Sampling Techniques for Forest Inventories

Using remotely sensed data to construct and assess forest attribute maps and related spatial products

“…Current approaches to model-based inference originated in the context of survey sampling and can be attributed to Mátern (1960), Brewer (1963) and Royall (1970). Given the origins of modelbased inference in survey sampling, it is not surprising that forestry applications have often been in the context of forest inventory (Rennolls, 1982;Gregoire, 1998;Kangas & Maltamo, 2006;Mandallaz, 2008;McRoberts, 2006McRoberts, , 2010aStåhl et al, in press). An important aspect of the modelbased approach is that when the model is correctly specified, the estimator is unbiased (Lohr, 1999); however, when the model is misspecified, the adverse effects on inference may be substantial (Royall & Herson, 1973;Hansen et al, 1983).…”

Design-based methodological advances to support national forest inventories: a review of recent proposals

“…The procedure of neglecting non-forest points in the second phase is adopted in most NFIs but it is not sufficiently focused in most familiar textbooks (de Vries 1986, Schreuder et al 1993, Gregoire & Valentine 2008, Mandallaz 2008. Even if the procedure is suitable from both theoretical and practical point of view because avoids the waste of sampling effort outside forest, it may provide downward bias in both extent and total estimation.…”

“…Accordingly, in the first phase a unique spatial design for selecting points (from which plots are centred) suffices for both extent and total estimation. As pointed out by Gregoire & Valentine (2008) and Mandallaz (2008), in this framework estimation reduces to a twodimensional Monte Carlo integration. Thus the key problem is how to effectively select the first-phase points to better perform integration.…”

“…Keywords: Two-phase Strategies, Aerial Information, Non-forest Resources, Missing Data, LiDAR, Calibration Weighting First phase Gregoire & Valentine (2008) and Mandallaz (2008) provide excellent introductory treatments on the issue of sampling discrete objects (forest trees in the present case) scattered over a region by means of plots or transects (plots in the present case) to estimate a total. The authors emphasize that these sampling strategies may be conveniently re-formulated as spatial strategies for sampling the continuous populations of points constituting the study area and estimating an integral within.…”

“…Usually, NFIs are performed using an intensive first phase which is carried out on satellite imagery or aerial photos and a second phase which is carried out by ground inspections, possibly adopting the information acquired in the first phase as auxiliary information. A novel methodology is adopted based on Monte Carlo integration methods (Gregoire & Valentine 2008, Mandallaz 2008, which leads to a very general estimation strategy valid for any second-phase sampling scheme. In section 3, some recent proposal are considered in which the aerial information acquired in the first NFI phase is exploited to investigate non forest resources, such as woodlots, tree-rows and isolated trees outside the forest.…”