Validation of a small-area method for estimating the spatial distribution of the degree of tree damage

This article presents a validation of a method, proposed in Podlaski (2003), for estimating the degree of tree damage. This method belongs to a group of small-area methods. To estimate the degree of tree damage in individual P₃ sub-blocks of the SINUS system of information on the natural environment, a classification based on the degree of defoliation was used. Spatial distribution of the degree of tree damage was estimated through unrestricted simple random sampling. The maximum total estimation error for proportions of fir and beech with a damage degree of zero, and also with second and third degrees jointly, reached a maximum of 35.6% for fir, and 33.8% for beech. P₃ sub-blocks characterised by the highest proportion of the healthiest trees (at least 15% being undamaged trees) occurred in patches in the case of fir, and in larger, compact areas in the case of beech. Because spatial diversification in the health of fir, and to a smaller degree in the case of beech, occurs at small scales, small-area methods for estimating tree condition are necessary. An analogous scheme, combining small-area with large-area methods (utilising sub-blocks of the network of large-area systems in a small-area inventory, and establishing a common database), could be used on a wider scale; for example, by supplementing the program of estimating tree health carried out by the United Nations Economic Commission for Europe (Lorenz et al. 2002) with local inventories.