The convergence of artificial intelligence (AI), including large vision models (LVMs) and transformers, with advanced quantitative remote sensing is revolution-izing the ways in which we monitor and manage agricultural landscapes and...
moreThe convergence of artificial intelligence (AI), including large vision models (LVMs) and transformers, with advanced quantitative remote sensing is revolution-izing the ways in which we monitor and manage agricultural landscapes and vegetation. As we face unprecedented challenges from climate change, volatile global markets, and a growing population, AI-driven analysis of data from multi-constellation satellite (e.g., Sentinel, Landsat-9, NISAR), aerial, and in situ sensors offers a transformative pathway to enhance global food security and promote sustainable ecosystem management. These intelligent systems provide critical, real-time insights into crop health, photosyn-thetic efficiency, and biomass productivity, enabling a paradigm shift toward precision agriculture and climate-smart resilient practices.
This Special Issue aims to explore the latest advancements in AI, machine learning (ML), and physics-informed neural networks (PINNs) and their wide ranging applica-tions in agriculture and vegetation studies. We seek to move beyond traditional mapping toward the quantitative retrieval of biophysical traits. We invite authors to provide sub-missions that showcase innovative methodologies, cutting-edge sensor applications (e.g., Hyperspectral, LiDAR, SAR, and Thermal fusion), and novel modeling approaches that contribute to the digital transformation of agriculture in a new era of geospatial big data and carbon neutrality.
This Special Issue welcomes original research articles, reviews, and technical notes on topics including the following:
• Novel AI/ML and Foundation Model Development: Creation of geospatial foundation models, self-supervised learning, and vision transformers (ViT) for processing multi-temporal remote sensing data.
• Large-Scale Vegetation Mapping and Biodiversity: Application of AI for high-resolution crop type mapping, species identification, and invasive spe-cies detection using PlanetScope, WorldView, and Sentinel-2.
• Quantitative Biophysical and Biochemical Parameter Retrieval: Leveraging AI for the retrieval of Leaf Area Index (LAI), chlorophyll content, nitrogen use ef-ficiency (NUE), sun-induced fluorescence (SIF), and canopy water content.
• Precision Agriculture and Regenerative Farming: AI-driven applications for real-time yield forecasting, soil organic carbon (SOC) estimation, carbon farming, and the detection of disease, pests, and nutrient deficiencies.
• Forestry, Carbon Sequestration and ESG Metrics: Monitoring deforestation (REDD+), forest degradation, and regrowth; estimating above-ground bio-mass (AGB) and carbon stocks for carbon credit verification.
• Urban Ecology and Green Infrastructure: Mapping urban heat island mitiga-tion, quantifying ecosystem services, and monitoring the health of urban veg-etation using high-spatial-resolution (HSR) data.
• Time-Series Analysis and Phenological Change Detection: Advanced meth-ods (e.g., LSTM, RNNs) for analyzing long-term data records to detect pheno-logical shifts, land-use change, and climate-induced disturbances.
• Multi-Sensor Data Fusion and Interoperability: Innovative techniques for the synergy of Optical-SAR fusion (e.g., Sentinel-1/2, NISAR), hyperspectral (PRISMA/EnMAP), LiDAR (GEDI), and UAV-to-satellite cross-scaling.
• Vegetation Stress and Climate Resilience Monitoring: Early detection of drought, heat stress, and anthropogenic pressure using thermal infrared (TIR) and multi-source indicators.
• Cloud Computing and Scalable Big Data Workflows: Development of auto-mated pipelines on Google Earth Engine (GEE), Microsoft Planetary Comput-er, and AWS for planet-scale monitoring.
• Explainable AI (XAI) and Model Interpretability: Research focused on making complex AI models transparent, physically consistent, and trustworthy for scientific and policy applications.
• Generative AI and Synthetic Data Augmentation: Using generative adversari-al networks (GANs) and diffusion models to simulate realistic imagery and labels for training in data-scarce environments.