Expanding Forest Inventories: A Comparison of Traditional Methods of Remotely Estimating Woody Crown Cover

The Forest Service’s Forest Inventory and Analysis program (FIA) provides a census of the nation’s forests, with sample plot locations based on a vast point grid. In the Interior West, more than 60% of plot locations do not meet the FIA definition of forest land. These nonforest areas include various land uses and cover types, including urban forests and sparse pinyon-juniper woodlands. In the past, very little data was recorded for these nonforest plots. The Interior West FIA unit is initiating a remotely-sensed inventory for the entire grid, including nonforest points. Variables will include land use, canopy cover, and past canopy cover at disturbed sites. To populate these variables, we compared the precision and efficiency of three methods of estimating canopy cover: dot-grid samples, ocular estimation, and image segmentation. We also assessed the accuracy of these methods by comparing our estimates against field-recorded crown cover data.