Box 1. Formal Application of Robust Decision Making to Adaptation

A comprehensive, formal application of RDM was undertaken by Lempert and Groves (2010) for Southern California’s Riverside County Inland Empire Utilities Agency (IEUA).

This study examined how climate change might affect IEUA’s 2005 Urban Water Management Plan (UWMP), a static 25 year plan to meet projected demand based on central projections of supply and demand, looking at uncertainty related to climate change and also the region’s growth and socioeconomic development trends using RDM.

The analysis followed the step by step process outlined earlier, i.e.:

In collaboration with IEUA stakeholders, it selected key performance measures (e.g. annual water demand, cost of supply, etc.) and then developed alternative management strategies. The latter included static strategies considering existing and augmented IEUA 2005 plans, as well as adaptive strategies based on a decision-tree framework within a Water Evaluation And Planning (WEAP) model environment (a water balance and management model).

The study then identified significant uncertainties. These included six key areas:
1. Future climate change (temperature, precipitation);
2. Future water demand;
3. Impact of climate change on imported supplies;
4. Response of groundwater basin to urbanisation and changes in precipitation patterns (percolation);
5. Achievement of management strategies (recycling program and groundwater replenishment);
6. Future costs (annual cost increases in imported supplies and efficiency).

The strategies’ performance across scenario futures and uncertainties was modelled using WEAP, starting with the original strategy, and working through a succession of 5-year signpost periods. The signposts evaluated the average difference between projected supply and demand to determine if the strategy should adopt an alternative course of actions, i.e. adopting an iterative approach.

The outputs were then input into the Computer Assisted Reasoning (CARs) program and analysed to determine the relative performance of various strategies over time. Performance was measured using projected present value (PV) costs in USD billions against PV shortage costs.

Based on the results of the CARs analysis, key trade-offs between the various strategies were summarised and robust alternative strategies recommended for consideration by stakeholders.

The analysis highlighted that a number of uncertainties would increase operating costs significantly (large declines in precipitation, larger-than-expected impacts of climate change on the availability of imported supplies and, reductions in percolation of precipitation into the region’s groundwater basin). In response, RDM analysis identified eight response strategies, four static and four adaptive. In each of the scenarios explored, it was found that the adaptive strategy leads to fewer vulnerable states than the static version.

The findings also showed that accelerating efforts in expanding the size of one of the agency’s groundwater banking programs and implementing its recycling program, while monitoring the region’s supply and demand balance and making additional investments in efficiency and stormwater capture if shortages are projected provides a promising robust adaptive strategy — and eliminates more than 80% of the initially-identified high-cost outcomes.

Case Study 2: Robust Decision Making for Climate Uncertainty

Dessai and Hulme (2007) present an example of the application for RDM to look at climate uncertainty for water supply in one of the driest regions in England, the East Suffolk and Essex (ES&S) Water Resource Zone (WRZ).

This area is vulnerable to future climate change, and a potential drying signal, with potential impacts on water security, as measured by the average available headroom (the difference between water available for use and demand) relative to target headroom (the minimum buffer allowed between supply and demand).

The analysis focused on the implications of uncertainty from climate change on proposed adaptation actions at a local/regional level, focusing in on water resource supply (not demand), and assessing the robustness of the existing 25 year plan (which had already built in adaptation to climate change using ensemble mean projections for alternative emission scenarios). The aim was to systematically assess the plan against the range of climate change projections and other uncertain parameters. The study included stakeholder consultation with local water managers.

The analysis focused particularly on isolating the threat that uncertain climate related parameters posed to supply-side security. They considered a series of climate uncertainties, including GHG emissions, climate sensitivity, carbon cycle, ocean diffusivity, aerosol forcing, regional climate response and climate impacts, looking at the potential effects one at a time (rather than in combination as in formal RDM and the previous case study, where interactions between uncertainties are explored).

This allowed quantification of the uncertainty introduced by the parameters sampled in the assessment. In turn, this was used to analysis whether the existing adaptation options identified were robust to the range of climate uncertainties.

Overall the findings indicated that the existing water plan was robust, primarily because it had already built in climate change considerations using one of the drier climate models available at the time of plan development. The analysis also strongly indicated that the largest uncertainty introduced into adaptation planning came from the regional climate response.

Some additional analysis was undertaken to look at the potential interaction of different factors, i.e. the cumulative uncertainty. This highlighted that under extreme conditions, further investments would be needed.

The analysis did not take into account the uncertainty around other factors, such as from the loss of groundwater supplies due to pollution, borehole deterioration, leakage, etc. which are important in looking at overall robustness, but it provides an useful case study into the consideration of climate robustness.