Are Forecasts Still for Wimps?

Preface

The authors, along with Helen Ingram and Mark Houck, a water engineer at George Mason University, developed three case studies of water resource managers' awareness of and possible use of probabilistic forecasts¹ in the Pacific Northwest (PNW), the Los Angeles Basin, and the Potomac River Basin with funding from the National Oceanic and Atmospheric Administration (NOAA) (Lach et al. 2003, 2005, 2006; Rayner et al. 2005). These three case studies represented the ecosystems familiar to at least one of the social scientists, although we worked on all three cases together. While we were working on the project, there was a popular commercial claiming that the advertised SUV could overcome any weather conditions—snow, rain, sleet, wind—and, therefore, any "forecasts are for wimps." This slogan struck us as particularly apt when we started talking with water resource managers who had been dealing with extreme weather—floods, droughts, storms—for decades. They had built water infrastructure and systems—dams/dikes, storage, distribution—to withstand wide swings in temperature and precipitation, ensuring that water resources could be delivered regardless of the forecast or actual weather. While all the authors are knowledgeable about water resources, Dr. Ingram brought a wealth of knowledge about the political and institutional arrangements necessary for delivering water resources to a wide variety of constituents. Her understanding and experience helped make the original research insightful and provocative in challenging how new forms of information can be integrated into existing decision routines.

Introduction

In the years since our original research on the use of climate change information in water resource decisions, knowledge about and evidence for climate change continue to grow (e.g., National Research Council 2007, 2009, 2010). The availability of ensemble-based climate forecasts,² for example, has increased with the aim of better quantifying the inherent uncertainties in the weather/climate system (e.g., Doblas-Reyes et al. 2005). Yet, the National Research Council and others (Lemos et al. 2012; National Research Council 2009, 2010) report a continuing gap that echoes what we found years ago between the production of climate information and the use of such information for decision making. In our earlier research we heard that climate information products, such as the seasonal climate predictions, offered the promise of improved management but were not sufficiently "reliable" or spatially precise for use in the water resources sector. In addition, institutional and ecological complexities of individual water agencies made it difficult to produce the types of tailored information our respondents believed would be useful for local decisions. Analysis of the three case studies in California, the Pacific Northwest, and metropolitan Washington, D.C., identified several additional reasons behind managers' reluctance to use the forecasts:

• • Traditional reliance on large built infrastructure like dams, storage, and distribution systems to mitigate the effects of weather and/or climate

• • Organizational conservatism and complexity

• • Mismatch of temporal and spatial scales of forecasts with management needs

• • Political disincentives to innovation

• • Regulatory constraints

We also heard from our respondents that even if the forecasts could be improved, incorporating new climate products into existing decision routines would be determined by two factors: (1) a radical change in system behavior including physical or management characteristics and (2) radical improvement in the reliability and friendliness of the models. They also suggested it would take at least 15 years for the process of incorporation, which have now elapsed since the original fieldwork was undertaken.

With support from the Oxford Martin Programme on Resource Stewardship (OMPORS; oxford.martin@ox.ac.uk), Rayner and a group of researchers are extending our understanding of the factors that influence climate information usability for decision making to an international set of water managers in England, South Australia, Belize, and Kenya. In addition, during fall 2015, we re-engaged with the Pacific Northwest water organizations we had talked with during our initial study. We were interested in learning whether climate and weather information and products were being integrated into decision making about water resources in the region. And, if they were, how this integration occurred. This paper discusses the findings from those...