Combinational Biophysical Composition Index (CBCI) for Effective Mapping Biophysical Composition in Urban Areas

The current spectral indices for estimating urban biophysical compositions still have several limitations. First, most of those indices can highlight one urban biophysical composition, such as impervious surfaces, vegetation, or water. Furthermore, most of those indices depend on short-and mid-wave infrared and thermal infrared bands and thus have limited availability in high-resolution imagery. Finally, the separability between impervious surfaces and bare soil still needs to improve. Therefore, this paper proposed a combinational biophysical composition index (CBCI) for effective highlighting the four-major urban biophysical compositions. Correlation analyses were performed to estimate the relationship between the CBCI values and biophysical compositions, and the spectral discrimination index (SDI) was conducted to test the separability degree between impervious surfaces and bare soil. Moreover, normalized difference vegetation index (NDVI), normalized difference building index, BCI, and the combinational build-up index were utilized to perform the comparative analyses. Moreover, this paper investigated the applicability of the CBCI in different spatial resolution remote sensing images. Our results indicated that the CBCI had a much closer relationship with impervious surface fraction than the other four indices and had a stronger correlation with vegetation abundances when compared with the NDVI. In addition, the CBCI had much higher SDI values than the other indices. This paper proved that the CBCI was an effective index for highlighting the four-major urban biophysical compositions and had good performance for separating impervious surfaces and bare soil.