TY - ABST

T1 - High-resolution mapping of tree mortality in European forests

AU - Cheng, Yan

AU - Oehmcke, Stefan

AU - Mosig, Clemens

AU - Beloiu, Mirela

AU - Kattenborn, Teja

AU - Abel, Christin

AU - Gominski, Dimitri Pierre Johannes

AU - Nord-Larsen, Thomas

AU - Fensholt, Rasmus

AU - Horion, Stéphanie

PY - 2024/3/11

Y1 - 2024/3/11

N2 - Tree mortality has escalated worldwide in recent years due to climate warming and unprecedented drought events. However, mapping tree mortality across forest ecosystems has not yet been achieved. Aerial photos provide opportunities to reveal the spatial and spectral characteristics of canopy death at local to landscape scales. In this work, we present a deep learning model for mapping tree mortality from aerial photos in various forested ecosystems across Europe. This model builds on a baseline model trained with data on dead tree canopies from California using sub-meter resolution aerial photos and allows the use of various spatial resolutions of the input images (ranging from 10 to 60 cm). By comparing our results to ground observations and/or state-of-the-art forest disturbance and loss products, we will discuss the advantages and limitations of aerial photo-based tree mortality mapping. The proposed framework can be used for large-scale mapping of tree mortality from multi-year aerial photos. The tree mortality maps provide detailed information that can help understand the mechanisms of tree mortality under climate change. Furthermore, aerial photo-based maps can serve as training labels for mapping pixel-level deadwood fractions from satellite images, which enables seamless spatial coverage and could be an essential step towards a global map of tree mortality.

AB - Tree mortality has escalated worldwide in recent years due to climate warming and unprecedented drought events. However, mapping tree mortality across forest ecosystems has not yet been achieved. Aerial photos provide opportunities to reveal the spatial and spectral characteristics of canopy death at local to landscape scales. In this work, we present a deep learning model for mapping tree mortality from aerial photos in various forested ecosystems across Europe. This model builds on a baseline model trained with data on dead tree canopies from California using sub-meter resolution aerial photos and allows the use of various spatial resolutions of the input images (ranging from 10 to 60 cm). By comparing our results to ground observations and/or state-of-the-art forest disturbance and loss products, we will discuss the advantages and limitations of aerial photo-based tree mortality mapping. The proposed framework can be used for large-scale mapping of tree mortality from multi-year aerial photos. The tree mortality maps provide detailed information that can help understand the mechanisms of tree mortality under climate change. Furthermore, aerial photo-based maps can serve as training labels for mapping pixel-level deadwood fractions from satellite images, which enables seamless spatial coverage and could be an essential step towards a global map of tree mortality.

U2 - https://doi.org/10.5194/egusphere-egu24-20213

DO - https://doi.org/10.5194/egusphere-egu24-20213

M3 - Conference abstract for conference

T2 - EGU General Assembly 2024

Y2 - 15 April 2024 through 19 April 2024

ER -