Conversion of Low-resolution Slope Gradient Based on Histogram Matching at Rolling Hilly Area of Northeast China
[Objective]Slope gradient data was accurately obtained to address the limitation of slope gradient underestimation using the freely downloaded 30 m resolution digital elevation model(DEM)for farmland in Northeast China,in order to provide important data support for quantitatively evaluating soil erosion in the rolling hilly regions.[Methods]A 5 cm resolution DEM was generated from drone survey images and resampled to obtain 1,5,and 12.5 m DEM resolutions.Combined with the 30 m DEM resolution,the optimal DEM resolution for slope gradient extraction in the study area was identified.Additionally,the histogram matching method was used to establish a slope gradient conversion model between the 30 m DEM resolution and the optimal DEM resolution for each slope gradient category.[Results]① The slope gradient distributions derived from the five DEM resolutions indicated that the 1 m and 5 m DEM resolutions exhibited a strong similarity to the slope gradient distribution of the 5 cm DEM.Given that the 5 m DEM resolution corresponds to a 1∶10,000 scale topographic map,the 5 m DEM resolution was optimal for constructing the slope gradient conversion model.② Using the histogram matching method,a univariate linear model and a univariate quadratic non-linear model were developed for slope gradient conversion between the 30 m and 5 m DEM resolutions across different slope gradient segments.The linear conversion model was suitable for slopes less than 7°,while the non-linear model was more appropriate for slopes greater than 7°.③ After applying both linear and non-linear conversion models,the frequency distribution of slope gradients extracted from the 30 m DEM resolution closely matched that of the 5 m DEM resolution,significantly improving covariance and correlation coefficients.This reflected that the slope gradients after conversion from the 30 m DEM resolution can accurately represent ground undulation;additionally,the optimization results from the non-linear conversion model were superior to those from the linear conversion model.[Conclusion]The 5 m DEM resolution is the optimal resolution for extracting slope data in the study area.The developed conversion model for low-to-high resolution slope gradients showed that the non-linear slope conversion model has a better optimization effect than the linear slope conversion model.
rolling hilly region in Northeast Chinadifferent DEM-resolutionshistogram matching methodslope gradient conversion model
国家科技期刊平台
NSTL
万方数据
