Remote sensing data are increasingly used to monitor the critical zone. This study argues that drones are transforming the remote sensing paradigm. While recent satellite constellations (e.g., Sentinel-1 and Sentinel-2) have revolutionized Earth observation through improved temporal resolution, most research efforts have focused on refining data processing methods to extract key environmental variables such as soil moisture (Gao et al. 2017), snow cover (Gascoin et al. 2019), and crop classification (Blickensdörfer et al. 2022). In parallel, drone technology has become more prevalent. Their high flexibility and superior spatial, spectral, and temporal resolutions enable the acquisition of high-quality remote sensing data over limited areas. To monitor essential critical zone and biodiversity variables effectively, it is crucial to adapt acquisition protocols. These protocols must be defined through interdisciplinary collaboration, considering the processes under study before data collection. They should determine the optimal spectral and temporal resolutions based on the size of the target object (e.g., leaves, trees, landscapes), its characteristics, and the ecological, hydrological, or climatic processes influencing its behavior over time (e.g., hourly, daily, or seasonally). Drones show great potential for applications such as viticulture, forest ecology, and habitat mapping, though their use remains exploratory and experimental. This paper presents three case studies demonstrating the capabilities of drones in critical zone monitoring and their role in shifting the remote sensing paradigm. Case Studies 1. Monitoring Vineyards for Temperature Variability The first example focuses on vineyard exposure to extreme air temperatures. The objective was to detect cold air pooling, which can cause frost damage to grapevines. We hypothesized that thermal imagery could effectively identify these cold air pockets. Drone flights were conducted every 20 minutes before and after sunrise in September 2017 in Saint-Émilion, southwestern France, under low wind conditions (<2 m/s). A fixed-wing eBee+ drone equipped with a Thermomap sensor was used, and temperature sensors were deployed within a 20-hectare vineyard to validate thermal measurements. The results (Fig. 1) indicate that cold air pools were detectable and their distribution was influenced not only by valley topography but also by anthropogenic structures such as walls and buildings, which affected airflows (dark areas in Fig. 1, left). After sunrise, air and surface temperatures diverged due to differences in emissivity. 2. Mapping Semi-Natural Habitats in Marshlands The second example explores habitat mapping in marshlands, where vegetation heterogeneity and composition are critical factors. The study was conducted in the Sougéal marsh (France), where hydrological regimes affect vegetation distribution. We hypothesized that integrating high-resolution Digital Surface Models (DSM) with multispectral and multitemporal datasets could improve classification of mesophilic, meso-hygrophilic, and hygrophilic habitats (Fig. 2). Monthly drone acquisitions were conducted between May 2017 and May 2018 using a fixed-wing eBeeX drone equipped with RGB SODA and Sequoia multispectral sensors under consistent atmospheric conditions. Spectral signatures from drone imagery were used as reference data to train classification models applied to Sentinel-2 imagery. The best classification results were obtained using Sentinel-2 data from April (Fig. 2, right), whereas classifications for other months were less accurate. The study also revealed that grazing activities influenced spectral responses and habitat classification (Alvarez-Vanhard et al. 2020). 3. Detecting and Mapping Ponds in Sand Dune Ecosystems The third case study aimed to detect and map ponds within a vegetated sand dune habitat in the Vauville Natura 2000 reserve (France), an area affected by summer droughts and declining water levels. Expanding the reserve to include nearby ponds could enhance biodiversity conservation. Leveraging the thermal inertia properties of water, we combined nighttime and daytime thermal data to distinguish water bodies from surrounding vegetation. A Trinity F90+ fixed-wing drone equipped with a Micasense ALTUM-PT thermal sensor was deployed in May 2024. One-hour flights covered 150 hectares at 9:30 PM (after sunset, minimizing emissivity effects) and at noon the following day, under cloud-free conditions. The extracted water bodies are highlighted in yellow Fig. 3. Conclusion This paradigm shift in remote sensing emphasizes the need to design data acquisition protocols tailored to specific research objectives, rather than solely improving data processing methods. While enhancing processing techniques remains important for accuracy, drones offer unique flexibility in selecting the timing, frequency, and types of sensors used for data collection. Additionally, drones facilitate the integration of multispectral, thermal, RGB, LiDAR, and other sensor-based data, broadening the scope of environmental monitoring. Ongoing applications include soil moisture characterization (Wu et al. 2019), forest phenology and health assessments (Ecke et al. 2024), plant ecology studies, invasive species detection (Singh et al. 2024), and ecosystem service mapping (Schenone et al. 2021). Furthermore, drone data can be used independently or in synergy with satellite imagery. Alvarez-Vanhard et al. (2021) identified four types of drone-satellite synergies that contribute to characterizing and monitoring the critical zone.
Read full abstract- All Solutions
Editage
One platform for all researcher needs
Paperpal
AI-powered academic writing assistant
R Discovery
Your #1 AI companion for literature search
Mind the Graph
AI tool for graphics, illustrations, and artwork
Journal finder
AI-powered journal recommender
Unlock unlimited use of all AI tools with the Editage Plus membership.
Explore Editage Plus - Support
Overview
2839 Articles
Published in last 50 years
Related Topics
Articles published on Plant Ecology
Authors
Select Authors
Journals
Select Journals
Duration
Select Duration
2963 Search results
Sort by Recency