Some very general steps for identifying and evaluating planned adaptations are given in Carter et al. (1994) and UNEP (1998). Somewhat more detailed procedures for evaluating anticipatory adaptation policies in the climate change context are outlined in Smith and Lenhart (1996) and Smith (1997). This approach addresses management of institutional processes and players and proposes net benefits and implementability as central evaluative criteria. Numerous other considerations are noted, including flexibility, benefits independent of climate change ("no regrets"), local priorities, levels of risk, and time frames of decisions. From a disaster management perspective, Tol et al. (1996) argue that policies must be evaluated with respect to economic viability, environmental sustainability, public acceptability, and behavioral flexibility. Tol et al. (1999) apply these observations in an examination of adaptation to increased risk of river floods in The Netherlands. They note several possible adaptations, but none could be accomplished without creating significant distributional and/or ecological impacts. None, therefore, would be feasible without enormous political will and institutional reform. Klein and Tol (1997) and UNEP (1998) describe methodologies for evaluation, including cost-benefit, cost-effectiveness, risk-benefit, and multi-criteria methods. Multi-criteria methods to evaluate possible adaptation options have been demonstrated for coastal zones (El-Raey et al., 1999) and agriculture (Mizina et al., 1999).

Fankhauser (1996) provides an economic efficiency framework in which adaptation actions are considered justified as long as the additional costs of adaptation are lower than the additional benefits from the associated reduced damages. Optimal levels of adaptation (in an economic efficiency sense) are based on minimizing the sum of adaptation costs and residual damage costs. Such studies require the definition of a base case that involves estimation of autonomous adaptations. These and other normative studies (e.g., Titus, 1990; Goklany, 1995) illustrate the range of principles and methods that have been proposed for identifying, evaluating, and recommending (planned) adaptation measures.

There are, however, few comprehensive estimates of the costs of adaptation. Mimura and Harasawa (2000) report estimates of 11.5-20 trillion Yen as the cost of maintaining the functions of Japanese infrastructure against a 1-m rise in sea level. Yohe and Schlesinger (1998) applied a cost-benefit rule to adaptation decisions across a sample of the developed coastline of the United States. With a 3% discount rate, their national estimates of the expected discounted cost of protecting or abandoning developed coastal property in response to sea-level rise that is based on a mean greenhouse emissions scenario is US$1.3 billion with foresight and US$1.8 billion without. Their estimates climb to more than US$4 billion and 5 billion, respectively, along the 1-m sea-level rise scenario that matches the Mimura and Harasawa study. Between 55 and 70% of these costs were attributed to planned adaptation. The remainder reflect estimates of residual damage associated with abandoning property with and without completely efficient autonomous adaptation. Indeed, the differences between the foresight and non-foresight estimates can be regarded as estimates of the incremental cost of incomplete autonomous adaptation in advance of planned responses.

On a more local scale, Smith et al. (1998) report cost estimates that are clearly sensitive to design and evaluation criteria. For example, none of the five flood protection strategies for the southernmost part of the Dutch Meuse (assuming 10% more winter precipitation and a warming of 2°C) would achieve economic benefits that exceed their costs of DGL 243-1,505 million, given a 5% discount rate. Moreover, only building quays would meet the benefit-cost standard with a 5% discount rate. Nonetheless, the government chose a wildlife renovation strategy on the basis of additional benefits for nature and recreation. Smith et al. (1998) also report that the cost of raising the Northumberland Bridge between Prince Edward Island and New Brunswick to accommodate a 1-m sea-level rise would be US$1 billion or 250,000, depending on whether the entire bridge or only the portion that spanned the shipping lanes were raised.

Klein et al. (1999) develop a conceptual framework of the process of planned adaptations, aimed at changing existing management practices in coastal zones. In this model, adaptation is a continuous and iterative cycle, involving several steps: information collection and awareness raising, planning and design (incorporating policy criteria and development objectives), implementation, monitoring, and evaluation (see Figure 18-6).

Research increasingly addresses how adaptation is considered in actual policy decisionmaking. Stakhiv (1996) and Frederick (1997), dealing with the U.S. water resources sector, conclude that existing institutions and planning processes can deal with climate change; such processes essentially represent adaptive management. As in many other sectors and circumstances, adaptation to climate change hazards in the coastal zone is part of ongoing coastal zone management. Adaptation to sea-level rise and extreme climate events is being included in Japanese coastal policies (Mimura and Kawaguchi, 1997), British shoreline management (Leafe et al., 1998), and Dutch law and coastal zone management (Koster and Hillen, 1995; Helmer et al., 1996; Klein et al., 1998).

Planning of adaptation invariably is complicated by multiple policy criteria and interests that may be in conflict (Hareau et al., 1999). For example, the economically most efficient path to implement an adaptation option might not be the most effective or equitable one. Moreover, decisions have to be made in the face of uncertainty (Lempert et al., 2000). Uncertainties that are pertinent to adaptation are associated with climate change itself, its associated extremes, their effects, the vulnerability of systems and regions, conditions that influence vulnerability, and many attributes of adaptations, including their costs, implementability, consequences, and effectiveness (Campos et al., 1996; Lansigan et al., 1997; Handmer et al., 1999; Murdiyarso, 2000).

Given these uncertainties, it is not surprising that adaptation strategies frequently are described as forms of risk management. For example, adaptations to deal with climate change impacts or risks to human health can be biological (acquired immunity), individual (risk-aversion options), or social (McMichael et al., 1996). Most social adaptation strategies are measures to reduce health risks via public health programs (Patz, 1996; McMichael and Kovats, 2000). Similarly, public adaptations via "disaster loss mitigation" (Bruce, 1999) are mainly risk management initiatives such as improved warning and preparedness systems, less vulnerable buildings and infrastructure, risk-averse land-use planning, and more resilient water supply systems. Nguyen et al. (1998), Hisschemöller and Olsthoorn (1999), and Perez et al. (1999) also describe adaptations to climate change and extremes as modifications to existing risk management programs. As de Loë and Kreutzwiser (2000) and others point out, it remains unclear whether practices designed for historical climatic variability will be able to cope with future variability.

To recognize these uncertainties, decision tools to help evaluate adaptation options include risk-benefit and multi-criteria analyses (Klein and Tol, 1997). Such evaluations are further complicated by the existence of secondary impacts related to the adaptation itself. For example, water development projects (adaptations to water supply risks) can have significant effects on local transmission of parasitic diseases, including malaria, lymphatic filasiasis, and schistosomiasis (Hunter et al., 1993; McMichael and Kovats, 2000). Improved water supply in some rural areas of Asia has resulted in a dramatic increase in Aedes mosquito breeding sites and, consequently, outbreaks of dengue (WHO, 1997). Langen and Tol (1999) provide examples of technical response options to climate hazards that are counterproductive in the longer term. Existing resource management programs do not necessarily consider changed risks or recognize local interests and inequities (Primo, 1996). Wilhite's (1996) analysis of programs in the United States, Australia, and Brazil shows the ineffectiveness of reactive crisis management approaches and the need for proactive and cooperative planning.

Nonetheless, it is widely accepted that planned adaptations to climate risks are most likely to be implemented when they are developed as components of (or as modifications to) existing resource management programs or as part of national or regional strategies for sustainable development (Campos et al., 1996; Magalhães, 1996; Theu et al., 1996; Mimura, 1999a; Apuuli et al., 2000; Munasinghe, 2000; Osokova et al., 2000).

IPCC Homepage