Rangeland monitoring using remote sensing

Monitoring vast landscapes has, from the beginning of rangeland management, depended on people's judgements. This is no longer tenable, but a more effective method has yet to be devised. The problem is how to do an economical inventory that will detect ecologically important change over extensive land areas with acceptable error rates. The error risk is a function of adequate sample numbers and distribution for each indicator monitored. Of all of the indicators identified for monitoring, ground cover and its inverse, bare ground, may be the most discussed. Ground-cover measurements address soil stability and watershed function which are first-priority ecological concerns; are well adapted to remote sensing frameworks thus allowing extensive, unbiased, economical sampling; and, the measurements, especially when done by computer image analysis, have the potential to reduce or avoid the human-judgement factor. Data collection through remote sensing appears the most logical approach to acquiring appropriately distributed information over large areas in short time periods and on random sites far removed from easy ground access. The value of satellite and high-altitude sensors for landscape-level evaluations, such as plant community distribution, is well established but these tools are inadequate for inventory and measurement of details needed for valid conclusions about range condition. New advances in low-altitude remote sensing may give us the ability to accurately measure bare ground and perhaps other indicators. Combining information from high and low-altitude sensors appears to offer an optimal path for developing a practical system for cost-effective, data-based, rangeland monitoring and management.