Towards Monitoring Biodiversity from Space

Human impacts change the state of global biodiversity, calling for protection efforts to ensure ecosystem functions and services to humanity. Forest ecosystems provide humanity with ecosystem services via food and resources, regulate air and water, inspire art and culture, and contain the majority of Earth’s biodiversity. Biodiversity is critical for forest well-being and functioning, such as productivity and resilience to disturbances. However, pressures such as land use change, affect the state of forests and the life they harbor. To counteract these developments, the Parties to the Convention on Biological Diversity committed to promoting sustainable development and conservation of biodiversity by releasing the Kunming-Montreal Global Biodiversity Framework. Alongside defining targets and goals to protect biodiversity, the monitoring framework provides indicators to report and track the state of biodiversity. Biodiversity and biodiversity change do not perfectly align with species count. Biodiversity is a multidimensional concept describing multiple aspects of the variation of life on Earth. Focusing on one dimension of biodiversity (species diversity, genetic diversity, trait-based functional diversity) can lead to biased results that may induce partial protection efforts. Changes in trait-based functional diversity or genetic diversity might affect ecosystem functions and services before they are visible in species richness. In forest ecosystems, findings suggest that trait-based functional diversity might better predict forest functions than species richness alone. Therefore, there is a need to understand the distribution, speed, and direction of biodiversity change, as well as the causes and drivers. Field-based biodiversity data is often sparse regarding spatial, temporal, and thematic coverage. Satellite-based remote sensing or Earth Observation acquires global reflectance data of the Earth’s surface, including inaccessible areas and regularly revisited locations of interest. Therefore, Earth Observation data has the potential to contribute to biodiversity monitoring by informing and complementing field-based biodiversity research. Multiple publicly available satellite missions could support global biodiversity monitoring today and in the future. This thesis explores the potential of satellites for biodiversity assessment and monitoring via trait-based functional diversity assessments from passive-optical remote sensing data in temperate forest ecosystems. The studies presented in this thesis demonstrate and explore the potential of trait-based functional diversity in a scalable, efficient, and continuous method for assessing biodiversity from space. Until now, trait-based functional diversity approaches have mainly been studied using airborne imaging spectroscopy data. We upscaled methods and applications based on airborne imaging spectroscopy data at a temperate mixed forest in Switzerland to satellite data. In comparison, satellite data show different acquisition geometry regarding solar-to-sensor angle and spatial, spectral, and temporal resolution. Data processing and acquisition geometry affect trait-based functional diversity calculations. Before upscaling, we quantified atmospheric, topographic, and anisotropic effects via correction methods on airborne imaging spectroscopy data and discussed implications for satellite data. This allowed us to understand and quantify the impact of the changing acquisition geometry, data processing, and shadow treatment on the resulting functional diversity estimation. The two Sentinel-2 satellites operated by the European Space Agency acquire publicly available data of 10 m or 20 m resolution and ten spectral bands. We presented a scalable approach using three spectral indices designed for Sentinel-2 and studied the impact of spatial resolution by upscaling high-resolution airborne imaging spectroscopy data. We found that the overall variance of the three mapped traits was reduced by around 50% at 20 m pixel size compared to 6 m, the average crown diameter at Laegern forest. The study site at Laegern forest in the center of the northern Swiss plateau is a well-studied forest site, ideal for expanding to larger areas on the Swiss plateau. On a larger research area on the Swiss plateau, we subsequently used the maps on trait-based functional diversity to study biodiversity-ecosystem function relationships in large-scale approaches in a non-experimental setting. We showed that forest drought response was linked to trait-based functional diversity at landscape scales. The high temporal resolution of satellite-based observations, such as Sentinel-2, establishes the basis for future monitoring of functional diversity from space. By building on functional diversity maps, usually derived during peak greenness, we explored the potential of Earth Observation data for monitoring using satellite time series data. We studied seasonal and interannual variation in spectral indices and functional diversity metrics and built the groundwork for future approaches to monitoring functional diversity using Earth Observation data. Overall, this thesis contributes to understanding trait-based functional diversity assessment from satellite data and provides best practices for future approaches. Future efforts are still required to close the gap between remote sensing technologies and field-based biodiversity research. Given the urgent need for timely and global biodiversity monitoring and conservation strategies, the presented contributions are relevant to biodiversity research.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.