Will managing for climate variability also manage for climate change? - Andrew Moore
Download as PPTX, PDF1 like294 views
This document summarizes a study examining how different adaptation policies for a sheep grazing system in southern Australia may perform under climate change. The study uses models to simulate the system under historical weather from 1970-2009 and projected future climates from 2010-2099. It finds that climate change impacts will not be masked by natural variability and that adaptation policies like incremental changes or those based on long-term forecasts may outperform traditional or step-change approaches. Inter-decadal variability is also an important source of uncertainty. The best policy depends on the specific climate scenario, and future work could blend incremental and forecast approaches while considering ensembles of climate projections.
Will managing for climate variability also manage for climate change? - Andrew Moore
- 1. Will managing for climate variability also manage for climate change? A southern Australian grazing system as an example CSIRO Climate Adaptation Flagship Andrew Moore
- 2. Adaptation – how to get there from here? • Pannell (2010) on adaptation policy in agriculture: • Climate change will be slow, uncertain, masked by variability • “At any point in time, farmers will use those practices that suit current perceived conditions ... the pace of climate change is predicted to be easily slow enough for pre-emptive action to be unnecessary...” • To what extent do these arguments hold in livestock production? • Constraints on the rate of management change, imposed by: • Longer-term consequences on (e.g.) flock or herd structure • Less scope to adapt key profit drivers to seasonal conditions • Is recent history a good guide to “current perceived conditions”? • If so, at what time scale?
- 3. A case study using the GRAZPLAN models Dual-purpose Merino enterprise at Lucindale • Self-replacing flock + 1st cross lambs • Mediterranean climate, duplex soils • Annual grass-clover pastures • Confinement feeding in poor years Simulated with the GRAZPLAN models • Historical weather data 1970-2009 • 2010-2099 projections from ECHAM5/MPI-OM and NCAR-CCSM3 under SRES A1B • 4 realizations/GCM • Downscaled after Zhang (2009) • Stochastic prices for each realization Stocking rates & joining dates were varied
- 4. Adaptation policies Traditionalist Continue with the optimal policy from 1970-2009 (7.5 ewes/ha, mid-Jan) Incremental Each year, make a small change in either stocking rate or joining date based on the last 5 years’ performance Step-change Every 15 years, change to the policy that would have performed best over the previous 15-year period Forecast Smoothly changing policy that maximizes expected performance across all realizations (i.e. using an accurate long-term forecast) • Each policy evaluated for 2 GCMs x 4 realizations • “Profit” here is gross margin less an operator allowance • Risk-averse producer assumed when defining “performance” • Conditional value-at-risk used as the risk measure
- 5. “Traditional” policy: impacts of climate change • ECHAM5/MPI-OM, averaged over 15-year running averages Comparing GCMs, 4 realizations realizations, 15-year running realizations, averages • Inter-decadal variability is a major source of uncertainty across realizations • Expected pasture production stays fairly stable • Clover-dominant pastures in the model (earlier in NCAR-CCSM3)
- 6. S.R. & joining dates over time: NCAR-CCSM Incremental Step Change Forecast
- 7. S.R. & joining dates over time: ECHAM5/MPI-OM Incremental Step Change Forecast
- 8. Long-run-average profit & its variability Traditional Incremental Step Forecast Change Mean Historical 290 1970-2009 Mean CCSM 267 242 252 262 2010-99 ECHAM5 236 265 244 250 RMSD CCSM 76 71 70 73 Across Years ECHAM5 63 53 77 73 RMSD CCSM 107 74 95 96 Across Realizations ECHAM5 96 157 120 111
- 9. Lessons learnt For this particular livestock production system: • climate change will be slow and uncertain • climate change won’t be masked by climatic variability • confinement feeding is a promising adaptation Inter-decadal variability is important The “incremental” and “forecast” approaches may work better than the “step change” approach • but no policy consistently outranked another over all 8 realizations • Blend the “incremental” & “forecast” approaches? Adaptation studies focussed on (e.g.) 2050 may need to work with ensembles of GCM realizations • We await CMIP5 with interest...
- 10. CSIRO Plant Industry Andrew Moore Project Leader, Southern Livestock 2030 Phone: +61 2 6246 5298 Email: Andrew.Moore@csiro.au Web: www.csiro.au Thank you This project is supported by funding from the Australian Government Department of Agriculture, Fisheries and Forestry under its Climate Change Research Program, Meat & Livestock Australia, Dairy Australia and Australian Wool Innovation