VARIACIÓN ESPACIAL DEL DIÁMETRO COMO RESPUESTA A DIFERENTES INTENSIDADES DE MUESTREO EN UNA CUENCA FORESTAL

Abstract

Management of forest resources implies an adequate knowledge of their characteristics and spatial distribution, although it is common to find problems when implementing this process. This paper compares two interpolation techniques: 1) deterministic (inverse distance weighted) and 2) geostatistical (ordinary kriging). Both techniques were used to determine the spatial variation of the mean diameter in trees from the same forest region. Data were gathered throughout a forest inventory carried out in El Carrizal watershed in the Sierra de Tapalpa, Jalisco. 79 sites were sampled using a systematic sampling design and sub-samples series were randomly developed from this sample, to generate groups with different percentage sampling intensities (87, 75, 62, 49, 36 and 24%), subsamples series were randomly defined. · Based on the geostatistical analysis, the corresponding experimental variograms were generated for each sampling intensity. These variograms were fitted to the best model, which were used to develop the corresponding interpolations. Deterministic interpolations were made directly. The lower mean square error was the selection criterion to choose between the two interpolation techniques. After the selection of the interpolation technique and sampling intensity, the corresponding continua! surfaces were generated to show the spatial diameter response among trees. lt is concluded that both techniques have similar accuracy for the determination of spatial variation.