FIA Estimation in the New Millennium
Francis A. Roesch
Abstract:In the new millennium, Forest Inventory and Analysis (FIA) will deliver most of its database information directly to the users over the Internet. This assumption indicates the need for a GIS-based estimation system to support the information delivery system. Presumably, as the data set evolves, it will free FIA and the users from exclusive estimation within political boundaries. A data set of basal area measurements from a survey unit in Georgia is used to simulate one that might have been obtained had an annual inventory been conducted over a 5-year time interval. The simulated data set was used to investigate various estimators and any potential spatial correlation of basal area. The presence of spatial correlation, coupled with a desire to fulfill user needs to obtain estimates over individually defined elements of the spatial-temporal cube, forms the basis for an argument that a real-time GIS-based estimation system should be developed as the main information delivery vehicle for FIA. 1 Paper presented at the Second Annual Forest Inventory and Analysis (FIA) Symposium, Salt Lake City, UT, October 17–18, 2000. 2 Mathematical Statistician, USDA Forest Service, Southern Research Station, Asheville, NC 28804. INTRODUCTION As we approach the next millennium, it is apparent that we cannot consider how we might improve Forest Inventory and Analysis (FIA) estimation without first asking: “What are the major products of FIA likely to be?” That is, through which routes will we deliver the bulk of our inventory information? Most likely, we will deliver most of our information directly from the database over the Internet, not in the paper reports that have historically taken about 2 years to publish. To use the Internet effectively and efficiently, we must build an estimation system to adequately support the delivery of information that is more sensitive to the needs of its users. Insights into the needs of these users can come from myriad sources, but none so compelling as the reports of the two Blue Ribbon Panels, BRP I (Anonymous 1992) and BRP II (Anonymous 1998). These panels were formed specifically to provide suggestions for improving the FIA program. For instance, a concern over the potential misuse of FIA data resulted in the following statement from BRP I: “To maintain the credibility of the program, FIA, working together with experienced biometricians, must issue clear direction on the scientifically valid uses of FIA data without creating disincentives to innovation and advancement of technology” (Anonymous 1992). The best way to communicate scientifically valid uses of the information is to develop a system that can provide estimates in as many usable forms as possible. In this manner, FIA will provide scientifically defensible mechanisms from which to make estimates. FIA may still challenge inferences drawn from the estimates, but if the estimates themselves are sound, the scientific community can debate the validity of various resulting inferences. The second Blue Ribbon Panel reiterated and expanded the recommendations of the first in one specific recommendation: “Better analysis is necessary for improving customer service. More analysis of FIA data would be useful in improving and increasing customer service. While some FIA customers have the capability and inclination to analyze raw data themselves, other customers rely on outside sources to summarize and analyze the data for them” (Anonymous 1998). In addition, the second Blue Ribbon Panel charged FIA to “Produce the most current resource data possible.” The overwhelming consensus among panel members was that timeliness of resource data is of paramount importance: “Strengthening of Forest Service research and expertise in Geographic Information Systems (GIS), and collaboration with other agencies, could deliver immediate benefits. We urge the Forest Service to: “Reallocate funding within the Forest Service in order to reach the goal of timely resource data established in the first Blue Ribbon Panel report. Fully integrate GIS technology into the inventory process. Aggressively support and promote the annual inventory systems being established in the North Central and Southern FIA units. Based upon results from these efforts, establish a model for annual inventory to be adopted nationwide” (Anonymous 1998). To fully comprehend the needs of the users, we must first identify those users. FIA users include State foresters, university researchers, National Forest System employees, Forest Service researchers, military bases, other government and State agencies, forest industry, forestry consultants, and members of conservation and environmental groups. Their needs are as diverse as the groups themselves.