| The relationship between Landsat Multispectral Scanner (MSS) and lodgepole pine biophysical data were investigated for portions of Grand Teton National Park, Wyoming. Correlation and regression analyses were employed to highlight the significant relationships between forest biophysical factors and satellite spectral response. Tree age, basal area, and height were highly correlated with the green and red channels of the Multispectral Scanner and the Normalized Difference Vegetation Index (NDVI). Age and height were strongly correlated with the infrared MSS channels. Tree density was not highly correlated with any of the satellite bands or NDVI. Regression equations for the estimation of age, basal area, and height accounted for at least 61% of variance in these factors using satellite data as independent variables. A combination of several biophysical factors was found to influence spectral reflectance from lodgepole pine forests.; Landsat Multispectral Scanner and field data were also used to model basal area in Grand Teton National Park using three techniques: regression of field and satellite data, kriging of field data, and cokriging of the field and satellite data. The vegetation index NDVI was the satellite derived factor used in the modeling process. For the basal area variable, the cokriging model had the highest modeling efficiencies, followed by the kriging model, and then the quadratic regression model. Modeling efficiency ranged from 0.34 to 0.92 for the estimation techniques implemented. Plots of observed vs. estimated values showed that as the complexity of the estimation process increased (from quadratic regression to kriging to cokriging), the ability to duplicate the observed values also increased. |
