You are here:
EFFECTS OF LANDSCAPE CHARACTERISTICS ON LAND-COVER CLASS ACCURACY
Citation:
Smith, J H., S. V. Stehman, J D. Wickham, AND L. Yang. EFFECTS OF LANDSCAPE CHARACTERISTICS ON LAND-COVER CLASS ACCURACY. REMOTE SENSING OF ENVIRONMENT 84(3):342-349, (2003).
Impact/Purpose:
Our research objectives are to: (a) develop new methods using satellite remote sensor data for the rapid characterization of LC condition and change at regional to national scales; (b) evaluate the utility of the new NASA-EOS MODIS (Moderate Resolution Imaging Spectrometer) leaf area index (LAI) measurements for regional scale application with landscape process models (e.g., biogenic emissions and atmospheric deposition); (c) provide remote sensor derived measurement data to advance the development of the next generation of distributed landscape process-based models to provide a predictive modeling capability for important ecosystem processes (e.g., nutrients, sedimentation, pathogens, etc.); and (d) integrate in situ monitoring measurement networks with UAV and satellite based remote sensor data to provide a continuous environmental monitoring capability.
Description:
Utilizing land-cover data gathered as part of the National Land-Cover Data (NLCD) set accuracy assessment, several logistic regression models were formulated to analyze the effects of patch size and land-cover heterogeneity on classification accuracy. Specific land-cover class relationships were established at both the level I and 11 classification schemes. Results indicate that the general pattern of the relationship between class accuracy and landscape structure was as expected: as heterogeneity increases, accuracy decreases, while as patch size increases, accuracy also increases. However, the magnitude of the effects of the variables varied by land-cover class resulting in some classes having patch size more important, while others had land-cover heterogeneity more important. Comparison of the results of the two classification schemes revealed that the heterogeneity odds ratios for the level I classes were smaller than those for the level 11 classes. In addition, their patch size odds ratios were greater than those at level 11. These results indicate that the level I classes are more sensitive to changes in the variables than the level 11 classes. Interaction between the two landscape variables was found to be significant for only a single class, indicting that the influence of heterogeneity was not impacted by the sample being in a small, or large patch. The information provided by the landscape variables remained significant even in the presence of regional dummy variables indicating that their impact was not due to regional differences in the mapping, assessment processes, or landscape characteristics Results of these analyses provides land-cover class specific quantitative information useful for extracting classification error information for a location on the map based on that location's land- cover class, patch size and land-cover heterogeneity.