Remote Sensing Precision Requirements For FIA Estimation

Abstract

In this study the National Land Cover Data (NLCD) available from the Multi-Resolution Land Characteristics Consortium (MRLC) is used for stratification in the estimation of forest area, timberland area, and growing-stock volume from the first year (1999) of annual FIA data collected in Indiana, Iowa, Minnesota, and Missouri. These estimates show that with improvements in the classification, ground plot location and geo-rectification methods, we should be able to meet the accuracy standard for timberland area estimates (3 percent per million acres of timberland) in most areas. To meet the accuracy standard for growing-stock volume estimates (5 percent per billion cubic feet of growing stock) under the base sampling intensity (one plot per 5,937 ac on a 5-year cycle) that FIA has implemented in the North Central Region, we must be able to create meaningful volume classes from TM imagery or find other means to improve our estimates. Improvements in classification methods are most important in areas where forest land makes up a small portion of the total land area. Simulations based on observed FIA data, an existing classification of Landsat TM data, and various assumptions are used to examine the expected accuracy of FIA estimates when a complete cycle of annual inventories (5 years) has been completed.

Citation

Hansen, Mark H. 2001. Remote Sensing Precision Requirements For FIA Estimation. In: Reams, Gregory A.; McRoberts, Ronald E.; Van Deusen, Paul C., eds. 2001. Proceedings of the second annual Forest Inventory and Analysis symposium; 2000 October 17-18; Salt Lake City, UT. Gen. Tech. Rep. SRS-47. Asheville, NC: U.S. Department of Agriculture, Forest Service, Southern Research Station. pp. 43-51