Agent based models of water abstraction reform policy options
Water abstraction is the process of removing water from natural sources like rivers, lakes and aquifers; it is regulated in England and Wales through a system of licences. Pressures from climate change and population growth mean that the current licensing system is no longer considered fit for purpose1.
The Government has committed to working closely with stakeholders to reform the abstraction system making it more flexible while retaining a high level of protection for water ecosystems. The Water Reform Abstractor Behaviour Model is central to this process. This agent-based economic behavioural model is being implemented for seven catchments across England and Wales.
The work is considering the different benefits, costs and risks of each reform option, and where possible quantifying the level and distribution of these impacts. Options include retention of the current system and reforms designed, among other things, to link abstraction limits to flow and make trading of abstraction rights easier. Models are run over the period 2025-2050 and explore the impacts of reform options under a range of future socio economic and climate scenarios. Results will be aggregated separately for England and Wales to inform the policy Impact Assessment.
Each Catchment Model (CM) comprises a novel combination of a hydrological model (HM) linked to an abstractor behaviour model (ABM).
The HM is a 1km gridded implementation of CatchMOD2 with additional ground water modules developed to provide appropriate ground to surface interactions. It estimates river flows and groundwater levels in each grid cell at a point in time and passes these to the behavioural model.
The ABM is an agent-based model developed within a proprietary tool3. It calculates the abstraction and discharge in each grid location at a daily time-step based on abstractors’ (agents’) demand requirements and decisions, and passes this information back to the hydrological model. It determines how abstractors might trade, interact and influence each other, and whether they will make investment decisions that affect the future availability of water and land use. It establishes the costs to abstractors associated with water abstraction and investment decisions.
Choices about adaptation, response to price signals, weather events, and interaction with other abstractors are incorporated into each agent’s behaviour. Agent heterogeneity is driven by water needs, location, relationships with other abstractors, local water scarcity and inherent decision characteristics such as propensity to innovate or follow neighbours, risk aversion etc.
Model development and validation is collaborative, being informed by literature review, detailed consultations with abstractors and sector representatives, catchment-based workshops and input from experts. Oversight is provided by a Project Board, an Abstractor Group, and independent peer review.
The work is funded by Defra.
1. Environment Agency, Case for Change, http://www.environment-agency.gov.uk/research/planning/135501.aspx
2. Cloke, H. L., Jeffers, C., Wetterhall, F., Byrne, T. Lowe, J. and Pappenberger, F. (2010) Climate impacts on river flow: projections for the Medway catchment, UK, with UKCP09 and CATCHMOD Hydrological Processes, 24, pp 3476–3489
3. Delphi Datasheet, http://www.embarcadero.com/images/dm/datasheet/Delphi-XE3-datasheet.pdf