| Intensifying public interest in forests and the development of new monitoring technologies have induced major changes in the methods for updating forest monitoring systems. This thesis reviews the available methods for projecting and propagating forest plot and stand information. Since developing a methodology to produce forest maps based in forest inventory data would be very useful for many forest management and planning purposes, emphasis is focused on propagating forest sample information through the landscape for producing locally useful maps of forest variables such as cover type, stand density and timber volume. Considerable effort in combining forest monitoring systems information, remote sensing, and geographic information systems disciplines has been made in Nordic Countries. One of the applications developed from this work is the k-nearest neighbors (kNN) method for forest estimation and mapping. In the kNN method, the information contained in field sample units is propagated to the entire population using a similarity function. The result is a form of post-stratification. We examine the use of the kNN method for producing wall-to-wall basal area, volume and cover type maps in the context of the USDA Forest Service's Forest Inventory and Analysis (FIA) monitoring system. Several variations within the kNN method were tested, including: distance metric, weighting function, and number of neighbors. Also, specific procedures to incorporate ancillary information and image enhancement techniques were tested. The final products from this thesis are basal area, volume and cover type maps based on FIA and other inventory data. Additionally, each of these maps were characterized by estimations of their overall accuracy using cross-validation and other bootstrap based techniques. |
