Fig 1.
(a) Digital Elevation Model (DEM) from Copernicus (ESA); (b) False-color composite of Landsat 8 OLI imagery (USGS). Images were processed by the authors to illustrate terrain.
Table 1.
Percentage Area of Each Land Cover Type.
Table 2.
Selected vegetation indices and formulas.
Table 3.
Selected Terrain Correction Methods and Formulas.
Fig 2.
Near-infrared (NIR), Red and Green composites for spring (a–d) and winter (e–h) before and after topographic correction.
Fig 3.
Relationships between reflectance and the cosine of the solar incidence angle in spring.
Blue, Red and NIR reflectance before and after topographic correction using the SCS + C, SCSCTS, and PLC models on April 16, 2020. R2 denotes the coefficient of determination between reflectance and cos i, where lower values indicate weaker topographic dependence.
Fig 4.
Relationships between reflectance and the cosine of the solar incidence angle in winter.
Red and NIR reflectance before and after topographic correction using the SCS + C, SCSCTS, and PLC models on November 10, 2020. R2 denotes the coefficient of determination between reflectance and cos i, where lower values indicate weaker topographic dependence.
Fig 5.
Percentages of outlier reflectance pixels under different topographic correction methods.
(a) April 16, 2020; (b) November 10, 2020.
Fig 6.
Distribution of shaded and sunlit slopes in rugged terrain.
(a) April 16, 2020; (b) November 10, 2020.
Fig 7.
Absolute differences in reflectance between sunlit and shaded slopes before and after topographic correction.
Corrections were performed using the SCS + C, SCSCTS, and PLC models. (a) April 16, 2020; (b) November 10, 2020.
Fig 8.
Weighted RDMR values of median reflectance under different topographic correction methods.
Results are summarized by land-cover types. (a) April 16, 2020; (b) November 10, 2020.
Fig 9.
Vegetation indices in spring before and after topographic correction.
NDVI, EVI, NIRv, and RDVI derived on April 16, 2020 using the SCS + C, SCSCTS, and PLC models.
Fig 10.
Vegetation indices in winter before and after topographic correction.
NDVI, EVI, NIRv, and RDVI derived on November 10, 2020.
Fig 11.
Relationships between vegetation indices and the cosine of the solar incidence angle in spring.
NDVI, EVI, NIRv, and RDVI before and after topographic correction on April 16, 2020. R2 represents the coefficient of determination between vegetation indices and cos i, with lower values indicating improved topographic correction.
Fig 12.
Relationships between vegetation indices and the cosine of the solar incidence angle in winter.
NDVI, EVI, NIRv, and RDVI before and after topographic correction on November 10, 2020. R2 represents the coefficient of determination between vegetation indices and cos i, with lower values indicating improved topographic correction.
Fig 13.
Outlier pixels across different slope classes in April (1) and November (2).
Blue (a), Red (b), and NIR (c) reflectance bands are shown.