A spectral-spatial-temporal attention network for tree species mapping using DESIS hyperspectral imagery

Accurate mapping and monitoring of forest tree species are crucial for understanding ecosystem dynamics [1], assessing biodiversity [2], and enabling sustainable forest management [3]. Tree species adapt their morphology and phenology to the environment [4], leading to variability in spectral signatures across geographic regions. Furthermore, the spectral reflectance of a given tree species varies significantly with growth stages and seasons [5], making the classification based solely on RGB data extremely challenging. At the local level, spectral variability also closely correlates with stand structure factors such as crown size, stand density, and gap sizes. This results in varying signal reflectance from different parts of the same crown, further complicating tree species classification [6]. Thus, we proposed a spectral-spatial-temporal constrained deep learning method, an end-to-end multi-head attention-based network, to automatically extract deep features for tree species mapping. Employing this model on multi-temporal hyperspectral imagery from the DLR Earth Sensing Imaging Spectrometer (DESIS), we produced a 30 m resolution forest species distribution map of the Harz Forest in Germany. DESIS, a VNIR sensor aboard the International Space Station, captures detailed Earth images upon request, offering extensive spectral data across 235 bands ranging from 400 to 1000 nm [7]. Our methodology leverages the comprehensive spectral information provided by DESIS, enhancing the tree species mapping accuracy. Utilizing the reference data from TreeSatAI Benchmark Archive [8], we prepared 134,886 hyperspectral data patches, each labelled with tree species information. The evaluation involved assessing the F1-score, Jaccard index, Hamming loss, and accuracy for various tree species using National Forest Inventory (NFI) data plots. The results reveal the potential of deep learning using hyperspectral data in the precise and automated mapping of forest tree species distribution, thereby supporting evidence-based decision-making in sustainable forest management.

 

[1] Welle, Torsten, et al. "Mapping dominant tree species of German forests." Remote Sensing 14.14 (2022): 3330.

[2] Grabska, Ewa, David Frantz, and Katarzyna Ostapowicz. "Evaluation of machine learning algorithms for forest stand species mapping using Sentinel-2 imagery and environmental data in the Polish Carpathians." Remote Sensing of Environment 251 (2020): 112103.

[3] Xie, Bo, et al. "Analysis of regional distribution of tree species using multi-seasonal sentinel-1&2 imagery within google earth engine." Forests 12.5 (2021): 565.

[4] Chuine, Isabelle. "Why does phenology drive species distribution?." Philosophical Transactions of the Royal Society B: Biological Sciences 365.1555 (2010): 3149-3160.

[5] Hesketh, Michael, and G. Arturo Sánchez-Azofeifa. "The effect of seasonal spectral variation on species classification in the Panamanian tropical forest." Remote Sensing of Environment 118 (2012): 73-82.

[6] Ferreira, Matheus Pinheiro, et al. "Tree species classification in tropical forests using visible to shortwave infrared WorldView-3 images and texture analysis." ISPRS journal of photogrammetry and remote sensing 149 (2019): 119-131.

[7] de los Reyes, Raquel, et al. "The Desis L2a Processor And Validation Of L2a Products Using Aeronet And Radcalnet Data." The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences 46 (2022): 9-12.

[8] Ahlswede, Steve, et al. "TreeSatAI Benchmark Archive: A multi-sensor, multi-label dataset for tree species classification in remote sensing." Earth System Science Data Discussions 2022 (2022): 1-22.