Robust adaptation decision-making under uncertainty: Real Options Analysis for water storage

Planning for climate change adaptation is challenging due to the inherent uncertainty associated with future climate changes. Although we have a range of climate projections, even these cannot provide a definitive picture of the future, with little certainty regarding the timing, magnitude and location of change. As a result there is increasing interest in approaches that can accommodate uncertainty better. A range of approaches exist and are being developed across several disciplines. Each has advantages and disadvantages and is suited for different types of decisions.

In this study we focus on one of these approaches, Real Options Analysis (ROA), and chose water storage for irrigation as an example to demonstrate its use in uncertain futures.

ROA is most suited to large, one-off investment decisions, such as water storage. Reservoirs are a significant investment, and without fully considering the range of future water availability, may turn out to be either too large, or not economically viable. Conversely, without their existence to smooth out water variability, significant production losses may be experienced in the future.

While we use a case study location in Canterbury, the intention with this analysis is to demonstrate the process of ROA, the type of information developed, and how it can be used to support decision-making under uncertainty. The approach can be applied to any type of large irreversible investment decision.

Using the full range of climate model and warming scenario combinations available in New Zealand (24), we develop estimates of water requirements out to 2090 for the chosen location. We specify two decision points: 2018 and 2050, and construct a decision tree with 256 potential paths. We develop a method to use the range of 26 GCM/RCP combinations available in New Zealand to generate estimates of likelihood. Using a backward induction technique, we identify the most cost-effective storage size. In this particular example, the most cost-effective option was to construct a reservoir for the most conservative climate change outcome in 2018, and in 2050. The decision to build for the smallest climate change is sensitive to a number of factors however, particularly the discount rate and the milk price, both of which result in larger reservoirs being more cost-effective in the first time period. The decision is also sensitive to assumptions regarding the likelihood of future climate changes – when the higher and lower climate scenarios are assigned a lower likelihood, this changes the decision to a larger storage size in 2018.

This analysis illustrates the benefits of ROA when the future is uncertain – by enabling decision-makers to adjust their decisions over time rather than locking themselves into a decision made now that has long-term consequences.

We used an example of water storage on a dairy farm to illustrate how ROA can be used, using the climate data available in New Zealand. This method is suitable for application across a range of investment decisions in New Zealand, where the initial cost is large and the investment is at least partially irreversible.

We believe that using ROA for these types of decisions will enable more cost-effective investment than cost-benefit analysis or other methods that use only a single climate scenario.