Canopy Height Mapping by Sentinel 1 and 2 Satellite Images, Airborne LiDAR Data, and Machine Learning

Abstract

Continuous mapping of vegetation height is critical for many forestry applications, such as planning vegetation management in power transmission line right-of-way. Satellite images from different sensors, including SAR (Synthetic Aperture Radar) from Sentinel 1 (S1) and multispectral from Sentinel 2 (S2), can be used for producing high-resolution vegetation height maps at a broad scale. The main objective of this study is to assess the potential of S1 and S2 satellite data, both in a single and a multisensor approach, for modeling canopy height in a transmission line right-of-way located in the Atlantic Forest of Paraná, Brazil. For integrating S1 and S2 data, we used three machine learning algorithms (LR: Linear Regression, CART: Classification and Regression Trees, and RF: Random Forest) and airborne LiDAR (Light Detection and Ranging) measurements as the reference height. The best models were obtained using the RF algorithm and 20 m resolution features from only S2 data (cross-validated RMSE of 4.92 m and R2 of 0.58) or multisensor data (cross-validated RMSE of 4.86 m and R2 of 0.60). Although the multisensor model presented the best performance, it was not statistically different from the single-S2 model. Thus, the use of only S2 to estimate canopy height has practical advantages, as it reduces the need to process SAR images and the uncertainties due to S1 noise or differences between the acquisition dates of S2 and S1.

Effect of the Synergetic Use of Sentinel-1, Sentinel-2, LiDAR and Derived Data in Land Cover Classification of a Semiarid Mediterranean Area Using Machine Learning Algorithms

“…Montesano et al [41] used Landsat, SAR and LiDAR to improve the estimation of forest biomass. More recently, Chen et al [42], Morin et al [43] or Torres de Almeida et al [44] combined S1, S2 and LiDAR to estimate forest aboveground biomass. Spectral and LiDAR information have also been combined to analyze urban areas [45,46].…”

“…One of the conclusions that can be drawn from these studies is that the more sources, the better results. However, an increase in the number of sources indeed increases the dimensionality of the problem, so feature selection based on importance [49,50] or comparing different subsets extracted from the main data [44] is usually performed Although S1 and S2 have been previously combined to improve land cover classification accuracy, the inclusion of LiDAR metrics in the feature dataset has been mainly focused to the estimation of specific metrics in specific environments such as forest, agricultural and urban areas. The aim of this study was to find out whether a synergetic use of different type of sensors, S1, S2 and LiDAR, and features extracted as indices and texture measures may enhance classification accuracy in a semi-arid Mediterranean area when trying to separate covers of different nature, but similar spectral characteristics, at a regional scale.…”

Comparative Study on Remote Sensing Methods for Forest Height Mapping in Complex Mountainous Environments

“…2023, 15, 2275 9 of 17 in the Sentinel-2 surface reflectance product, vegetation indices are calculated using only the green, red, vegetation red edge, near-infrared, water vapour, and shortwave infrared bands from the acquired Sentinel-2 dataset. The specific vegetation indices are selected based on the highly correlated characteristic variables for forest canopy height inversion, as described in [37][38][39][40]. In addition, to account for the effects of terrain, weather, and altitude on tree growth, factors such as elevation, slope, mean annual temperature, and mean annual precipitation are selected [41].…”

Estimation of tree height using radar remote sensing in urban settings: a preliminary result

“…The Sentinel-1 based height estimation was not possible for individual trees because of the spatial resolution. It is evident that Sentinel-1 data could give good results for large scale forest height applications 22,23 . However, it needs further investigation of SAR data with high spatial resolution (i.e., in the order of centimetre to few metres), to account the spatial limitation of the satellite data for individual tree scale.…”