Resumen

ESCANDON CALDERON, Jorge et al. Evaluación de dos métodos para la estimación de biomasa arbórea a través de datos LANDSAT TM en Jusnajab La Laguna, Chiapas, México: estudio de caso. Invest. Geog [online]. 1999, n.40, pp.71-84. ISSN 2448-7279.

Two approaches to estimate arboreal biomass with remote sensing (LANDSAT TM) are evaluated In the first approach a multi-spectral supervised classification with six bands was applied The classification of the vegetation types is based on biomass composition of the dominant tree species and canopy height. Eight vegetation types could be distinguished. According to this approach the total tree biomass amounted to 1 073 x 10³ t (902 to 1 220 x 10³t). In the second approach a Normal Differentiated Vegetation Index (NDVI) of the band combinations TM4/TM3, TM4/TM5 and TM4/TM7 was used, A regression equation was developed to relate arboreal biomass with NDVI. Using these equations, the total biomass was estimated at 1 164 x 10³ T (490 to 2 409 x 10³T) for TM4/TM3; at 515 x 10³t (331 to 757 x 10³t) for TM4/TM5 and 726 x 10³t (398 to 1 210 x 10³t) for TM4/TM7. The average biomass estimation of the NDVI using TM4rTM3 is similar to the estimation using the classification approach, but the 95% confidence Interval is wider. Meanwhile, the biomass estimation of the NDVI using TM4/TMI5 and TM4/TM7 was lower than the biomass estimation from the classification approach, but both showed a narrow 95% confidence interval. The results of this study Indicate that it is possible to estimate within a reasonable confidence interval the tree biomass of pine-oak forest using an ordination approach with NDVI As such, remote sensing could be used to estimate temporal and spatial changes in aboveground biomass.

Palabras llave : Tree biomass; remote sensing; supervised classification; Normal Differentiated Vegetation Index; regression models.