A method for using Landsat time series for monitoring young plantations in boreal forests

A time series of Landsat Enhanced Thematic Mapper Plus (ETM+) imagery was used to investigate its potential contribution to characterization of young forest plantations in northern Sweden (latitude 65 degrees N, longitude 16 degrees E). The time series image data were first bandwise relatively normalized, using image statistics from forest areas. It was shown that the difference in reflectance development between different types of forest plantations could be characterized by fitting a linear regression model through the bandwise spectral mean values for each stand, starting 5 years after the final felling. The coefficients of this regression model can be used to improve statistically based characterizations of young forest, compared to using data from the most recent image only. This result is useful for screening stands that might be in need of precommercial thinning (PCT), as well as for improving forecasts of expected forest development for whole forest-dominated landscapes consisting of several estates.