Development of a Prototype EnKF-WRF-CH4 Inverse Modeling System for Reginal to Local Methane Emission Retrieval

Atmospheric methane (CH4) is the second-most important anthropogenic greenhouse gas (GHG) after carbon dioxide (CO2), playing a crucial role in the Earth's climate system. The imbalance between methane sources and sinks is evident from the continuous rise in atmospheric methane concentrations, which significantly contributes to global temperature changes and is a key driver of recent global warming trends. Since methane emissions originate from natural and anthropogenic sources, identifying opportunities to reduce anthropogenic emissions has become a priority for policymakers aiming to stabilize global methane concentrations and mitigate climate change. To achieve this, precise and reliable quantification of methane emissions is essential for informing effective climate policies and strategies.
In this work, we present the development of an iterative 4DEnSRF (a variant of the Ensemble Kalman Filter, EnKF) data assimilation system designed to estimate surface methane emission fluxes and atmospheric methane concentrations dynamically. The EnKF system is integrated with the WRF-based weather prediction-CH4 transport/dispersion model (WRF-CH4), forming an EnKF-WRF-CH4 inverse modeling system. This system can assimilate weather observations and multi-scale methane measurements, including those from dense ground observation networks, tower measurements, mobile and aerial platforms, and satellite data, providing the necessary constraints for emission flux inversion. The system will first be tested with Oklahoma's Anadarko basin for oil and gas production.
To evaluate the performance of our inverse modeling system, we conduct observing system simulation experiments (OSSEs) that aim to isolate and investigate the impacts of data assimilation configurations and CH4 observation network configuration and data quality on flux inversion. A dense network of low-sensitivity surface CH4 sensors and vehicle-based high-sensitivity mobile measurements (e.g., Li-Cor sensor) will be tested. The OSSEs will also explore the effects of different types of perturbations and inflation to the emission flux ensemble, observation density and error, assimilation window length, covariance localization radius, model grid spacing, and weather conditions. Detailed results will be presented at the conference.