Derivation of tasseled cap transformation coefficients for SDGSAT-1 Multispectral Imager at-sensor reflectance data

Abstract

The tasseled cap transformation (TCT) is a widely used technique for reducing remote sensing multispectral data into three tasseled cap (TC) components – brightness, greenness, and wetness – while retaining essential information for various applications. We derived the TCT coefficients for 7-band SDGSAT-1 Multispectral Imager data for the first time by leveraging established Sentinel-2 TCT coefficients. This was achieved through Principal Component Analysis (PCA) for dimensional reduction of SDGSAT-1 data and the Procrustes Analysis (PA) method for aligning the principal components’ eigenvectors with the directions of Sentinel-2 TC components. A comparison between the new SDGSAT-1 coefficients and those of Sentinel-2 and Landsat-8 revealed a strong correlation, demonstrating similar characteristics for brightness, greenness, and wetness components. Given the established applications of TCT, the SDGSAT-1 TCT could significantly facilitate the use of SDGSAT-1 Multispectral Imager data for vegetation monitoring, water body analysis, and change detection. This study not only presents the derivation of SDGSAT-1 TCT coefficients but also highlights the effectiveness of the PA method in deriving TC wetness component coefficients that are sensitive to water bodies and vegetation, even for multispectral data lacking the moisture-sensitive shortwave-infrared (SWIR) band.

Keywords

• SDGSAT-1
• tasseled cap transformation
• principal component analysis
• Procrustes analysis