ensemble 3 -1 7.5 Ensemble 4 6.5

Yield (t ha Yield (t 5.5

4.5

3.5 0 20 40 60 80 100 120 Random seed So, the average result of 100-seed simulations

7 Climate change projections

Precipitation (mm) Baseline

2080 ensemble

8 Climate change projections Mean temperature (degC)

Baseline

2080 ensemble

9 DSSAT – Soil

 ISRIC-WISE soil dataset (Batjes, 2006) • 5 × 5 minutes • 14 FAO soil categories

10 DSSAT – cultivating strategies

Scheme Planting manner density: 6 plants m-2 date: controlled by soil temperature Fertilization nitrogen fertilizer (kg ha-1) applied by county Irrigation controlled by the total irrigation quota (350mm) and frequency

11 Calibration – cultivar parameters The initial ranges of 5 gene coefficients (P 1, P 2, P 5, G 2, G3 )

Produce 30 trial geno-types by the uniform design method  Cultivar parameters P1, P2, P5 Yield Simulation

G2, G3 Calculate Ie ; Find out the genotypes with the two smallest Ie s

The i -th ranges of gene coefficients

 11-site observations If (the differences between the ranges of these two genotypes are smaller than 5% of the initial No yield, upper range) then sowing date, Yes

harvest date The final range of gene coefficients

The coefficients are the average of the final range

12 Calibration – cultivar selection

Late Early

P1 280 270 Late P2 0.3 0.3

Early P5 790 700

G2 720 720

G3 8.5 8.5

13 Calibration – 11 sites (1996- 2000)

Site Cultivar Ysim/Yobs Msim/Mobs Changling Late 0.55 0.94 Nong’an Late 0.92 1.03 Yushu Late 0.87 1.06 Lishu Late 1.05 1.05 Ji’an Late 0.94 0.96 Early 0.87 1.06 Yongji Early 0.99 0.95 Early 1.23 1.13 Early 1.05 1.01 Early 0.94 0.94 Huadian Early 0.90 1.00

14 Validation – Compare with the county census yield (1990-2002)

 Average  Standard deviation

a b

15 Climate scenarios

 Climate change in Jilin Province • Growing season (Apr.-Sept.) • Temp. • Prec.

25 1000 (mm) Prec ) 20 800 ℃ • Prec/PET 15 600 10 400 5 200 0 0 SongyuanChangchun Siping Jilin Liaoyuan

2.0 1.6 1.2 P/PET=1.0

P/PET Temp ( P/PET 0.8 0.4 0 Baicheng SongyuanChangchun Siping Jilin Liaoyuan Baishan Tonghua Yanji

P/PET Prec Temp

16 Results – spatial pattern of yield change

Baseline 2020

2050 2070

17 Results – uncertainties among scenarios

8 8 Baicheng Jilin Baishan 7 7 6 6 5 5 4 4 3 3 2 2 1 2020s 2050s 2070s 2020s 2050s 2070s 2020s 2050s 2070s 1 0 0

) 8 8

-1 Liaoyuan Tonghua 7 7 6 6 5 5 4 4 Yield (t ha (t Yield 3 3 2 2 1 2020s 2050s 2070s 2020s 2050s 2070s 2020s 2050s 2070s 1 0 0 8 8 Changechun Siping Yanji 7 7 6 6 5 5 4 4 3 3 2 2 1 2020s 2050s 2070s 2020s 2050s 2070s 2020s 2050s 2070s 1 0 0 A1B A1FI A1T A2 B1 B2 Baseline Median

18 Results – the probability of yield reduction

5 10 15 20 30 40 50 5 10 15 20 30 40 50

100% 100% 80% Baicheng Jilin 80% 60% 60%

40% 40%

20% 20%

0% 0% 0 10 20 30 40 50 60 70 80 90 0 10 20 30 40 50 60 70 80 90

100% 100% 80% Songyuan Liaoyuan 80% 60% 60% 2020

Probability 40% 40%

20% 20% 2050 0% 0% 2070 0 10 20 30 40 50 60 70 80 90 0 10 20 30 40 50 60 70 80 90

100% 100% 80% Changechun Siping 80% 60% 60%

40% 40%

20% 20%

0% 0% 0 10 20 30 40 50 60 70 80 90 0 10 20 30 40 50 60 70 80 90

Yield reduction %

19 Results – changes in sowing date

Baseline 2020

2050 2070

20 Results – changes in flowering date

Baseline 2020

2050 2070

21 Results – changes in growing season

Baseline 2020

2050 2070

22 Results – yield and growing phase

8 160 8 200 Baicheng Dunhua 7 140 7 180 47 160 6 34 32 120 6 22 28 60 30 29 140 5 100 5 120 4 43 80 4 71 68 61 100 44 41 41 54 3 60 3 80 Yield (t ha-1)Yield (t ha-1)Yield

Duration (days) 60 Duration (days) 2 40 2 40 54 59 56 53 66 67 70 66 1 20 1 20 0 0 0 0 Baseline 2020s 2050s 2070s Baseline 2020s 2050s 2070s 8 160 8 200 Tongyu Huadian 7 140 7 180 160 6 34 120 6 33 30 28 140 30 37 51 65 5 100 5 120 4 43 43 43 43 80 4 100 65 3 60 3 61 55 53 80 Yield (t ha-1)Yield (t ha-1)Yield

Duration (days) 60 Duration (days) 2 40 2 58 56 56 54 40 1 20 1 51 50 51 49 20 0 0 0 0 Baseline 2020s 2050s 2070s Baseline 2020s 2050s 2070s

Sowing - Tasseling begins Tasseling begins - Filling begins Filling begins - Filling ends Auto Irr Control Irr

23 Results – adaptation analysis: irrigation

10 Baicheng 8 ) -1 6 Baseline

4 Yield (t ha (t Yield 2 2020 2050 2070

0 350 450 550 650 750 350 450 550 650 750 350 450 550 650 750 Irrigation quota (mm) 10 Tongyu 8 ) -1 Baseline 6

4 Yield (t ha (t Yield 2 2020 2050 2070

0 350 450 550 650 750 350 450 550 650 750 350 450 550 650 750 Irrigation quota (mm)

25-75th Percentile 10-90th percentile Median

24 Conclusion

 Impacts on maize yield west & central areas eastern mountain areas  Impacts on phenology shortened growing season in the west and central (especially the reduced grain filling phase)  Adaptation measures Increase in total irrigation (20- 50 years) New cultivar with longer growing season (>50 years)

25 Thank You

26