A Deep Learning Approach to Estimate Canopy Height and Uncertainty by Integrating Seasonal Optical, SAR and Limited GEDI LiDAR Data over Northern Forests

Abstract

Accurate forest canopy height estimation is essential for evaluating aboveground biomass and carbon stock dynamics, supporting ecosystem monitoring services like timber provisioning, climate change mitigation, and biodiversity conservation. However, despite advancements in spaceborne LiDAR technology, data for northern high latitudes remain limited due to orbital and sampling constraints. This study introduces a methodology for generating spatially continuous, high-resolution canopy height and uncertainty estimates using Deep Learning Regression models. We integrate multi-source, multi-seasonal satellite data from Sentinel-1, Landsat, and ALOS-PALSAR-2, with spaceborne GEDI LiDAR as reference data. Our approach was tested in Ontario, Canada, and validated with airborne LiDAR, demonstrating strong performance. The best results were achieved by incorporating seasonal Sentinel-1 and Landsat features alongside PALSAR data, yielding an R-square of 0.72, RMSE of 3.43 m, and bias of 2.44 m. Using seasonal data instead of summer-only data improved variability by 10%, reduced error by 0.45 m, and decreased bias by 1 m. The deep learning model's weighting strategy notably reduced errors in tall canopy height estimates compared to a recent global model, though it overestimated lower canopy heights. Uncertainty maps highlighted greater uncertainty near forest edges, where GEDI measurements are prone to errors and SAR data may encounter backscatter issues like foreshortening, layover, and shadow. This study enhances canopy height estimation techniques in areas lacking spaceborne LiDAR coverage, providing essential tools for forestry, environmental monitoring, and carbon stock estimation.

Figures (7)

• Figure 1: Map of the study area. The highlighted region is the study area for which airborne LiDAR data is available from the Forest Resources Inventory Leaf-on LiDAR dataset of the Government of Ontario.
• Figure 2: Pipeline of the proposed methodology. a) Descriptions of the data acquisition process. b) Illustration of the data preprocessing strategy followed to extract the training samples. c) Diagram of the neural network optimization procedure phase, and d) illustration of the inference approach using ensemble of neural networks.
• Figure 3: Overview of the ResUNet network. (a) The left (downward) branch is the encoder part of the architecture. The right (upward) branch is the decoder. The last convolutional layer has as many channels as there are distinct classes. (b) Building block of the ResUNet network. Each unit within the residual block has the same number of filters with all other units.
• Figure 4: GEDI and ALS harmonization methodology. a) pipeline of the harmonization strategy. b) Illustration of the GEDI and ALS interception tiles. c) Zonal statistics energy-based process to harmonize ALS samples with corresponding GEDI.
• Figure 5: Relationship of canopy height from GEDI, this study and META with harmonized ALS reference data. The 1:1 line between observed and predicted values is shown in dashed black line and the estimated regression in bold red line. Red color density plot indicates a high concentration of samples, while green and blue colors represent medium and low-density ranges, respectively.