Traditional wisdom on coping with drought (Mortimore 1998), complemented by cutting-edge science and information technology (for example, drought forecasting and climate change scenarios), holds great potential for sustainable resource management. If we can have better information about the near future, we can plan better how to deal with it. Although drought alone cannot be held responsible for causing food insecurity (Sen 1981), some desert regions face food insecurity and increases in excess mortality during prolonged drought periods. USAID has initiated a famine early warning system project (FEWS NET: http://www.fews.net/), which provides drought early warning and vulnerability information for drought-menaced African countries, both in semi-arid drylands and in deserts.

The objectives of interventions triggered by the early-warning information range from saving lives in response to immediate emergencies (for example, emergency food programs and livestock health interventions), to saving both lives and livelihoods by reducing exposure to risk and developing diverse opportunities to generate income (such as, crafts and other off-farm employment) other than from agriculture.

While the activities of FEWS NET tend to focus on the most vulnerable human populations, an example from a more robust setting is the Rangeview project (http://rangeview.arizona. edu) that sprang from an initiative to make geospatial technology accessible to a range of users. Launched in 2000 to provide rangeland managers in the American Southwest with satellitederived information about the status and trends of vegetation greenness, it evolved into a decisionsupport tool and has meanwhile been extended to cover the entire United States. Similar initiatives, making use of the Internet as an inexpensive means of information exchange, could be beneficial to rangeland management in desert regions around the world.

This type of tool helps natural resource managers to understand what is happening now, and compare it with what has been happening over time and come to an appreciation of where we are at any point in time. As our ability to understand and model the global climate system improves, we are able to develop increasingly useful seasonal weather forecasts for large regions and long-term scenarios. These can be used to plan for adapting to climate variability and change, which are expected to be increasingly inevitable during the next decades (Dessai and others 2005).

Technical knowledge and reliable forecasts alone, however, are insufficient; they need to be implemented to the benefit of people under a given set of circumstances. Climate change adaptation planning therefore must include the identification of vulnerable population groups and the exploration of effective and affordable livelihood strategies during times of climatic stress. Perhaps most importantly, there need to be systems in place that have the political will and institutional capacity to act on the most likely scenarios, or at least to accommodate some outcomes.