GeoAI for polar vegetation mapping and hydrological interactions: A systematic review

Remote sensing (RS) and Artificial intelligence (AI) are increasingly applied to monitor vegetation and hydrology in the Arctic and Antarctic, where logistical and environmental constraints make fieldwork difficult. These technologies offer new opportunities to track ecological change, but the extent, consistency, and methodological quality of current applications have not been systematically reviewed. This study presents the first PRISMA 2020 based systematic synthesis of AI enhanced RS, collectively termed GeoAI, applied to Arctic and Antarctic environments (2005-2025; 116 studies). Publication activity has expanded significantly since 2018, driven by the convergence of uncrewed aerial vehicle (UAV), multispectral imaging, satellite archives, and deep learning (DL). Bibliometric and conceptual-network analyses reveal a rapid shift from isolated ecological monitoring toward integrated, data-fusion frameworks linking vegetation, hydrology, and climate processes. Classical machine learning approaches remain foundational, while DL-based convolutional neural-network architectures are emerging as powerful tools for fine-scale segmentation and prediction. Most studies still operate at the landscape scale, with few achieving full UAV-to-satellite integration, exposing persistent spatial-resolution and validation gaps. Vegetation hydrology coupling is reported in most cases, though subsurface and process-based monitoring remain limited. Spectral-index analysis reveals a persistent reliance on greenness metrics, yet there is a growing shift toward pigment, moisture, and cryptogam-sensitive indices that more accurately capture plant physiological function and microclimatic interactions. This review establishes the empirical foundation for next-generation polar monitoring, emphasising hierarchical UAV-to-satellite fusion, open benchmark datasets, and explainable, ecologically grounded AI as essential pathways for scalable, climate-adaptive conservation of Earth's fastest-changing regions.