Accurate methane (CH₄) emission estimates are essential for attributing regional CH₄ sources and quantifying their spatial and temporal variability, particularly in high-emission regions of China under stringent mitigation policies. Here, we apply CTDAS-WRF, a regional inversion framework that couples the CarbonTracker Data Assimilation Shell (CTDAS) with WRF-Chem, to quantify CH₄ emissions over East Asia. Through Observing System Simulation Experiments (OSSEs), we find that the system can recover ~61 % of the imposed prior-to-truth adjustment in the Yangtze River Delta (YRD), 68 % in South Korea, and 93 % in northern Japan, indicating good performance in observation-rich regions. Using a composite prior for anthropogenic emissions based on EDGAR_2024 and CAMS v6.2, the 2022 inversion yields total posterior emissions of 7.530.09 Tg yr^-1 in the YRD, which is 11.0 % lower than the prior (8.46 Tg yr^-1). Prior anthropogenic emissions are overestimated by 12.7 % (0.9 Tg yr^-1), while prior natural emissions are also higher by 0.03 Tg yr^-1. Daily posterior emissions exhibit a much stronger seasonal cycle and peak in summer, with natural sources and rice cultivation dominating the summertime underestimate in the prior. Independent evaluation at in situ station further supports the posterior improvement, with the daily mean absolute error reduced by 3.4 ppb (11.0 %) relative to the prior simulation. This work highlights the importance of accounting for seasonal variability in YRD emissions and provides a basis for future inversions that integrate in situ and satellite observations to better constrain the CH₄ budget of East Asia.