1. Predicting the Effects of Climate Change and Water Resources and Food Production in the Kennet Catchment Richard Skeffington, Aquatic Environments Research Centre Phillip Jones and Richard Tranter, Centre for Agricultural Strategy Potential Application to China? Integrated Project to evaluate the Impacts of Global Change on European Freshwater Ecosystems University of Reading

2. The Kennet Catchment 1137 km 2 Geology: chalk with clays at the East end Maximum altitude 297m above sea level Mean annual rainfall (1961-90): 759 mm Mean annual runoff (1961-90): 299 mm Theale

3. Kennet Agriculture Largely arable

4. Kennet Agriculture 2 Largely arable and livestock production

5. Kennet Land Use There are some urban areas (this is Reading) It is probably not very like China!

6. Problems on the Kennet 1. A low flow problem – the upper reaches can almost dry up in a dry summer 2. A (potential) nitrate problem – increasing concentrations Photo: Helen Jarvie

7. Modelling Agricultural Change CLIMATE CHANGE Change in river flows and composition Change in agriculture in catchment Changes in world agriculture Changes in crop prices and demand SOCIOECONOMIC CHANGE: population, global trade policies etc Is it possible to model these outcomes? …with any credibility?

8. © University of Reading 2008www.reading.ac.uk19 October 2015 Predicting the effect of climate change on water resources and food production Modelling land use impacts

9. Overview The socio-economic change scenarios IPCC SRES futures UKCIP refinements for UK BLS world food trade model The climate change scenarios HadCM3 projections The economic/land use model (CLUAM)

10. The SRES storylines Scenarios selected were: –A2 – low globalisation/market based solutions –B2 – low globalisation/sustainability led Local Stewardship Conventional Development Autonomy Community Interdependence Consumerism National Enterprise World Markets Global Sustainability

11. Climate change scenarios AOGCM HadCM3 (UK Hadley Centre’s 1 third generation coupled Atmosphere- Ocean Global Circulation Model) –This used with the A2 & B2 SRES scenarios to project to 2100 –Our modelling scenarios sample 2020 and 2050 1 Hadley Centre for Climate Prediction and Research (part of UK Meteorological Office)

12. Basic Linked System (BLS) -1- International Institute for Applied Systems Analysis (IIASA) Framework for analysing the world food trade system The BLS is an applied general equilibrium (AGE) model system – All economic activities are represented 34 national and/or regional geographical components – 18 eighteen single-country national models – 2 region model – 14 country groupings

13. Basic Linked System (BLS) -2- Market clearance (production and uses must balance) The model is recursively dynamic, ie, working in annual steps – For given prices calculate Global net exports and imports – Check market clearance for each commodity – Revise prices. When markets are balanced, accept prices as world market solution for year and proceed to next year This process is repeated until the world markets are simultaneously cleared in all commodities

14. BLS outputs Production levels (volumes) Market prices Technology change (yields) LUAM also requires climate-driven yield changes

15. Climate induced yield changes Two stage process: – Meta analysis of existing data on UK- specific crop yield changes due to climate change – Decisions on where crops would not grow due to climate limit

16. The CLUAM An LP model of England & Wales agriculture Range of major land using agricultural enterprises included –Outputs (revenue) –Inputs (incur costs) Land base partitioned by CEH Land Classification system Model objective maximize gross margin, –Subject to various constraints

17. Results – Agricultural Change

18. Livestock Numbers

19. Modelling Agricultural Change CLIMATE CHANGE Change in river flows and composition Change in agriculture in catchment Changes in world agriculture Changes in crop prices and demand SOCIOECONOMIC CHANGE: population, global trade policies etc

20. Downscaling in Space and Time This work has used the UK Climate Impacts Programme (UKCIP02) Scenarios, derived as follows. The INCA-N model for predicting nitrate and flow works on a daily time step & requires daily temperature, rainfall and evapotranspiration. HadCM3 c.300 km grid HadAM3H c.120 km grid HadRM3 c.50 km grid SRES Scenarios (4 future climates, including A2 and B2) “Experiments” run by the Hadley Centre Global Models European Model Monthly Time Step

21. More Downscaling HadRM3 c.50 km grid Monthly Kennet Catchment 5 km grid, Daily EARWIG Environment Agency Rainfall and Weather Impacts Generator Stochastic “weather generator” giving daily values for: Rainfall Potential evapotranspiration (Penman –MORECS or FAO) Min and Max temperatures (and others) Actual evapotranspiration estimated by a simple spreadsheet model constrained by soil water deficit.

22. EARWIG: Mean Monthly Temperatures Annual means: Base (1961-90) 9.2  C 202010.2  C 2050 B211.0  C 2050 A211.3  C

23. EARWIG: Mean Monthly Rainfall Annual Totals: Base 759 mm 2020s 787 mm 2050s 757 mm

24. How does INCA work?.. Each sub-catchment has 6 land uses: Urban; Forest; Arable + Oilseeds; Grassland; Unfertilised; Not covered by CLUAM. Catchment divided into sub-catchments

25. Land Cell: Hydrological Model Quick flow QuickSoil Groundwater Quick flow Throughflow Groundwater flow PAET Hydrological Model Abstraction (e.g. for water supply)

26. INCA-N Soil Processes

27. Land Uses and Fertiliser Inputs Each land use parameterised separately for all the above ScenarioArableGrassNot in CLUAM Urban, Forest Unfert. 1990180261500 Socio-2020 A2180263500 Econ.2020 B2180249500 2050 A2162276500 2050 B2180259500 Socio-2020 A2180269500 Econ. +2020 B2180248500 Climate2050 A2162272500 change2050 B2180283500 N Fertiliser in kg N ha -1 yr -1

28. IN-STREAM PROCESSES in INCA

29. Annual Hydrology Low summer rainfall protects the river from extra evaporation – to some extent

30. Period of River Recharge Shortens Consecutive months without hydrologically-effective rainfall

31. What Happens to Nitrate? 60-year realisation of nitrate in the R. Kennet: baseline climate EU Drinking water standard: 11.3 mg/L

32. Mean Nitrate Concentrations Crops in reference state (1990) Crop changes due to socio-economic factors only Crop changes due to socio-economic & climate change

33. Variation in Nitrate: 2050 A2 Socio-economic change makes a difference – adding climate change has no effect

34. Variation in Nitrate: 2050 B2 Socio-economic change makes small difference – adding climate change increases it

35. Other Modelling Work Same river, same climate scenario Different downscaling method, INCA parameterisation Nitrate increases in response to climate change!

36. Uncertainty

37. Conclusions It is possible to predict the effects of climate change on river flows and water quality, but a long chain of models and assumptions is required; Different assumptions can lead to radically different outcomes; These start at the top of the model chain – some GCMs give a substantial increase in rainfall by 2050 when downscaled to this catchment; The SRES Scenarios are looking a bit dated – need an “Energy Security” scenario? Better confidence on the hydrological predictions than the water quality – need to understand the effects of temperature and hydrological change on nitrogen cycle processes much better than we do; The work shows that potentially, changes in the world agricultural system can affect water quality at the catchment scale, but it is hard to predict what that influence might be in individual cases; Might have more predictive power at a more aggregated scale

38. Implications for China The methodology would be transferable, but the results of course are not; Technological and economic change is likely to be greater in China than the UK (?) and thus even more important as a driver of change; With current understanding, only worth doing at a highly aggregated scale May be more valuable in generating a set of plausible scenarios than in making predictions. THANK YOU