Background Hidden hazards in coalfield fire zones are characterized by limited detectability, rapid evolution, and multi-hazard coupling, posing major challenges to the safe coal mining and related ecological restoration.

Methods This study aims to achieve the fine-scale identification and risk control of hidden hazards in coalfield fire zones. Based on a systematic review of the genetic mechanisms and spatial distribution patterns of coalfield fire zones, this study summarizes the characteristics of three primary fire zone types: outcrop/surface fire zones, shallowly buried fire zones, and deeply buried underground burnt-out areas and fire zones within old coal pits. Multi-dimensional hazards caused by coalfield fire zones, including geological environment degradation, atmospheric pollution, and greenhouse gas emissions, are analyzed. Accordingly, a discriminant indicator system that accounts for both spatial structure and temporal evolution is established. Centered on technologies for detecting hidden hazards, this study presents a comprehensive summary of the fundamental principles, advances, and representative applications of mainstream technologies (i.e., dynamic monitoring based on space-air integrated remote sensing, temperature- and geochemical-field survey, electrical/electromagnetic prospecting, and elastic-wave survey) and two supplementary methods (e.g., ground-penetrating radar and magnetic survey). A comparative analysis is conducted on the applicability and limitations of these different techniques from the perspective of detection depth, spatial resolution, timeliness, and cost. Meanwhile, the advantages and limitations of stereo-collaborative multi-source detection are derived. Accordingly, key bottlenecks currently faced in the detection of hidden hazards in coalfield fire zones are identified, involving the transparency of deep fire sources, resistance to complex environmental interference, multi-source data-physics fusion and joint inversion, four-dimensional dynamic monitoring, and intelligent interpretation and relevant uncertainty quantification.

Results To meet the demands for intelligent mining and transparent geology of coal mines, an overall technical framework is proposed for detecting hidden hazards in coalfield fire zones. This framework consists of multi-source stereo-collaborative detection, along with space-air-ground-borehole multi-source information fusion and intelligent inversion. First, a space-air-ground-borehole multi-modal spatiotemporal database and knowledge graph are to be constructed. Second, physics-constrained and data-driven methods for multi-source information fusion and intelligent inversion should be developed. Finally, it is necessary to establish a patrol-monitoring-verification-warning closed-loop system that integrates multi-source collaborative sensing, joint inversion and interpretation, three-dimensional transparent representation, risk assessment and early warning, and comprehensive treatment.

Conclusions The results of this study provide a systematic reference for the precise detection, dynamic monitoring, and prevention and control decision-making of hidden hazards in coalfield fire zones. Furthermore, the results are of great significance for safe coal mining, regional ecological restoration, carbon neutrality, and peak carbon dioxide emissions.