The Arctic has been one of the fastest-warming regions on the globe for the past few decades, as evidenced by the recent dramatic decline in summer sea ice cover. The consequences of this rapid warming — for instance, for the vast quantities of carbon locked away in the frozen soils and icy submarine sediments that permeate the region — is a key climate concern. Data from the Arctic, like those presented on page 64 of this issue, can therefore be of immense value to the climate science community. However, making measurements in such an inhospitable environment, particularly in the icy and increasingly storm-prone waters of the Arctic Ocean, is risky. In the case of the research presented in this issue, the rescue boat that set out to assist the researchers when they hit stormy waters sank, taking 11 of the crew with it.

On 24 August 2010, the researchers headed out to a spot in the southern Laptev Sea, to the east of the Lena River delta, to verify observations from the previous summer that storms can flush methane out of the sea water. But when a storm hit on 26 August, their ship got into difficulty and they sounded the alarm. A 14-strong rescue team set out to help, but their tug boat capsized en route. A nearby tanker managed to save three of the rescuers from the icy-cold waters. Tragically, the remaining 11 rescuers died. The researchers survived unscathed.

The data taken by the surviving research team, collected over a 10-year period and presented in this issue, show that submarine sediments on the Arctic shelf are releasing significant quantities of the potent greenhouse gas methane to the overlying ocean, and that storms and bubbles mediate the flux of much of this methane to the overlying atmosphere.

The death toll associated with this research expedition is particularly high. However, Earth science fieldwork often comes with some risk attached. For instance, tree snakes and saltwater crocodiles prey on the minds of researchers in the hot humid mangroves of the tropics (Nature Geosci. 4, 344; 2011), and the threat of piracy looms large for researchers in the South China Sea (Nature Geosci. 1, 875; 2008). Such risks make it all the more important that the hard-won data from these projects are analysed, shared and reanalysed to their full potential.