Subtropical lakes are important source of atmospheric methane (CH₄). This study aims to investigate spatial variations of CH₄ flux in Lake Taihu, a large (area 2400 km²) and shallow (mean depth 1.9 m) eutrophic lake in Eastern China. The lake exhibited high spatial variations in pollution level, macrophyte vegetation abundance, and algal growth. We measured the diffusion CH₄ flux via the transfer coefficient method across the whole lake. In addition, data obtained with the flux gradient and the eddy covariance methods were used in conjunction with the data on the diffusion flux to estimate the contribution by ebullition. Results from 3 years' measurements indicated high spatial variabilities in the diffusion CH₄ flux. The spatial pattern of the diffusion CH₄ emission was correlated with water clarity, dissolved oxygen concentration, and the spatial distributions of algal and submerged vegetation. In comparison to the transfer coefficient method, the eddy covariance and the flux gradient method observed a lake CH₄ flux that was 3.39 ± 0.58 (mean ± 1 standard deviation) and 1.95 ± 0.36 times higher in an open‐water eutrophic zone and in a habitat of submerged macrophytes, respectively. The result implied an average of 71% and 49% ebullition contribution to the total CH₄ flux in the two zones. The annual mean diffusion CH₄ flux of the whole lake was 0.54 ± 0.30 g m⁻² yr⁻¹. Our CH₄ emission data suggest that the average CH₄ emission reported previously for lakes in Eastern China was overestimated.

Atmospheric CH₄ is the second-largest anthropogenic contributor to global warming. However, its emissions, components, spatial–temporal variations and projected changes still remain largely uncertain from city to national scales. CH₄ emissions from waste treatment (including solid waste landfills, solid waste incineration and sewage) account for >50 % of total anthropogenic CH₄ emissions at the city scale, and considering the high temperature sensitivity of CH₄ emission factors (EFs) to biological-process-based sources such as waste treatment, large differences will occur when estimating future CH₄ emissions under different global warming scenarios. Furthermore, the relationships between temperature and waste treatment CH₄ emissions have only been studied in a few site-specific studies and lack representativity for whole cities, which contain various biophysical conditions and show heterogeneous distribution. The above factors cause uncertainty in the evaluation of city-scale CH₄ emissions (especially from waste treatments), and projected changes still remain unexplored. Here we conduct the first tower-based CH₄ observation network with three sites in Hangzhou, which is located in the developed Yangtze River Delta (YRD) area and ranks as one of the largest megacities in China. We found the a priori total annual anthropogenic CH₄ emissions and those from waste treatment were overestimated by 36.0 % and 47.1 % in Hangzhou, respectively. In contrast, the total emissions in the larger region, i.e., Zhejiang Province or the YRD area, were slightly underestimated by 7.0 %. Emissions from waste treatment showed obvious seasonal patterns following air temperature. By using the linear relationship constructed between monthly waste treatment CH₄ emissions and air temperature, we find the waste treatment EFs increase by 38 %–50 % with temperature increases of 10 ^∘C. Together with projected temperature changes from four climate change scenarios, the global-warming-induced EFs in Hangzhou will increase at the rates of 2.2 %, 1.2 %, 0.7 % and 0.5 % per decade for IPCC AR5 (Intergovernmental Panel on Climate Change, the Fifth Assessment Report) RCP (Representative Concentration Pathway) 8.5, RCP6.0, RCP4.5 and RCP2.6 scenarios, respectively. And the EFs will finally increase by 17.6 %, 9.6 %, 5.6 % and 4.0 % at the end of this century. Additionally, the derived relative changes in China also show high heterogeneity and indicate large uncertainty in projecting future national total CH₄ emissions. Hence, we strongly suggest temperature-dependent EFs and the positive feedback between global warming and CH₄ emissions should be considered in future CH₄ emission projections and climate change models.

Urban areas are global methane (CH₄) hotspots. Yet large uncertainties still remain for the CH₄ budget of these domains. The Yangtze River Delta (YRD), China, is one of the world's most densely populated regions where a large number of cities are located. To estimate anthropogenic CH₄ emissions in YRD, we conducted simultaneous atmospheric CH₄ and CO₂ mixing ratio measurements from June 2010 to April 2011. By combining these measurements with the Weather Research and Forecasting and Stochastic Time‐Inverted Lagrangian Transport models and a priori Emission Database for Global Atmospheric Research emission inventories, we applied three “top‐down” approaches to constrain anthropogenic CH₄ emissions. These three approaches included multiplicative scaling factors, flux ratio, and scale factor Bayesian inversion. The posteriori CH₄ flux density estimated from the three approaches showed high consistency and were 36.32 (±9.17), 35.66 (±2.92), and 36.03(±14.25) nmol·m⁻²·s⁻¹, respectively, for the duration of the study period (November 2010 to April 2011). The total annual anthropogenic CH₄ emission was 6.52(±1.59) Tg for the YRD region based on the average of these three approaches. Our emission estimates were 30.2(±17.6)%, 31.5 (±5.6)%, and 30.8 (±27.4)% lower than the a priori Emission Database for Global Atmospheric Research v432 emission inventory estimate. The scale factor Bayesian inversion results indicate that the overestimate was mainly caused by two source categories including fuel exploitation and agricultural soil emissions (rice cultivation). The posteriori flux densities for agricultural soil and fuel exploitation were 10.68 and 6.34 nmol·m⁻²·s⁻¹, respectively, and were 47.8% and 29.2% lower than the a priori inventory. Agricultural soil was the largest source contribution and accounted for 29.6% of the YRD CH₄ budget during the study period.

The methane (CH₄) budget and its source partitioning are poorly constrained in the Midwestern United States. We used tall tower (185 m) aerodynamic flux measurements and atmospheric scale factor Bayesian inversions to constrain the monthly budget and to partition the total budget into natural (e.g., wetlands) and anthropogenic (e.g., livestock, waste, and natural gas) sources for the period June 2016 to September 2017. Aerodynamic flux observations indicated that the landscape was a CH₄ source with a mean annual CH₄ flux of +13.7 ± 0.34 nmol m⁻² s⁻¹ and was rarely a net sink. The scale factor Bayesian inversion analyses revealed a mean annual source of +12.3 ± 2.1 nmol m⁻² s⁻¹. Flux partitioning revealed that the anthropogenic source (7.8 ± 1.6 Tg CH₄ yr⁻¹) was 1.5 times greater than the bottom‐up gridded United States Environmental Protection Agency inventory, in which livestock and oil/gas sources were underestimated by 1.8‐fold and 1.3‐fold, respectively. Wetland emissions (4.0 ± 1.2 Tg CH₄ yr⁻¹) were the second largest source, accounting for 34% of the total budget. The temporal variability of total CH₄ emissions was dominated by wetlands with peak emissions occurring in August. In contrast, emissions from oil/gas and other anthropogenic sources showed relatively weak seasonality.