Climate Change Scenarios for Agriculture

Climate Change Scenarios for Agriculture

Sam Gameda and Budong QianEastern Cereal and Oilseed Research Centre

Agriculture and Agri-Food CanadaOttawa, Canada

2

Objective

• Review some of the climate change scenarios being developed for agricultural impact and adaptation assessments

• Present AAFC research on climate variability and change in a Canadian context

3

Climate Change Scenarios

• Range of efforts on developing and using climate change scenarios– Global, Regional

• IPCC AR4• EU, ENSEMBLES (PRUDENCE, STARDEX, MICE)

– National• US

– Effects of Climate Change on Agriculture, Land Resources, Water Resources and Biodiversity (2008) Climate Change Science Program

– Climate Change Impacts for the Conterminous US (Climatic Change 2005 (Vol 69))

• UK, Climate Impacts Program– UKCIP02, UKCIP08

• Developing Countries– UNDP Climate Change Country Profiles (52 countries)

4

Highlights

• ENSEMBLES – Research and Application Function– Probabilistic estimate of uncertainty in future climate

• Seasonal, decadal, +

– Tool for statistical downscaling– Regional climate data sets

Work linked toEvaluation of, and recommendation on, systematic errors in GCM and RCM modelling

- higher resolution dynamical and/or statistical downscaling to provide projections and hindcasts

5

Highlights

• PRUDENCE– High resolution climate change scenarios for 2071-2100 for Europe using

regional climate models– Estimates of variability and level of confidence in the scenarios

• STARDEX– Intercomparisons of statistical, dynamical and statistical-dynamical

downscaling methods• Reconstruction of observed extremes• Construction of scenarios of extremes

• MICE– Direct use of climate models

• Evaluate capacity of climate models to reproduce observed extremes

6

Highlights

• UK Climate Impacts Program– Scenarios Gateway page

• Guidance on scenario use and development• Access to maps and datasets

• Canada – Climate Change Scenarios Network (CCSN)– Network of researchers providing scenarios and advice to the

impacts and adaptation community– Provision of CRCM output– On-line automated statistical downscaling tool, based on

SDSM

7

General Characteristics

• Global and regional changes in mean values– Annual, seasonal, (monthly)

• Useful for determining broad changes, e.g.– Growing season length– Moisture availability– Broad vulnerability to pests, disease

• Limitations for determining crop dynamics, pest hazard cycles

8

AAFC Climate Change Scenarios Research

• Background on climate change and scenario development

• AAFC weather generator• Findings on agroclimatic indices and extremes• Links to crop response

11

Downscaling

• Output from the nearest GCM grid point used at times to evaluate impacts of future climate

• However, downscaling is required to construct realistic regional or local scenarios from GCM outputs

• There are two main approaches to downscaling - dynamical and statistical

12

Dynamical approaches

• High-resolution atmosphere-only GCMs• Nested regional climate models (RCMs)• Formulated using physical principles• Computationally expensive• Parameterization schemes for processes at sub-

grid scales may be operating outside the range for which they were designed

13

Statistical approaches

• Regression-type models.• Weather generators.• Weather classification.• Analogue methods.• Computationally inexpensive; function at finer scales than

dynamical methods; applicable to parameters that cannot be directly obtained from RCM outputs

• Require observational data at the desired scale for a long enough period; assume that the derived cross-scale relationships remain valid in a future climate; cannot effectively accommodate regional feedbacks and can lack coherency among multiple climate variables under some approaches.

16

AAFC-WG

• An unconditional weather generator• Richardson-type weather generator• Precipitation occurrence is simulated by a two

state second order Markov chain• Precipitation amounts, temperatures and

radiation are simulated by empirical distributions of the observed data

17

AAFC-WG (continued)

• Validated for diverse Canadian climates• Has been compared with other stochastic

weather generators – LARS-WG• Evaluated for extreme daily values• Developed schemes for perturbing weather

generator parameters based on GCM-simulated changes in the statistics of daily climate variables

18

Climate Change Scenarios Data

Two sets of daily climate scenarios data• CGCM1 IS92a GHG+A and HadCM3 A2.• On 0.5°grids for south of 60°N• For the future time period of 2040-2069• Values of daily Tmax, Tmin, P and Rad• Generated by AAFC-WG

19

Scenarios data for ecodistricts

• Two data sets for CGCM1 IS92a (GHG+A) and HadCM3 A2.

• Developed with the “delta” method.• For the future period of 2040-2069.• Daily Tmax, Tmin, precipitation and Rad.• Centroids of ecodistricts where daily weather

data are available at neighbouring stations.

20

Some evaluations using the scenarios data

• Agroclimatic indices (e.g. frost-free days, last frost day in spring and first frost day in fall, GDD, EGDD, CHU, precipitation deficit)

• Annual and growing-season extreme values of daily Tmax, Tmin and precipitation, their 10yr, 20yr and 50yr return values

• Relative changes to 1961-1990 baseline climate

21

Last Frost in Spring(2040-2069)

CGCM1

HadCM3

22

Last Frost in Spring(Changes)

CGCM1

HadCM3

23

First Frost in Fall(2040-2069)

CGCM1

HadCM3

24

First Frost in Fall(Changes)

CGCM1

HadCM3

25

Frost-Free Days(Changes)

CGCM1

HadCM3

26

Precipitation Deficit/Surplus(2040-2069)

CGCM1

HadCM3

27

Precipitation Deficit/Surplus(Changes)

CGCM1

HadCM3

28

JJA Number of days Tmax≥30˚C (2040-2069)

CGCM1

HadCM3

29

JJA Number of days Tmax≥30˚C (Changes)

CGCM1

HadCM3

30

DJF Number of days Tmin≤-20˚C (2040-2069)

CGCM1

HadCM3

31

DJF Number of days Tmin≤-20˚C (Changes)

CGCM1

HadCM3

32

Corn yields increase about 0.6 t ha-1 for each increase of 100 CHU

y = 0.00583x - 8.23R2 = 0.86P<0.001

4

6

8

10

12

14

2200 2400 2600 2800 3000 3200 3400

CHU

Yie

ld (

t ha

-1)

CHU versus grain corn yields in eastern Canada

33

CHU versus soybeans yields in eastern Canada

Soybean yields increase about 0.13 t ha-1 for each increase of 100 CHU

y = 0.00133x - 0.68R2 = 0.74P< 0.001

2.0

2.5

3.0

3.5

4.0

4.5

2200 2600 3000 3400 3800

CHU

Yie

ld (

t ha

-1)

34

Barley yields versus Effective Growing Degree-Days above 5ºC (EGDD)

Increasing EGDD by 400 units reduces yield of 6-row and 2-row barley about 0.6 and 0.4 t ha-1, respectively

2 - Row Barley

y = -0.00098x + 5.84

R2 = 0.16 (NS)

3.0

3.5

4.0

4.5

5.0

5.5

1000 1200 1400 1600 1800 2000 2200

EGDDY

ield

(t

ha-1

)

6 - Row Barley

y = -0.0016x + 7.57

R2 = 0.26 P < 0.013

3.0

3.5

4.0

4.5

5.0

5.5

6.0

6.5

1000 1200 1400 1600 1800 2000 2200

EGDD

Yie

ld (t

ha

-1)

Barley yields versus water deficit (PE – P)

6 - Row Barley

y = -0.0001x2 + 0.0253x + 4.19

R2 = 0.18

3.0

3.5

4.0

4.5

5.0

5.5

6.0

6.5

-20 20 60 100 140 180

DEFICIT (mm)

Yie

ld (

t h

a-1

)

2 - Row Barley

y = -0.000145x2 + 0.0258x + 3.48

R2 = 0.41

3.0

3.5

4.0

4.5

5.0

5.5

-20 20 60 100 140 180

DEFICIT (mm)Y

ield

(t

ha

-1)

Average corn yields vs CHU – USA Locations1 = Illinois2 = Nebraska3 = Indiana4 = Iowa5 = Ohio6 = Missourii = irrigated

(based on average yield of top 10 hybrids in field trials, 4 to 8 yrs data, 1994-2001)

Agriculture and Agri-Food Canada

6i

6i

6i6i

6i

6i

6

6

6 6

6

6

6

6

55

5

5

5

5 5

5 55

44

4

4

4 4

43

33 3

3

2

2

2

2i

2i2i2i

1i 1i1

1i1

1i

1

6

10

14

3000 3500 4000 4500 5000

CHU

Yie

ld (t

/ha)

1 = Illinois2 = Nebraska3 = Indiana4 = Iowa5 = Ohio6 = Missouri(based on average yield of top 10 hybrids in field trials, 4 to 8 yrs data, 1994-2001)

Agriculture and Agri-Food Canada

Corn yields vs Water Deficits – USA Locations

6

6

66

6

6

6

6

55

5

5

5

5 55

5 5

44

4

4

44

43

333

3

2

2

2

1

11

6

10

14

100 150 200 250 300 350

Water Deficit (mm)

Yie

ld (t

/ha)

38

Planned scenarios

• New data sets for CGCM3 (A2, A1B, B1) and HadCM3 (A1B, B1)

• For the future period of 2040-2069• Gridded and/or ecodistrict scales• Continuous 2000-2099 data for a range of

stations• Daily Tmax, Tmin, precipitation and Rad.

39

Summary

There is a need for a shift from scenarios based on annual, seasonal, or monthly climate values, to daily ones for agricultural impact assessments.

AAFC-WG was suitable for the development of daily climate scenarios, and scenarios of extremes.

Earlier last frost in spring and later first frost in fall with a longer growing season are projected.

There would be an increase in crop heat units under climate change. Larger precipitation deficits can be expected, especially in the Prairies. An increase in extremely hot days in summer is foreseen. Increased crop heat units will likely result in increased production of corn and

soybeans, but decreased barley yields.

Crop response may be more sensitive to extremes. We will be carrying out more studies on the impacts of climate extremes. We will make use of crop modelling for this purpose.

THANK YOU!