Geography Reference
In-Depth Information
burden, with hundreds of data points for more than 1,000 households, we were forced to
reduce the scope of this indicator—and others—to something more manageable.
Data Operationalization and Binning
Once a set of impacts has been agreed upon, how should we go about collecting the data
to measure them? Some impacts are quite readily captured in a model, whereas others are
much more difficult to examine systematically. For example, the hydrologists on our team,
tasked with measuring “flow regime” (BP5), a biophysical system factor, could simply go
out and collect water samples from upstream and downstream of a given dam project or
even use a predictive sediment model that uses flow rate and other data to estimate sedi-
ment load. But the data-collection process can be much more difficult with complex and
nuanced impacts such as “cultural change” (SE2). Once the data necessary for measur-
ing each impact was collected, we were tasked with converting it to a scale, a process we
came to call “binning.” We settled on a four-point scale: 0 = no impact, 1 = small impact,
2 = medium impact, 3 = large impact. But hard choices had to be made at every step of
the process. For example, if one determines that a given impact is “large,” the obvious
question is, “Large relative to what?” When our research group convened a symposium on
the Columbia River and was touring the Bonneville Dam, the docent from the U.S. Army
Corps of Engineers told us about the river's volume, the technology behind the dam's tur-
bines, and other key facts. About one year later, we convened a similar workshop in Maine,
where we toured a small flow-control dam on the Kennebec River. We all had a good laugh
when one of the hydrologists in the group informed us that the flow volume of the Kenne-
bec was about a thousand times smaller than that of the Columbia. Such field experiences
underscored the importance of thinking about scale in the modeling effort and encouraged
us to allow enough flexibility in the model to capture different hydropower-development
scenarios, from the mammoth Three Gorges Project to more modest projects. As a result,
the scientists and policy makers who use the model will need to agree upon the scale of
their modeling effort as a precondition to their work.
Measuring Change
One of the problems of understanding the effects of dams is that such effects often unfold
over an extremely long time horizon. The life cycle of a dam—from design to construction,
operation, and finally decommissioning—typically lasts for many decades. Our research
team tried fora very long time to analyze and measure diachronic “change” within the vari-
ous biophysical, socioeconomic, and geopolitical systems affected by dams, but we came
to realize that doing so is a tall order. For example, if we set out to measure changes in so-
