A simulation of image-assisted forest monitoring for national inventories.

摘要: The efficiency of national forest monitoring efforts can be increased by the judicious incorporation ancillary data. For instance, a fixed number ground plots might used to inform larger set annual estimates observing smaller proportion each year while augmenting estimate with data in order reduce overall costs maintaining desired level accuracy. Differencing successive geo-rectified remotely sensed images conceivably provide change at scale and accuracy conducive improvement temporally relevant attribute estimates. Naturally, degree is highly dependent on relationships between spatial-temporal scales plot observations estimation. In this paper, observation for source are explore value three different levels information available from image-change Four populations simulated sampled under four sampling error structures. results show that image (ICE) significantly bias harvest mortality improved estimation sometimes, but not always, contribute better standing volume.