The Effect of Topographic Correction on Canopy Density Mapping Using Satellite Imagery in Mountainous Area

Abstract

One of the main factors contributing to radiometric distortion on remote sensing data is a topographic effect, but it can be reduced by applying the topographic correction. This study identifies the effect of topographic correction on canopy density mapping in the Menoreh Mountains, Indonesia. Topographic correction methods examined in this research are C-Correction, Minnaert, and Sun-Canopy-Sensor+C (SCS+C). Canopy density estimation was approached using vegetation indices, i.e., Normalized Difference Vegetation Index (NDVI), Modified Soil Adjusted Vegetation Index (MSAVI), Aerosol Free Vegetation Index (AFRI) 1.6, and AFRI 2.1 derived from Landsat-8 OLI imagery. We evaluated the performance of topographic correction by visual and statistical analysis before comparing the accuracy of canopy density estimation of different vegetation indices and correction methods. The results showed that topographic normalization could increase the accuracy of canopy density mapping. The most significant improvement is the model using MSAVI, which increases by 1.207% using the Minnaert method to reach 86.692% accuracy. Meanwhile, NDVI, AFRI 1.6, and AFRI 2.1 have less significant improvement with the maximum increase of 0.241%, 0.057%, and 0.032% using the SCS+C method, reaching the accuracy of 88.980%, 83.303%, and 82.308%, respectively. The accuracies were slightly improved since the algorithms have already reduced the effect of topography, which are categorized as ratio vegetation indices. SCS+C is the best topographic correction method because of not only the appropriate assumption of canopy normalization but also its consistency and better accuracy in canopy density estimation, among other methods.

Assessing the Influence of UAV Altitude on Extracted Biophysical Parameters of Young Oil Palm

“…Ground measured CPA diameter and tree height were used to validate UAV derived parameters. Standard error of estimation (SE) with a 95% confidence level was used to assess the accuracy of UAV extracted CPA diameter and tree height with the ground information [39,40]. For NDVI and NDRE, root mean squared deviation (RMSD) was calculated to determine the absolute difference between each UAV altitude [41].…”