Are forecasts still for wimps?

Journal article

More than 15 years ago, Lach, Rayner, and Ingram interviewed water resource managers around the US to determine if and how they were using probabilistic climate forecasts in decision making. At that time, we heard that the forecasts were too unreliable and difficult to integrate into existing decision routines. Instead, we saw what we called a “domestication” of the forecasts that postponed use until sometime in the future – say 15 years or so – after more research was done and forecasts were refined so they were more usable. We recently went back to talk with people in the same organizations in the Pacific Northwest to see what that 15 years of domestication yielded. We found this time that technological advancements have made real-time monitoring and short-term weather forecasts (three-five days) very much a factor in water decision making, although the forecasts remain a “black box” to most users with little understanding of how they are produced. Seasonal forecasts were often referenced as context or background to planning processes while the large global change models that have been scaled down to smaller regions were used for infrastructure planning by technically trained employees. We also heard about processes of co-production, whereby users and producers of forecasts work together to select model parameters, define assumptions, and interpret results.

Authors: Lach, D; Rayner, S

• Publication status: Published
• Peer review status: Peer reviewed
• Publisher: University of Arizona
• Journal: Journal of the Southwest
• Volume: 59
• Issue: 1-2
• Pages: 245-263
• Publication date: 2017-07-01
• Acceptance date: 2016-11-22
• DOI: 10.1353/jsw.2017.0013