| Abstract: | Digital elevation models (DEMs) and their derived parameters play an important role in geosciences. When upscaling, the influence of the upscaling procedure on the accuracy of DEMs' derived parameters has not been systematically studied. In this study, slope is taken as an example to explore the effect of upscaling procedure on the accuracy of DEMs' derived parameters. The meaning of upscaling is converting the data from finer resolution to coarser resolution. The two upscaling procedures of EES (elevation to elevation first, then elevation to slope) and ESS (elevation to slope first, then slope to slope) combined with four aggregation rules of maximum, mean, median, and minimum were used to obtain the slope result at the desired resolution in plain, hill, and mountain areas. From the experimental results, it can be concluded that the upscaling procedure significantly affected slope accuracy at the desired resolution. To obtain the slope result with higher accuracy during upscaling, it was recommended that the upscaling procedure of ESS combined with the mean or median aggregation rule was adopted whatever the terrain complexity. This study can be applied for the selection of upscaling procedure and aggregation rule in the upscaling of slope and it warns us to pay more attention to the upscaling procedure when obtaining the DEM‐derived parameters at multi‐resolutions. |
