Studying geoengineering is emerging as one of the most important tasks facing humanity. Climate scientists are taking the necessary first steps: defining the problem and deciding on how to conduct the research.

A pair of articles in today’s issue of Science — a Policy Forum item and a Perspectives item — contribute to these efforts by raising questions about research into solar radiation management. The policy item addresses the political issues of geoengineering research, and the perspectives item presents an unsettling picture of what it will take to accurately test geoengineering.

On the policy side, Jason J. Blackstock and Jane C. S. Long argue that stakeholders need collectively to define acceptable risk, to determine if, when and where to conduct geoengineering research, and decide how to manage the research. “Such questions require a broadly accessible, transparent, and international political process,” the authors write.

They also call for all researchers and research organizations to forswear climactic impact testing unless it’s approved by a broad international process; they also advance the notion that all solar radiation management research should be in the public domain.

The difficult and unfinished work of building an international framework for curbing carbon emissions, and the checkered history of global treaties like the ban on the militarization of space, make prospects for developing the necessary international political process for governing geoengineering research uncertain at best.

These questions could be moot, however, if the arguments set forth by Alan Robock, Martin Bunzl, Ben Kravitz and Georgiy L. Stenchikov hold up. “Geoengineering cannot be tested without full-scale implementation,” the researchers write.

The researchers have identified two problems with limited field testing of solar radiation management. First, a geoengineering deployment would require repeated injections of aerosols into the stratosphere, which would cause previously injected particles to grow larger. The larger particles would be less effective, and it would take a full-scale deployment to measure the change.

Second, getting the effects of an experiment to rise above background noise would require aerosol injections equivalent to a Mount Pinatubo eruption every four years for at least a decade — in other words, a full-scale deployment.

This suggests that, for the time being at least, geoengineering research should be confined to computer modeling and laboratory experiments.

Geoengineering has hubris written all over it. The notion that we can control a system that we don’t fully understand — especially such a large nonlinear system — is a dangerous idea.

Someday we might actually gain rudimentary control of the climate, and we might determine that geoengineering is necessary to combat global warming. But we are nowhere near that day. The problem is, carrying out solar radiation management is hardly a daunting task. It’s simply a matter of injecting sulfur particles into the stratosphere. In this way, a single country or even a large corporation could unilaterally alter the planet’s climate.

Most climate scientists, whatever their views on the eventual need for geoengineering, argue that we don’t know enough to proceed today and we need to devote additional study to geoengineering to understand how it would work.

Studying geoengineering is critical for several reasons. We need to know more before we can confidently recognize and understand the effects of our actions. If we are to launch a geoengineering effort, rightly or wrongly, we should at least make informed decisions about how to do it. And studying geoengineering could advance our understanding of unintended human effects on the climate.

Perhaps most importantly, studying geoengineering should give us a basis for deciding which is the lesser of two evils: geoengineering -- or the irreversible course of global warming.