During the seven decades of its existence, the Soviet Union was a notorious terra incognita for Western geographers and map-makers. Information improved when remote-sensing satellites began to circle the globe, but the vast lands of Russia and her former satellite states only became accessible to Western scientists after the end of the cold war.

More than 20 years on, the region continues to surprise. The demise of the Soviet Union and the socio-economic disruption left in its wake resulted in a rural exodus and substantial changes in land use. A study by researchers in Germany (F.Schierhornetal.Glob.Biogeochem.Cycleshttp://doi.org/qg8;2013) now suggests that the area of cropland abandoned since 1990 is much larger than most people in the West would have guessed (see page 342).

The information that redrew the map was not supplied by sophisticated satellite-borne sensors. Instead, the authors analysed the annual sowing statistics for crops between 1990 and 2009 — which were still hard to come by — and deduced that the area used for grain production in the former Soviet breadbaskets of western Russia, Belarus and Ukraine has decreased by more than one-quarter. This decline is substantially greater than the estimate made, mainly from satellite data, in 2009 by the UN Food and Agricultural Organization.

Why does this decrease matter? Most immediately, millions of hectares of fertile land are lying idle while a growing world population demands more food. As urbanization, desertification and increasing water scarcity constrain the extension of arable land worldwide, this unsown land might look like a precious resource. But as nature gives with one hand, it takes away with the other. These abandoned areas have become important carbon sinks. Substantial amounts of organic carbon have accumulated in their untilled soils and natural vegetation over the past two decades. If intensive agriculture were to resume, this carbon would be rapidly released into the atmosphere as carbon dioxide, contributing to global warming.

The abandoned Eastern European croplands highlight the perennial trade-off between agriculture and climate protection. And they show yet again that the debate over what to do about this needs to be better-informed by more robust data.

Russian soils and forests are a major part of the global terrestrial carbon sink. Russia’s last-minute signing of the 1997 Kyoto Protocol on Climate Change (and its possible involvement in a future international agreement) owe much to the prospect of substantial gains from the sale of carbon credits. The possibility that Russia’s natural carbon sink has increased is potentially lucrative news for a country whose leaders and scientists often keep a low profile when it comes to global warming.

Millions of hectares of fertile land are lying idle while a growing world population demands more food.

However, the true size of the Eastern European — and the global — terrestrial carbon sink remains disturbingly uncertain. Discrepancies in various estimates of its size based on changes in land use point to shortcomings in regional and global carbon accounting — a discipline on which any new international climate regime will fundamentally rely. Whether or not the revised estimates on the extent of post-Soviet land abandonment are correct, they underscore the fact that satellite observations of land-use dynamics, in Russia or elsewhere, are no credible guide to a region’s carbon balance.

A matter so central to predicting the rate of global warming deserves more attention. But existing remote-sensing technology offers relatively coarse observations of land cover and land-use change, which means that assessments are often little more than good guesses.

Space-borne sensors such as the Moderate Resolution Imaging Spectroradiometer (MODIS) instruments on NASA’s Terra and Aqua satellites do provide rough but useful land-cover classifications. But over dry regions such as the south Russian steppes, spectral analysis does not discriminate well. Moreover, the spectral fingerprint of an area tells nothing about its past and current management, such as the application of fertilizer, which affects a soil’s carbon-sequestration capacity.

In the absence of reliable satellite observations, land-use dynamics need to be continuously monitored on the ground.A NASA-funded project on land cover and land-use change in western Russia is setting the right tone by incorporating the results of field surveys with remote sensing and statistical modelling. Similar field studies would be desirable in other countries, such as Brazil, Argentina, China and India, where land use is undergoing major transitions.

Ultimately, only improved satellite observations can provide the global data sets required to understand the elusive global carbon sink. The European Space Agency’s €400-million (US$551-million) BIOMASS radar mission, selected in spring as Europe’s next Earth Explorer mission and scheduled for launch around the end of the decade, could make a real difference — if mainly in the tropics. Emerging space nations, there is a great opportunity to be seized.