How and why the fate of individuals, populations, and species vary across space and time is a fundamental question in ecology. Currently, a prominent gap exists in our knowledge on the local dynamics of individuals, their drivers, and how they scale to the dynamics of species distributions across space (centimeters to biomes) and time (years to millennia). Closing the scale gap is essential for understanding vegetation dynamics under global change and their related biodiversity, ecosystem, and societal consequences. The aim of this project is two-fold: (1) investigating the degree to which drone-based remote sensing contribute to closing the scale-gap and add distinctive insight on vegetation dynamics by analyzing a unique data set of ground-based observations of vegetation composition and drone-imagery sampled across Greenland, and (2) establishing and consolidating the infrastructure and competences needed for mastering unique and innovative applications of drone-based remote sensing to answer questions in ecology. The project substantially contributed to establishment of the UAS4Ecology Lab, a research facility using Unmanned Aerial System (UAS) technology in combination with novel sensor technology (see BIOSENS) to address ecological questions.
With climate warming, shrub cover is expected to expand upward in elevation. Detecting vegetation change with imagery requires that models are transferrable across time and space. To quantify the likelihood with which we can detect vegetation change in the future, we assessed the transferability of vegetation classification across 108 plots randomly stratified across altitudes (Kolyaie et al. in press). We find good transferability of Arctic shrub cover classification, which is promising for vegetation monitoring using image classification of ultra-high spatial resolution imagery acquired with hand-held cameras or from drones. Quantification of shrub cover and land-cover classification are essential for monitoring and change detection purposes, as well as upscaling of various ecosystem processes. In Karami et al. (2018), we combined Landsat 8 data with drone-based and field-based observations collected across Greenland and were able to produce a land-cover classification map with a resolution of 30 m across Greenland.
Shrub dynamics is periodically affected by insect outbreaks. Linking information from wood anatomy to satellites we have mapped the spatiotemporal extent of outbreaks in Western Greenland (see key paper under sDYN). This project contributes to theme  Fundamental Biodiversity Dynamics and theme  Ecoinformatics and New Technologies of BIOCHANGE.