Représentativité locale des placettes d'inventaire en vue de l'estimation de variables dendrométriques de peuplement

Abstract

(underestimated). When the stands are older and when the simulated plots are larger (0.05-0.10 ha), our results suggest that the larger plot size and the regularization of the structure of the stand induced by silvicultural practices diminish both the variance and the bias of the random error. These results indicate that before using a growth model in order to project regional inventory data it is necessary to check the compatibility between the data used to build the model and the data gathered in an operational survey. maritime pine / sampling method / inventory / modelling / census

Forecasting wood resources on the basis of national forest inventory data. Application to Pinus pinaster Ait. in southwestern France

“…Three explanations are possible for this discrepancy: (i) sampling errors: the original height growth model had been built on the basis of the stem analysis of dominant trees. In contrast, the dominant height was estimated by NFI on small sample plots, which might result in a similar underestimation problem [35] yet described for c 0 [40]. However, this bias, if any, is likely to be small; (ii) measurement errors: NFI measures height increments on standing trees by counting whorls and shoots and measuring (from the ground) the length of the shoots: this method is not very precise for tall trees, and an overestimation is possible.…”

“…Since Salas et al [40] had demonstrated that NFI sampling scheme resulted in biased estimates of c 0 (underestimation due to plot size), N and G (for young stands only because of census threshold), that these biases could be corrected, the variables N, G and c 0 were corrected before testing the original model of the basal area increment of dominant trees (Eq. (2-2.2)).…”

“…Because our objective was to predict all the timber produced in the region and in subsequent years, it was necessary to estimate the total density and basal area of the stands. For this reason in an earlier study, 37 new large temporary plots were employed to estimate accurately these stand characteristics [40].…”

“…700 m 2 . Salas et al [40] have shown: (i) that the design and plot size used by NFI resulted in a high coefficient of variation (CV) of the estimates of stand features such as density (N) and basal area (G); (ii) that average and dominant circumference (c g and c 0 ) had a lower coefficient of variation, but that c 0 was biased with an average underestimation of ca. -2 cm.…”

“…Salas et al [40] indeed showed that it is necessary to correct N, G and c 0 estimated from NFI plots because of their small size and of the minimum tree census threshold (gbh 24.5 cm). Total N and G were thus estimated using the following equation:…”