Upscaling Wetland Methane Emissions From the FLUXNET‐CH4 Eddy Covariance Network (UpCH4 v1.0): Model Development, Network Assessment, and Budget Comparison

Abstract

Wetlands are responsible for 20%–31% of global methane (CH4) emissions and account for a large source of uncertainty in the global CH4 budget. Data‐driven upscaling of CH4 fluxes from eddy covariance measurements can provide new and independent bottom‐up estimates of wetland CH4 emissions. Here, we develop a six‐predictor random forest upscaling model (UpCH4), trained on 119 site‐years of eddy covariance CH4 flux data from 43 freshwater wetland sites in the FLUXNET‐CH4 Community Product. Network patterns in site‐level annual means and mean seasonal cycles of CH4 fluxes were reproduced accurately in tundra, boreal, and temperate regions (Nash‐Sutcliffe Efficiency ∼0.52–0.63 and 0.53). UpCH4 estimated annual global wetland CH4 emissions of 146 ± 43 TgCH4 y−1 for 2001–2018 which agrees closely with current bottom‐up land surface models (102–181 TgCH4 y−1) and overlaps with top‐down atmospheric inversion models (155–200 TgCH4 y−1). However, UpCH4 diverged from both types of models in the spatial pattern and seasonal dynamics of tropical wetland emissions. We conclude that upscaling of eddy covariance CH4 fluxes has the potential to produce realistic extra‐tropical wetland CH4 emissions estimates which will improve with more flux data. To reduce uncertainty in upscaled estimates, researchers could prioritize new wetland flux sites along humid‐to‐arid tropical climate gradients, from major rainforest basins (Congo, Amazon, and SE Asia), into monsoon (Bangladesh and India) and savannah regions (African Sahel) and be paired with improved knowledge of wetland extent seasonal dynamics in these regions. The monthly wetland methane products gridded at 0.25° from UpCH4 are available via ORNL DAAC (https://doi.org/10.3334/ORNLDAAC/2253).

Recent increases in annual, seasonal, and extreme methane fluxes driven by changes in climate and vegetation in boreal and temperate wetland ecosystems

“…in previous studies (e.g., Chang et al, 2023;McNicol et al, 2023;Ouyang et al, 2023), the use of gap-filled data is an additional source of uncertainty in our predictions.…”

“…In the absence of multidecadal FCH 4 measurement time series, our ML predictions can help unearth the pathways in which flux seasonality and extreme events may have contributed to annual increases over the last four decades. The daily ML predictions underpinning this analysis can investigate these detailed temporal trends and evolutions that cannot be detected with global scale models with coarser timesteps (Chang et al, 2023;McNicol et al, 2023). Days of extreme FCH 4 may play a more vital role under climate change, and their characterization allowed us to detect changes in contributions to total seasonal fluxes over a period of 40 years.…”

Characterizing Performance of Freshwater Wetland Methane Models Across Time Scales at FLUXNET‐CH4 Sites Using Wavelet Analyses

“…In particular, a more advanced intercomparison protocol would help distinguish structural and parameterization limitations by (a) testing multiple parameterization schemes for major wetland processes (e.g., CH 4 production rate and transport); (b) running the models with inputs from FLUXNET-CH 4 local meteorological condition and local site information such as slope, drainage, and vegetation characteristics; and (c) including longer-term records and spatially representative observations with full uncertainty characterization from EC tower measurements. In addition, incorporating wavelet analysis into a more comprehensive framework that includes evaluation of other key variables and machine learning-based estimates (Bansal et al, 2023;McNicol et al, 2023) may help identify the factors influencing its performance at specific time scales more effectively. Modeling global-scale wetland CH 4 emissions is essential for accurately quantifying the contribution of wetland-CH 4 feedback to ongoing climate change within the contemporary global CH 4 budget, given their increasing role as potential contributors to the rise in atmospheric CH 4 concentration in recent years (Peng et al, 2022;Zhang et al, 2023).…”

Permafrost Carbon: Progress on Understanding Stocks and Fluxes Across Northern Terrestrial Ecosystems

“…A few pan-Arctic studies have used statistical and machine learning models to upscale recent or current C fluxes at high spatial resolutions across larger domains or higher temporal resolutions (Jung et al, 2020;McNicol et al, 2023;Natali et al, 2019;Peltola et al, 2019a;Virkkala et al, 2021a). Earlier approaches often used simpler empirical upscaling of flux measurements (e.g., Bartlett & Harriss, 1993).…”