Comment on Tompalski et al. Combining Multi-Date Airborne Laser Scanning and Digital Aerial Photogrammetric Data for Forest Growth and Yield Modelling. Remote Sens. 2018, 10, 347

Abstract

Tompalski et al. (2018) propose “template matching” as a (required) intermediate step to use remote sensing-based predictions of forest attributes as inputs of the Growth and Yield Projection System (GYPSY) for the simulations of forest stand dynamics in Alberta, Canada. Yet, the feasibility of the approach can be criticized for many points that call for experimental verification. The approach cannot be fully replicated based on the description of the paper. Nevertheless, an experimental implementation with synthetic data indicates that the quality of the projections may vary considerably depending on parameter assumptions for the templates, and the projections may include discontinuities between the observed and projected forest attributes. The approach is poorly motivated given that the effects described above are largely avoidable, if the underlying GYPSY models are run without the template matching step. The R-codes used for the analyses are provided as supplementary data for an interested reader wishing to evaluate the conclusions made above. A semantic analysis indicates further problems with multi-date data on a wall-to-wall grid. The projections obtained by template matching should be exposed to criticism for their realism and benchmarked against other approaches prior to using template matching as proposed by Tompalski et al.

Keywords

• forest inventory
• growth
• uncertainty
• error
• model criticism
• evaluation
• validation