Forest canopy change information over two-, four-, and six-year periods was extracted digitally from multitemporal Landsat Thematic Mapper (TM) imagery over a 415 km(2) area in north central Minnesota. This paper discusses the complementary character of the pixel-based change information with respect to stand-based data that traditionally have provided the foundation for management prescriptions. Change maps generated from TM data, and change overlays created from traditional stand maps updated with aerial photography, were site-specifically cross-verified using contingency matrices and geographical information system (GIS) modeling, supported by intensive held verification. It was demonstrated clearly that a majority of the so-called digital change classification inconsistencies were not errors, but embodied a powerful source of sub-stand information with the potential to significantly impact the sustainability of the resource management. While the pixel approach does not directly encompass the stand concept (stand representing the actual management unit), qualitative and quantitative data on change at the stand level could be readily extracted from the TM imagery, provided pre-change stand information was available. For example, for the six-year monitoring cycle, 714 out of 759 stands identified as ''changed'' (canopy depletion or increment) in the traditional stand maps were also digitally classified as such, whereby change was detected in at least 75% of the pixels which made up the stands. New information being made available to the resource manager ranged from location and areal extent of sub-stand changes, e.g. trespasses, flooding, insect infestations, and species-specific dieback, to the detection of local failures in tree establishment and/or population development.
