Mapping of Moisture-Driven Ecosystems using hyper- and multispectral VNIR/SWIR Imaging and SLAM LiDAR scanning
Research areas: Remote Sensing
Principal investigators: Dr. Christoph Hütt, Prof. Dr. Georg Bareth
Project Info: Phase 3
The primary focus of this research project is to upscale and regionalize knowledge gained in the first two funding periods by utilizing the capabilities of the recently available hyperspectral satellite data, coupled with advanced Unmanned Aerial Vehicles (UAV), to examine the moisture dynamics of the Atacama Desert environment. Our research is centered around three main objectives (Fig. 1):
- First, our detailed analysis of Tillandsia landbeckii focuses on uncovering small-scale variations in its distribution to indicate local humidity and fog patterns. In collaboration with subprojects Z02, Z03, and D04, this study builds on the foundational data from subproject B01. We aim to provide new insights through in-depth spatial mapping and analysis, concentrating on how these plants adapt to changes in moisture dynamics. This work is critical for enhancing our understanding of localized moisture dynamics in these habitats and, finally, for upscaling carbon (C) and nitrogen (N) stocks in the Atacama using advanced satellite imagery.
- Second, Lichens are crucial in the desert ecosystem, indicating moisture availability through hydration-induced metabolic activity. Collaborating with subprojects B02, Z02, and Z03, we use UAV-based surveying for their monitoring sites with lichen occurrence. This builds on our previous work mapping gypsum-rich material in the Atacama, demonstrating the feasibility of remote sensing for biogenic crusts. However, our current focus extends beyond mere mapping; we aim to integrate the knowledge we gained about lichens from the earlier phases into a broader ecological and environmental framework. By doing so, we seek to understand their broader implications on desert ecosystems, particularly regarding biogeographical patterns and climatic adaptability.
- The third goal is to map Huidobria chilensis and Huidobria fruticosa in close cooperation with subproject B1. Based on the manually measured reference data provided by B1, we will train and validate models to use remotely sensed data, such as reflectance variations measured by the UAV camera and plant height and volume, obtained from 3D modeling to approximate traits such as plant biomass and plant vitality, plus their temporal variations. Then, those models can be applied to each plant in the study area, providing biomass and vitality information for thousands of plants, which is not feasible manually. Furthermore, our remote sensing methods provide valuable information about each plant location based on the relief, such as exposition, height, and other related parameters. Moreover, by integrating these detailed local observations with broader satellite data, our approach enhances the understanding of Huidobria chilensis and Huidobria fruticosa in their specific habitat and sheds light on larger ecological patterns. By synthesizing fine-scale and satellite data, we can extrapolate our local observations to gain insights into wider environmental dynamics and trends observed at a regional level.
To meet these objectives, our approach involves using high-resolution visible (VIS), near-infrared (NIR), and shortwave infrared (SWIR) imagery from UAVs and laser scanning (LiDAR - Light Detection and Ranging; SLAM Simultaneous Localization and Mapping) that can be used handheld or mounted on a UAV for detailed 3D and topographical data acquisition. Furthermore, we incorporate state-of-the-art satellite imagery (e. g., EnMAP, WorldView-3, Pléiades-Neo, Planet). Our successful trial runs during a Z02 field campaign (2023) have proven the viability of deploying advanced equipment in challenging desert conditions. In this multifaceted approach, we will fill existing knowledge gaps and provide a deeper understanding of the matter cycle of the Atacama Desert ecosystem, its relationship with moisture, and its impacts and interactions from a local up to a regional scale.