摘要： Tasseled cap transformation (TCT) is a commonly used remote-sensing technique and has been successfully used in various remote sensing-related applications. However, the TCT coefficient set is sensor-specific, and therefore, in this article, we developed the TCT coefficients specifically for Sentinel-2 multispectral instrument at-sensor reflectance data. A total of ten synchronous image pairs of Sentinel-2 and Landsat-8, collected from the different parts of the world, were used for this approach. Instead of using the traditional Gram–Schmidt orthogonalization (GSO) method, we derived the coefficients using a principal component-based Procrustes analysis (PCP) method. This was done by rotating principal component axes of Sentinel-2 data to align to Landsat-8 Operational Land Imager tasseled cap axes via the Procrustes analysis. The results show that the TCT coefficients derived from our PCP method can effectively enhance brightness, greenness, and wetness characteristics of the Sentinel-2 imagery. The results have also been compared with those of a previous study that derived the Sentinel-2 TCT data using the GSO method instead. Landsat-8 and moderate resolution imaging spectroradiometer (MODIS)-derived TCT data have been used as references for comparison. The comparison shows that the PCP method outperforms the GSO method, as the PCP’s results generally have lower root mean square errors and higher correlation coefficients (R) when compared with the corresponding TC components of Landsat-8 and MODIS data, respectively. The greatest difference between the PCP and GSO methods lies in the wetness component. The PCP method can correctly highlight vegetation and soil moisture, as well as water features in the component.