We investigated the seasonal variability of the relationships between land surface temperature (LST) and land use/land cover (LULC) variables, and how the spatial and thematic resolutions of LULC variables affect these relationships. We derived LST data from Landsat-7 Enhanced Thematic Mapper (ETM?) images acquired from four different seasons. We used three LULC datasets: (1) 0.6 m resolution land cover data; (2) 30 m resolution land cover data (NLCD 2001); and (3) 30 m resolution Normalized Difference Vegetation Index data derived from the same ETM? images (though from different bands) used for LST calculation. We developed ten models to evaluate effects of spatial and thematic resolution of LULC data on the observed relationships between LST and LULC variables for each season. We found that the directions of the effects of LULC variables on predicting LST were consistent across seasons, but the magnitude of effects, varied by season, providing the strongest predictive capacity during summer and the weakest during winter. Percent of imperviousness was the best predictor on LST with relatively consistent explanatory power across seasons, which alone explained approximately 50 % of the total variation in LST in winter, and up to 77.9 % for summer. Vegetation related variables, particularly tree canopy, were good predictor of LST during summer and fall. Vegetation, particularly tree canopy, can significantly reduce LST. The spatial resolution of LULC data appeared not to substantially affect relationships between LST and LULC variables. In contrast, increasing thematic resolution generally enhanced the explanatory power of LULC on LST, but not to a substantial degree.
This study evaluates and compares the performance of four machine learning classifiers-support vector machine (SVM), normal Bayes (NB), classification and regression tree (CART) and K nearest neighbor (KNN)-to classify very high resolution images, using an object-based classification procedure. In particular, we investigated how tuning parameters affect the classification accuracy with different training sample sizes. We found that: (1) SVM and NB were superior to CART and KNN, and both could achieve high classification accuracy (>90%); (2) the setting of tuning parameters greatly affected classification accuracy, particularly for the most commonly-used SVM classifier; the optimal values of tuning parameters might vary slightly with the size of training samples; (3) the size of training sample also greatly affected the classification accuracy, when the size of training sample was less than 125. Increasing the size of training samples generally led to the increase of classification accuracies for all four classifiers. In addition, NB and KNN were more sensitive to the sample sizes. This research provides insights into the selection of classifiers and the size of training samples. It also highlights the importance of the appropriate setting of tuning parameters for different machine learning classifiers and provides useful information for optimizing these parameters.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.