In this Section:

	Background
	Project Overview
	Project Components:
	Seasonal Climate Outlooks
	Historical Water Supply Outlooks
	Ensemble Water Supply Outlooks
	Decision-Support Tools
	Status of Project
	References
	Related Links
	Contacts

Research and Products:
Forecast Evaluation

Background

Forecasts are a natural intersection between science and decision-making.

Virtually every resource management decision requires a forecast, whether the forecast consists of an implicit assumption that the future will be much like the recent past or is the output generated by a series of complex dynamical numerical models. With every decision, individuals must balance their potential risks and rewards, while considering their uncertainty about the future.

From a science perspective, forecasts are where "the rubber meets the road." With every prediction, forecasters and scientists have a chance to test the ability of their scientific concepts and computer models to handle conditions not contained in historic data. Sometimes those conditions can be well beyond historic data, such as during times of extreme drought or large floods.

Every resource management decision offers researchers and forecasters an opportunity to assess whether their science can offer useful information for real-world applications. Every forecast offers decision makers an opportunity for assessing whether science can reduce their uncertainty about future conditions, risks, and rewards.

Each time a forecast is issued or used, both decision makers and scientists must confront a series of issues.

• Is the forecast relevant to the decision maker? Is it for the right variable (e.g., precipitation vs. sea surface temperatures)? Is it for the right region and period of interest? Is it issued with enough lead time to be useful?
• Is the forecast accessible to the decision maker? Can it be obtained easily? Can it be understood easily, correctly, and reliably? Can the forecast be readily related to past historical conditions and recent observations?
• Does the forecast actually reduce uncertainty? Does the forecast reduce the potential for being surprised by unusual conditions? Does the forecast raise alarms about situations that ultimately don’t happen? Does the forecast accurately indicate the confidence that forecasters have in their prediction?

Sometimes decision makers find that there simply aren’t any forecasts available that address their situation. When it comes to forecasts related to weather and climate (including forecasts of streamflows, water supplies, and fire risk), however, there are many different forecast products that may be appropriate, leading to problems in deciding which forecasts are the right ones to use.