Siwa ifpri water&mktprojections_wmoc_washdc_june2015_final
0 likes546 views
This document summarizes approaches to modeling the linkages between water and agriculture in economic models. It provides examples of global and regional models that have incorporated water constraints and relationships. Key challenges include capturing both the physical quantities and spatial dimensions of water use, as well as limitations in data on irrigated areas and water allocation. The document concludes with a quantitative experiment simulating the impacts of groundwater declines in northern India on global food production, prices, and trade.
Siwa ifpri water&mktprojections_wmoc_washdc_june2015_final
- 1. Imposing constraints on water in market projections Accounting for environment in market models Siwa Msangi IFPRI World Market Outlook Conference Washington DC, USA 8-9 June, 2015
- 2. The linkage between water and food Next to land, water is one of the most important resources supporting agriculture • It’s not always managed (as in rainfed agriculture) – but remains essential to growth • In biophysical modelling of crop growth – water is a key yield-limiting factor (along with nitrogen) and is always made explicit • In economic modelling of agriculture – water tends to only appear in activity-based farm models (if at all) – esp if irrigation is involved
- 3. Various approaches and key challenges to modeling water and agriculture linkages Page 3
- 4. Water in Ag farm or market models Water fits more ‘naturally’ into farm-level ag models • It can be accounted for as one of the inputs to production (w/in an explicit yield or prodn function) • Even if yield is fixed and input use is accounted for by Leontief coeffs – water can be a constraint • Even w/o explicit pricing of water (which is rare in agriculture) – shadow values can be derived • W/in a multi-market framework – the yield-water relationship becomes more reduced-form and the representation of water tends to become less explicit – except for a few cases where water accounting modules have been added
- 5. Examples of mkt models with water Type of model Name (institution) Key features Global PE IMPACT (IFPRI) Inter-sectoral water alloc model determines avail for ag. Supported by hydrological balance model and calculation of yield-water relationship Watersim (IWMI) Approach largely based on IMPACT – different way of representing basin efficiencies and water alloc rules GLOBIOM (IIASA) Disagg prodn systems into rainfed/irrig, includes irrig tech & costs, EPIC captures water-balance & crop cons use (no non-ag use) Global CGE GTAP-W Numerous variants of GTAP-based models incorporating water • Calzadilla, Tol, Redahnz – now a formalized GTAP-W database • Mirage variant Regional PE CAPRI (IPTS/U Bonn) Exploratory effort to incorporate water (starting with 2 NUTS regions – Andalusia and mid-Pyrenees) – w/irrig & water use modules
- 6. Model examples with water (cont) Type of model Name (institution) Key features Regional PE USMP/REAP (USDA) Regional math programming model which can take resource constraints like water directly into account Various ag sector models (ASME – Egypt) Math programming models with explicit constraints and yield-water relationships SWAP model (Univ California) Uses explicit prodn function – linked to statewide hydrological model to obtain surface & GW availability Country-level PE-GE analysis Terry Roe & Xinshen Diao (various papers with co-authors 2000- 2005) Combines a top-to-bottom and bottom-up linkage b/w country-level CGE model for Morroco to illustrate: • The impact of trade policy changes on farm-level water use • The impact of changes in property rights & water mgmt on mkts, prices and other sectors at country-level
- 7. Key challenges to capturing water The question of ‘quantities vs prices’ applies to how one choses to capture water in ag farm or mkt models • Primal vs dual – many find it preferable to capture the physical aspects of water use in ag (kg per m3) rather than the cost side (given rare pricing of water) • The biophysical requirements of crop growth wrt water are fairly well-known and can be captured in agronomic modelling approaches (i.e. yld penalties) • There is a spatial as well as a temporal dimension to water – it matters when in the crop cycle the deficit happens, and where water is located relative to the crops that need it (for irrigation) • The water dimensions of livestock are typically not well- captured in any modeling (biophysical or economic) • Nothing on aquaculture either (esp for inland systems) - data • Water quality dimensions are also rarely addressed
- 8. Key challenges (cont.) As in all aspects of modelling agriculture – data is always a challenge (more so for some regions) • Getting a good handle of how much irrigated area there really is – can be challenging for some regions • Sometimes the definition of what really is irrigated is not straightforward (e.g. the Fadama in Nigeria) • Often times there are discrepancies between various sources of data (OECD, FAO Aquastat, Kassel/U Frankfurt) • This is a key piece of data to have if doing basin- level water modelling in order to determine the balance between ET, runoff, precip, deep percolation
- 9. Quantitative Experiment: Groundwater Depletion in India & Implications for Global Food Balance Page 9
- 10. Simple quantitative experiment • In this experiment, we simulate what would happen if the groundwater availability in northern India (Gujarat, Rajasthan, Haryana, Uttar Pradesh, Madhya Pradesh, Bihar) were to decrease dramatically over 2010-2020 • Essentially halving the water available for irrigation (since GW supplies ~50% irrigated area) • Simulated over the corresponding IMPACT basins (Indus, Ganges, Mahi-Tapti & Luni basins) • Observe the impact on food production, prices, consumption and malnutrition in India & the world Page 10
- 11. Ganges Mahi-Tapti Luni Indus Key Indian basins targeted in scenario
- 12. Global cereal production changes Page 12
- 13. Shifts in global cereal net trade to compensate Page 13
- 14. Increases in global cereal prices Page 14
- 15. Page 15 A more widespread problem in India would be even more dramatic ( “Too Big to Fail” !) Similar effects would be observed if the North China Plain were subjected to such a scenario This is purely illustrative of the importance of India to the global food balance – and the implications of it falling into deficit due to environmental impacts This underscores the importance of water for food Summary The impact of groundwater declines in northern India have a sizable impact on food production, trade and security in both India and the world Page 15
- 16. Conclusions • Water is one of the key constraints to ag – but not widely captured in ag mkt model • Need to take the first step – separate harvested area b/w irrigated & rainfed – If rainfed – enters as an exog factor – if irrigated – is part of on-farm mgmt decisions • Tend to prefer the biophysical approach to capture the impacts on yield • Ag water can be valued but is rarely priced – therefore a mkt-based alloc is more of a “scenario” rather than the baseline case
- 17. Thank you!