Greenhouse Gas Emissions from Hydroelectric Dams: Reply tO Rosa Et al.

Abstract

Ever since my first estimate indicating high emissions of greenhouse gases from Amazonian dams (Fearnside, 1995), Luis Pinguelli Rosa and coworkers have effectively made a career of trying to prove me wrong. However, the longer this debate goes on and the more information that becomes available, the greater the impacts are found to be. The latest attack (Rosa et al., 2005) serves to illustrate a series of misconceptions regarding the science and brings out some of the political context that surrounds the subject in Brazil. First, a variety of claims in the Rosa et al. (2005) letter are technically incorrect, and would mislead readers not familiar with the details of the previous rounds of this debate. Rosa et al. (2005) claim that I assume that the CH4 concentration is “uniform in the reservoir and constant for many years.” Neither assumption is made. The calculation in question (for Tucurui) only applies this concentration to the depth of the turbines – not as a concentration that is uniform throughout the reservoir. In fact, it is conservative in that the value is based on a measurement at 30 m depth, and is assumed to be the same at the 34.6 m depth at the turbine intakes, even though methane concentrations are well known to increase steadily with depth anywhere below the thermocline. Neither is it assumed to be constant for many years, nor to be unduly “extended through extrapolation.” The “extrapolation” referred to is only for 1 year, from 1989 (the year of Tundisi’s measurement) to 1990 (not 1991). The year 1990 was used for the estimate because this is the standard year for the initial national inventories of greenhouse gases under the United Nations Framework Convention on Climate Change. The proportions of water passed through the spillway versus the turbines are from 1991, the year closest to 1990 for which data were available. The graph (Graph 1) reproduced from Galy-Lacaux et al. (1999) shows the annual cycle of methane concentration variation at the Petit-Saut reservoir in French Guiana, together with the expected general decline over time as the reservoir ages. The March 1989 measurement from Tucurui (rainy season at age 4 years) corresponds to the fourth trough on the saw-tooth like graph. The decrease from year 4 to year 5 is approximately 0.5 mg/l (about 12.5%) – hardly enough to explain the supposed 19-fold exaggeration in methane emissions.