Quantifying Uncertainty in Ecosytem Studies (QUEST): A Research Coordination Network.

Uncertainty analysis is essential to scientific progress; it is needed for determining the significance of observed differences, for detecting trends over time and making predictions with known confidence, and for guiding research investments by identifying which components contribute the most to the overall uncertainty. In forested ecosystems, however, it can be difficult to identify, evaluate, and propagate the many sources of uncertainty, and thus uncertainties in estimates of ecosystem pools and fluxes have rarely been evaluated.  The difficulties are even greater belowground, where direct observations are often not possible and many measurements are destructive.

In the case of soil measurements, a power analysis can be used to identify the magnitude of a difference that would be statistically significant given the intensity of sampling and the spatial variation in the sample.  It is useful to report the magnitude of the difference that could have been detected, rather than to simply report that a change was not significant.  A power analysis of forest floor studies in North America found that few designs were able to detect changes smaller than 20%; some were not sufficient to detect a doubling of forest floor carbon.  Sampling designs can be improved to allow detection of smaller changes while minimizing sampling and analytical effort.

Quantifying Uncertainty in Ecosystem Studies (QUEST) is a Research Coordination Network recently funded by the National Science Foundation of the USA.  The mission of QUEST is to support activities that advance the application of uncertainty analysis in five topic areas important to ecosystem studies: atmospheric deposition, stream water export, biomass, soils, and ecosystem budgets.  QUEST has a website at http://www.quantifyinguncertainty.org, which includes sample code, links to papers, and news announcements. We will be developing discussion boards, wikis, and tutorials.  Please join us if you have questions or suggestions for uncertainty analysis: quantifyinguncertainty@gmail.com; @QUEST_RCN