New Projecting climate change impacts: how robust are our methods?uest.ntua.gr/adapttoclimate/proceedings/full_paper/... · 2015. 11. 27. · Generic assessment methods ... Project

Post on 10-Oct-2020

4 Views

Category:

Documents

0 Downloads

Preview:

Click to see full reader

Transcript

International Conference ADAPTtoCLIMATENicosia, Cyprus, 27-28 March 2014

Projecting climate change impacts: how robust are our methods?

Timothy CarterFinnish Environment Institute, SYKE

International Conference ADAPTtoCLIMATENicosia, Cyprus, 27-28 March 2014

Outline

Introduction: why focus on impacts? Methods of projecting future impacts Investigating uncertainties Are impact models "fit for purpose"? Conclusions

International Conference ADAPTtoCLIMATENicosia, Cyprus, 27-28 March 2014

Why focus on impacts?

International Conference ADAPTtoCLIMATENicosia, Cyprus, 27-28 March 2014

Possible objectives

To assess:

• biophysical effects of climate change (magnitude and rate)

• economic costs of climate change (damages, benefits, effectiveness)

• socio-political and cultural effects of climate change

• integrated effects of climate change, including feedbacks

• the need for adaptation

International Conference ADAPTtoCLIMATENicosia, Cyprus, 27-28 March 2014

The adaptation learning cycle

Source: PROVIA, 2013

International Conference ADAPTtoCLIMATENicosia, Cyprus, 27-28 March 2014

Monday 31 March 2014, Yokohama

Release of the IPCC Working Group II Fifth Assessment Report (AR5):Climate Change 2014: Impacts, Adaptation, and Vulnerability

International Conference ADAPTtoCLIMATENicosia, Cyprus, 27-28 March 2014

Methods of projecting future impacts

International Conference ADAPTtoCLIMATENicosia, Cyprus, 27-28 March 2014

Generic assessment methods• Expert elicitation• Analogues

• Historical periods/trends• Past extreme events• Geographical analogues

• Controlled experiments• Indicator mapping• Impact modelling

• Empirical/statistical• Process-based• Site-scale through to global-scale• Single sector through multi-sector to fully integrated

International Conference ADAPTtoCLIMATENicosia, Cyprus, 27-28 March 2014

Investigating uncertainties

International Conference ADAPTtoCLIMATENicosia, Cyprus, 27-28 March 2014

Uncertainties in impact model projections

Model uncertainties Input data used for calibration

Parameter uncertainties (within-model)

Structural uncertainties (between model)

Projection uncertainties Model sensitivity/robustness

Scenario assumptions (climate and non-climate)

Scenario application (e.g. number; downscaling methods)

International Conference ADAPTtoCLIMATENicosia, Cyprus, 27-28 March 2014

Scenarios for impact assessment

Scenarios of future climate, socioeconomic development, land use, other environmental factors

Earlier studies used IPCC scenarios (IS92 and SRES) New scenarios framework (RCPs/SSPs) developed

independent of IPCC

International Conference ADAPTtoCLIMATENicosia, Cyprus, 27-28 March 2014

Source: van Vuuren et al., 2014

RCP/SSP scenario matrix architecture

RCP: Representative Concentration PathwaySSP: Shared Socioeconomic Pathway

RCP

International Conference ADAPTtoCLIMATENicosia, Cyprus, 27-28 March 2014

Scenarios for impact assessment

Scenarios of climate, socioeconomic development, land use, other environmental factors

Earlier studies used IPCC scenarios (IS92 and SRES) New scenarios framework (RCPs/SSPs) developed

independent of IPCC Availability for impact assessment:

New access to AR5-assessed, CMIP5 climate data and scenarios from IPCC DDC

Basic (global) socioeconomic scenarios (ICONICS)

Mapping of new scenarios against SRES

Preliminary work on developing extended socioeconomic storylines and scenarios

International Conference ADAPTtoCLIMATENicosia, Cyprus, 27-28 March 2014

Scenario application

Number of scenarios (e.g. representativeness) Probabilistic projections (risk assessment) Downscaling methods Representing changes in averages and variability Reference period for observations/projections

International Conference ADAPTtoCLIMATENicosia, Cyprus, 27-28 March 2014

Uncertainties in impact model projections

Model uncertainties Input data used for calibration

Parameter uncertainties (within-model)

Structural uncertainties (between model)

Projection uncertainties Model sensitivity/robustness

Scenario assumptions (climate and non-climate)

Scenario application (e.g. number; downscaling methods)

Red cross: standard HadSM3 climate projection; default CATCHMOD versionLight blue curve: standard HadSM3 climate projection; 100 CATCHMOD versionsDark blue curve; default CATCHMOD version; 449 climate projectionsRed curve: all possible combinationsBlack curves: each climate projection; 100 CATCHMOD versionsGreen curves: each CATCHMOD version; 449 climate projections

Changes in median flow simulated with a hydrological model (CATCHMOD) when model parameter uncertainties are combined with the climateprediction.net (CP.net) ensemble of climate model projections

Source: New et al. (2007)

International Conference ADAPTtoCLIMATENicosia, Cyprus, 27-28 March 2014

Simulated winter wheat yield from eight crop models (M1–M8), ensemble mean (Mean) and observed (Obs)

Model runs for eight sites in Europe

N = number of growing seasons with observed yields

Whiskers: min & max yields

Vertical dashed lines: observed min, median and max yields

Source: Palosuo et al. (2011)

Observed yields

Multi-model mean yields

International Conference ADAPTtoCLIMATENicosia, Cyprus, 27-28 March 2014

Source: Schellnhuber et al., 2014

Assessing climate impacts across sectors

Inter-Sectoral Impact Model Intercomparison

Project (ISI–MIP)

International Conference ADAPTtoCLIMATENicosia, Cyprus, 27-28 March 2014

Are impact models "fit for purpose"?

International Conference ADAPTtoCLIMATENicosia, Cyprus, 27-28 March 2014

-2 0 2 4 6 8

-40

-20

020

40

DE, S_wheat, 2008, ARMOSA

Temperature change (C)

Prec

ipita

tion

chan

ge (%

)

750

1500

2250

3000

3750

450

0

525

0 6

000

675

0

750

0

825

0

750 1500

2250

3000 3750

450

0

525

0 6

000

675

0

750

0 8

250

75015002250300037504500525060006750750082509000975010500112501200012750135001425015000157501650017250180001875019500

Grain yield kg/haBaseline

Impact response surfaces

Source: Pirttioja et al. (in prep.)

International Conference ADAPTtoCLIMATENicosia, Cyprus, 27-28 March 2014

Lleida, ES, Winter wheat, 30-year mean yield change (%) relative to unperturbed

Source: Pirttioja et al. (in prep.)

ANONYMOUS MODEL: Lleida, ES, Spring wheat, 30-year mean yield change (%) relative to unperturbed

International Conference ADAPTtoCLIMATENicosia, Cyprus, 27-28 March 2014

Proposed co-ordinated cycle of model improvement and projection

International Conference ADAPTtoCLIMATENicosia, Cyprus, 27-28 March 2014

International Conference ADAPTtoCLIMATENicosia, Cyprus, 27-28 March 2014

International Conference ADAPTtoCLIMATENicosia, Cyprus, 27-28 March 2014

PDFs: sampled uncertainties

International Conference ADAPTtoCLIMATENicosia, Cyprus, 27-28 March 2014

Food per capita Water exploitation index

PDFs: sampled uncertainties

International Conference ADAPTtoCLIMATENicosia, Cyprus, 27-28 March 2014

Summary: impact modelling Intercomparison studies are needed, but be cautious

of model convergence Scale issues (global versus regional/local) Sensitivity analysis – are models "fit for purpose"? Distinguish scenario analysis from uncertainty

analysis Treatment of adaptation – implicit or explicit? Ongoing work (e.g. HELIX, IMPRESSIONS, RISES-AM,

JPIs, COST)

International Conference ADAPTtoCLIMATENicosia, Cyprus, 27-28 March 2014

Conclusions

International Conference ADAPTtoCLIMATENicosia, Cyprus, 27-28 March 2014

Conclusions (1/2) Impact assessment is crucial for identifying

adaptation needs New initiatives on scenario development (e.g.

ICONICS, IPCC DDC) Ongoing work is applying impact model ensembles

and intercomparison

International Conference ADAPTtoCLIMATENicosia, Cyprus, 27-28 March 2014

Conclusions (2/2)Some priority issues for future impact assessments include:

• Support for basic research and capacity building in model development and application

• Collection of empirical and experimental data to underpin model development

• Scenario development with stakeholders using the RCP/SSP framework at relevant scales

• Extension of model intercomparison exercises (sectors, scales, economic models)

• Full treatment of impact uncertainties• Improved treatment of impacts in integrated assessment models• Design and implementation of coordinated cycles of model

International Conference ADAPTtoCLIMATENicosia, Cyprus, 27-28 March 2014

That's it

International Conference ADAPTtoCLIMATENicosia, Cyprus, 27-28 March 2014

Notice

Colleagues are welcome to incorporate these slides into their own presentations, assuming they are correctly

acknowledged. However, the authors would also appreciate being informed prior to the extensive

use of this material in public meetings.

top related