Miami-Dade County Land Cover Map (2021)

The land cover map for the urban development area of Miami-Dade County was produced from 2-m resolution multi-spectral World View 2 data. It includes nine land-cover classes: Tree Shrub, Grass, Bare Ground, Wetland, Water, Buildings, Impervious Areas, Road Railroad, and Cropland. Data Source: WorldView-2 (eight band spectral resolution, 2m spatial resolution) data acquired on 11/28/2019 and 3/27/2020. Pre-Processing: All images were geometrically corrected using the rigorous orthorectification module in ENVI. All images were atmospherically correct with the Fast Line-of-sight Atmospheric Analysis of Hypercubes (FLAASH) procedure in ENVI. Land cover Detection: Multi-spectral reflectance values were used in the classification of eight land cover classes. Concrete, Asphalt, Truck Trailers on Asphalt, Bare Ground, Grass, Tree Shrub, Wetland and Water Training samples for each of the eight classes were digitized from aerial photography and interpretation of WV spectral signatures. The initial land-cover detection was performed using the random forest classification algorithm (Liaw Wiener, 2002; Svetnik et al., 2003) in the caret R-package (Kuhn Team, 2021), on the basis of spectral signatures extracted for each training sample form the WV data. Reclassification of Land-Cover: Using vector data layers, provided by Miami-Dade County and obtained from OpenStreetMap, a post-detection expert classifier evaluated each pixel to belong to either of the final nine classes. Vector layers that were used for the reclassification are buildings (polygons), edge of pavement (polylines converted to polygons), railroads (polylines buffered at 3 m distance), water bodies (polygons), and agricultural areas (polygons). Expert-Classifier Rules: Asphalt, Truck Trailers on Asphalt, and Concrete were reclassified as Impervious except for shrubs and trees. Areas covered by the edge of pavement or railroads were reclassified as Road Railroad except for Tree Shrub. Areas covered by the water mask were reclassified as Water except for Tree Shrub, Grass, and Wetland. Buildings dominated over all classes except for Tree Shrub. Cropland dominated over Grass and Bare Ground, but not over Tree Shrub. A resampled 2m resolution digital canopy height model (DCHM) derived from 2016 Lidar (Light Detection and Ranging) point cloud data was used to reclassify wetland areas with a height of at least 50cm as Tree Shrub. Morphological Filter: The final map, in order to remove spurious pixels and areas smaller than the class-specific minimum mapping unit (MMU), was spatially filtered using a 4-edge kernel applying a nearest neighbor replacement method. The final filtered classes and their minimum mapping units in square meters: Impervious (40), Roads / Railroad (40), Bare Ground (20), Buildings (8), Grass (20), Trees / Shrubs (8), Wetland (200), Water (200), Cropland (200). Map Accuracy: A design-based accuracy assessment of land-cover class stratified random samples (N = 564; multinomial distribution sampling based on a 95% confidence) estimated the bias adjusted overall accuracy of the map to be 87.4%, with a standard error of 1.5% which means that the 95% upper and lower confidence of the true accuracy is estimated to be between 85.9% and 91.3%. Class-specific map accuracies ranged from 67.2 ± 5.8% for bare ground to 98.5 ± 1.5% for the streets/railroads and cropland classes. Buildings were mapped with an adjusted accuracy of 95.5 ± 2.5%, grass with 80.6 ± 4.9%, trees/shrubs with 85.1 ± 4.4% and impervious with 86.6 ± 4.2%. Grass was predominantly misclassified as trees and shrub at 16.4% and vice versa at 11.9%, bare ground as impervious (19.4%), and impervious as buildings (9.0%).