A new, private initiative should help show which gases come from where

IN 1955 a young man called David Keeling started to measure the level of carbon dioxide in the Californian air. It seemed of little practical value, but he liked designing and building the equipment—and driving back and forth along the Pacific Coast Highway to his sampling site at Big Sur was fun. Scientists with a new-found interest in the world's carbon-dioxide levels soon learned of his work and gave him a job setting up monitoring stations in Hawaii and Antarctica for the Scripps Institute of Oceanography, in La Jolla. He continued to work there for almost 50 years, devoting his life to the monitoring effort. His son, Ralph, runs the carbon-dioxide programme at Scripps to this day.

In those 50 years measuring carbon-dioxide levels has gone from being a fun problem for a postdoc to a crucial issue for the planet. But the amount of effort put into it remains surprisingly small. America's National Oceanic and Atmospheric Administration (NOAA) runs the biggest network of monitoring sites. A dozen other countries run a few here and there as well, with an expanded European effort getting under way. However, the scientists involved have been pointing out for years that it would take a very small investment, in a scientific world of satellites and supercomputers, to make such networks a lot more capable. On January 12th, such an investment was at last revealed—but not by any of the governments to which the pleas had been addressed.

A private company in Maryland, known until recently as AWS Convergence Technologies but now called Earth Networks, has announced that over the next five years it will spend $25m installing 100 state-of-the-art carbon-dioxide and methane monitors around the world. Fifty will be sited in America. According to Pieter Tans, the doyen of the field at NOAA, the country currently has 17 or 18, so that will improve things by a factor of four. In some less-well-covered places things will improve even more.

This is not pure philanthropy. Earth Networks, which already runs a large system of weather stations and lightning monitors, is looking to expand, and it believes that there should be a market for greenhouse-gas data. American states trying to show that they are doing something about their emissions might be prepared to pay for "inverse modelling" work (which uses measured gas levels and the weather patterns of recent hours and days to work out where the gas is coming from) if it were detailed enough to give results for areas as small as single states. So might countries in Europe, where the company plans to put 25 monitors. The new monitoring stations should allow such granularity.

Although carbon dioxide is the more important greenhouse gas, methane measurements will be a more practical early application to test the market for this sort of data. Land masses, and their inhabitants, emit methane without then sucking any of it back up, which makes inverse modelling easier than for carbon dioxide (which has sinks, in the form of photosynthesising plants, as well as sources).

Dr Tans probably speaks for many of the scientists involved when he says he is cautiously optimistic about the news. The caution stems in part from concerns about how the company's policy on access to its data may change as its business model becomes clearer. To begin with, those data will be free to academic researchers. But that could change.

Another, more selfish worry is that governments which buy Earth Networks' products will close down their existing research programmes. Besides the resulting unemployment, some researchers fear this would waste an opportunity to use the new data to reveal the true workings of the sources and sinks of gases.

And new data there will be—possibly a lot of them—if the business shows signs of profit. Earth Networks will presumably expand, or competitors will move in, or both. Bob Marshall, the company's boss, says he could imagine 1,000 monitoring stations around the world by 2021. Couple such capability with new satellite measurements (America's Orbiting Carbon Observatory 2 is due to launch in a few years) and better measurements of the mass of plant life in forests and other ecosystems, and the planet's greenhouse-gas credits and debits will surely be understood far better.

Such expansion has risks in itself. Long-term monitoring requires expertise that comes only with years on the job, so the quality of data from new entrants is always a bit suspect. Here, though, Earth Networks has been canny. As well as working with NOAA, it has a partnership with Scripps to ensure its precision and quality control. There's no better pedigree.