Bradley Gay, Kimberley Miner, Nils Rietze, Benjamin Poulter, Neal Pastick, Charles Miller
Abstract Across the circumarctic, permafrost landscapes store approximately 1700 billion metric tons of organic carbon - nearly twice atmospheric levels - yet reservoir stability depends on the zero-curtain, a subsurface thermal plateau that emerges when latent heat maintains soil temperatures near 0°C during freeze-thaw phase transitions. The zero-curtain sustains liquid water through cryosuction within the active layer, enabling microbial activity to persist well into the cold season and regulating permafrost thermal stability. However, zero-curtain intensity, duration, and spatial extent remain inadequately quantified during transitional seasons when isothermal buffering exhibits maximum variability, limiting our understanding of permafrost-climate feedbacks. Here, we show that zero-curtain phenomena exhibit pronounced seasonal asymmetry, with extended vernal intensification (1000-4000 hours) relative to compressed winter occurrence (100-500 hours), and significant longitudinal variations, with moderate intensity patterns across the North American Arctic, enhanced vernal amplification in Siberia, reduced winter suppression in Fennoscandia, and a delayed vernal response in the Canadian Archipelago, with systematic amplification under warming conditions (2015-2024). GeoCryoAI, a hybridized physics-informed transfer learning framework, integrates 62.71 million in situ measurements and 3.3 billion remote sensing observations to quantify these dynamics at 30 m resolution, achieving 96.4% detection accuracy and 39% improvement over conventional models. Ablation studies confirm that the full GeoCryoAI architecture achieves 11.8% improvement over baseline MLP architectures (93.4% vs 81.6% in component validation experiments), with physics-informed constraints providing essential regularization for thermodynamic consistency. Mechanistic analysis reveals soil moisture-latent heat coupling drives 60-90% of duration variability, amplified 20-40% under warming. This framework establishes a NISAR-ready circumarctic monitoring protocol, enabling 3-6-month forecasts and quantitative constraints for Earth system models simulating carbon-climate feedbacks in a warming Arctic.