A Review of Source-Term Estimation for Continuous Methane Monitoring: From Data Acquisition to Modeling and Estimation

Accurately quantifying methane emissions in the oil and gas industries is essential for meeting global climate targets. Combining fixed continuous monitoring systems with source-term estimation (STE) offers a promising pathway. This article provides a systematic review of the technical framework, comprising two core steps: data collection and inverse modeling, with the aim of identifying key technological bottlenecks and outlining directions for future development. The main conclusions are as follows: (1) There is a fundamental misalignment between prevailing sensor-deployment strategies and the ultimate inversion objective. Current optimization criteria primarily seek to maximize detection probability while neglecting their direct coupling with the minimization of uncertainty in the inferred parameters, which constitutes a central bottleneck limiting overall system accuracy. (2) Model and algorithm selection involves pronounced trade-offs and interdependencies. The study identifies an inherent contradiction between the high fidelity and computational efficiency of dispersion models, which directly constrains the choice of estimation algorithms. Moreover, Bayesian approaches offer clear advantages over deterministic optimization when confronting ill-conditioned inverse problems arising from sparse data because they accommodate prior information and enable uncertainty quantification. On the basis of these findings, future research should focus on developing sensor-deployment theory oriented toward inversion accuracy, building high-fidelity surrogate models that balance accuracy and efficiency, and advancing probabilistic inversion methods for multisource, sparse data. These recommendations are intended to guide the large-scale implementation of fixed continuous monitoring systems in the oil and gas sector and to accelerate technological innovation and practice in the green, low-carbon transition.