Sir

On-the-ground monitoring is undervalued, as Euan Nisbet points out in his Commentary 'Cinderella science' (Nature 450, 789–790; 2007). Long-term monitoring data provide the critical foundation we need in order to develop an understanding of the processes at work. This, in turn, enables modelling studies and rationally based management decisions. That is why having 'ground truth' — information gathered on the spot — to combine with satellite observations and modelling is even more critical today.

Through the 1980s and 1990s, we saw a deterioration in many key long-term monitoring programmes, the best example being the reduction in the number of US Geological Survey gauging stations. In recent years there has been a push to increase such networks as the US Natural Resources Conservation Service's snowpack telemetry gauging stations. We can see how hard it is to construct long-term records on sea-level rise, because tide gauge records have seldom been continuous.

Monitoring is as important, I believe, as expanding the horizons of research. The data sets gained are key to the expansion of knowledge as well. I monitor the mass balance of more glaciers than any other programme in North America and have done so for 25 years without any federal money. This was crucial from the start — I was correctly informed that the federal government was not interested in funding long-term monitoring. As a result, I sought alternative funds that were sustainable but also enforced the use of cost-efficient techniques. Both have been key to maintaining the extensive annual fieldwork programme that is required to measure and report glacier mass balance.