Multiple approaches to measure the impact of research exist - the most common ones include citation metrics like the in-famous Journal Impact Factor or the relatively new Hirsch index, and the volume of research grants earned.

Several such measures have become important determinants of scientific careers, but a certain type of impact measures always tend to be forgotten: environmental measures. Now, in a recent contribution to the journal "Trends in Genetics" (not Open Access), biochemist Hervé Philippe takes a semi-quantitative approach to these matters. Then, he points out that scientists (even those that do not work with environmentally hazardous agents), by doing their job - research and disseminating its results to the scientific community and the public - use up natural resources (and produce waste) at a level well above the average citizen of our planet (this level has come to be known as the ecological footprint, although Philippe does not explicitly mention the concept).

This is an important article because it highlights that sustainability is not just a topic for politicians and perhaps some specialized group of scientists - no, it is a matter of concern for a biochemist, too, and indeed for anybody but in particular for scientists who are trained in the analysis of relatively complex structures. The paper is also unusual for a science article in that it contains public advice to policy makers - they should take the limitedness of natural resources into account and leave the eternal quest for economic growth, even replace it (at least temporarily) with a goal that he calls "degrowth" and which I would rephrase as "leading to a steady-state economy".

At a time when even a multinational oil company like Elf Total admits that fossil fuels are finite (English version here), it would thus seem very appropriate if scientists, science organizations and science funders would start to devote some of their current resources to investigating in detail what impact they have on the environment, and how this could be balanced, in the long run, with the actual research impact. The Ecological Footprint mentioned above was originally designed to determine the resource use of the global human society but it has since been adapted to lower-level scales, e.g. cities and industrial products, and so it should not be too difficult to get reasonably correct estimates for scientific research projects, too (especially considering that scientific research tends to be well-documented). If you know of any initiative in this direction, please post it here.