Search

Atmospheric methane’s rapid growth from 2006 to the present is unprecedented in the observational record. Isotopic evidence implies the growth is mainly driven by an increase in biogenically-sourced emissions, both from wetlands and ruminants, and waste. A significant part of methane’s current rise may come not from direct anthropogenic emissions and land use changes, but rather from a combination of natural biogenic feedback responses, occurring in response to the anthropogenic forcing. Although microbial emissions from agricultural and waste have increased between 2006-2020 by about 35 Tg/yr, perhaps 35-40 Tg/yr of the recent net growth in methane emissions may have been driven by natural biogenic processes, especially wetland feedbacks to climate change. Modelling comparison between the biogenic component of methane’s growth and isotopic shift in the 15 years from 2007-2022, and the global-scale climate reorganisations during the transitions from glacial to interglacial periods in the Pleistocene, shows that the modern growth event is comparable to or greater than the scale and speed of methane’s growth and isotopic shift during past glacial/interglacial termination events. It remains possible that current changes are related to decadal- or centennial-scale variability in precipitation and temperature and remain within the range of Holocene variability, or due to direct anthropogenic actions. But, though any current transition will differ greatly from the past glacial-interglacial changes, it is also possible methane’s remarkable growth and isotopic shift that began in 2006 may be a first indicator that a very large-scale reorganisation of the natural climate and biosphere system is under way.

Understanding the sources and sinks of methane (CH4) is critical to both predicting and mitigating future climate change. There are large uncertainties in the global budget of atmospheric CH4, but natural emissions are estimated to be of a similar magnitude to anthropogenic emissions. To understand CH4 flux from biogenic sources in the United States (US) of America, a multi-scale CH4 observation network focused on CH4 flux rates, processes, and scaling methods is required. This can be achieved with a network of ground-based observations that are distributed based on climatic regions and land cover. To determine the gaps in physical infrastructure for developing this network, we need to understand the landscape representativeness of the current infrastructure. We focus here on eddy covariance (EC) flux towers because they are essential for a bottom-up framework that bridges the gap between point-based chamber measurements and airborne or satellite platforms that inform policy decisions and global climate agreements. Using dissimilarity, multidimensional scaling, and cluster analysis, the US was divided into 10 clusters distributed across temperature and precipitation gradients. We evaluated dissimilarity within each cluster for research sites with active CH4 EC towers to identify gaps in existing infrastructure that limit our ability to constrain the contribution of US biogenic CH4 emissions to the global budget. Through our analysis using climate, land cover, and location variables, we identified priority areas for research infrastructure to provide a more complete understanding of the CH4 flux potential of ecosystem types across the US. Clusters corresponding to Alaska and the Rocky Mountains, which are inherently difficult to capture, are the most poorly represented, and all clusters require a greater representation of vegetation types.

Methane emissions from organic-rich soils in the Arctic have been extensively studied due to their potential to increase the atmospheric methane burden as permafrost thaws1–3. However, this methane source might have been overestimated without considering high-affinity methanotrophs (HAMs; methane-oxidizing bacteria) recently identified in Arctic mineral soils4–7. Herein we find that integrating the dynamics of HAMs and methanogens into a biogeochemistry model8–10 that includes permafrost soil organic carbon dynamics3 leads to the upland methane sink doubling (~5.5 Tg CH4 yr−1) north of 50 °N in simulations from 2000–2016. The increase is equivalent to at least half of the difference in net methane emissions estimated between process-based models and observation-based inversions11,12, and the revised estimates better match site-level and regional observations5,7,13–15. The new model projects doubled wetland methane emissions between 2017–2100 due to more accessible permafrost carbon16–18. However, most of the increase in wetland emissions is offset by a concordant increase in the upland sink, leading to only an 18% increase in net methane emission (from 29 to 35 Tg CH4 yr−1). The projected net methane emissions may decrease further due to different physiological responses between HAMs and methanogens in response to increasing temperature19,20. Models overestimate Arctic methane emissions compared to observations. Incorporating microbial dynamics into biogeochemistry models helps reconcile this discrepancy; high-affinity methanotrophs are an important part of the Arctic methane budget and double previous estimates of methane sinks.

We have constructed an atmospheric inversion framework based on TM5-4DVAR to jointly assimilate measurements of methane and δ13C of methane in order to estimate source-specific methane emissions. Here we present global emission estimates from this framework for the period 1999–2016. We assimilate a newly constructed, multi-agency database of CH4 and δ13C measurements. We find that traditional CH4-only atmospheric inversions are unlikely to estimate emissions consistent with atmospheric δ13C data, and assimilating δ13C data is necessary to derive emissions consistent with both measurements. Our framework attributes ca. 85 % of the post-2007 growth in atmospheric methane to microbial sources, with about half of that coming from the tropics between 23.5∘ N and 23.5∘ S. This contradicts the attribution of the recent growth in the methane budget of the Global Carbon Project (GCP). We find that the GCP attribution is only consistent with our top-down estimate in the absence of δ13C data. We find that at global and continental scales, δ13C data can separate microbial from fossil methane emissions much better than CH4 data alone, and at smaller scales this ability is limited by the current δ13C measurement coverage. Finally, we find that the largest uncertainty in using δ13C data to separate different methane source types comes from our knowledge of atmospheric chemistry, specifically the distribution of tropospheric chlorine and the isotopic discrimination of the methane sink.

Understanding the sources and sinks of CH4 is critical to both predicting and mitigating future climate change. There are large uncertainties in the global budget of atmospheric CH4, but natural emissions are estimated to be of a similar magnitude to total anthropogenic emissions. The largest sources of uncertainty in scaling bottom-up CH4 estimates stem from limited ground-based measurements and the misalignment between drivers of CH4 fluxes and current land use classifications. To understand the CH4 flux potential of natural ecosystems and agricultural lands in the United States (US) of America, a multi-scale CH4 observation network focused on CH4 flux rates, processes, and scaling methods is required. This can be achieved with a network of ground-based observations that are distributed based on climatic regions and landcover. To determine the gaps in physical infrastructure for developing this network, we need to understand the representativeness of current measurements. We focus here on eddy covariance (EC) flux towers because they are essential for a bottom-up framework that bridges the gap between point-based chamber measurements and airborne or satellite platforms, informing the remote sensing and modelling communities and policy decisions, all the way to IPCC reports. Using multidimensional scaling and a cluster analysis, the US was divided into 10 clusters that were distributed across temperature and wetness gradients. We evaluated the distance to the medoid condition within each cluster for research sites with EC tower infrastructure to identify the gaps in existing infrastructure that limit our ability to constrain the contribution of US biogenic CH4 emissions to the global budget. These gaps occurred across all EC flux tower networks and independently managed sites as well as in some environmental clusters. Through our analysis using climate, land cover, and location variables, we have identified priority areas to target for research infrastructure to provide a more complete understanding of the CH4 flux potential of ecosystem types across the US.

We study the drivers behind the global atmospheric methane (CH4) increase observed after 2006. Candidate emission and sink scenarios are constructed based on proposed hypotheses in the literature. These scenarios are simulated in the TM5 tracer transport model for 1984–2016 to produce three‐dimensional fields of CH4 and δ 13C‐CH4, which are compared with observations to test the competing hypotheses in the literature in one common model framework. We find that the fossil fuel (FF) CH4 emission trend from the Emissions Database for Global Atmospheric Research 4.3.2 inventory does not agree with observed δ 13C‐CH4. Increased FF CH4 emissions are unlikely to be the dominant driver for the post‐2006 global CH4 increase despite the possibility for a small FF emission increase. We also find that a significant decrease in the abundance of hydroxyl radicals (OH) cannot explain the post‐2006 global CH4 increase since it does not track the observed decrease in global mean δ 13C‐CH4. Different CH4 sinks have different fractionation factors for δ 13C‐CH4, thus we can investigate the uncertainty introduced by the reaction of CH4 with tropospheric chlorine (Cl), a CH4 sink whose abundance, spatial distribution, and temporal changes remain uncertain. Our results show that including or excluding tropospheric Cl as a 13 Tg/year CH4 sink in our model changes the magnitude of estimated fossil emissions by ∼20%. We also found that by using different wetland emissions based on a static versus a dynamic wetland area map, the partitioning between FF and microbial sources differs by 20 Tg/year, ∼12% of estimated fossil emissions., Key Points Increased fossil fuel emissions are unlikely the dominant driver for post‐2006 global CH4 increaseA significant decrease in the abundance of hydroxyl radicals (OH) cannot explain the post‐2006 global CH4 increaseCH4 emission attributions are sensitive to the uncertainties in OH fractionation, tropospheric Cl sink, and wetland areas

In the face of ongoing and projected precipitation changes, precipitation manipulation experiments (PMEs) have produced a wealth of data about the effects of precipitation changes on soils. In response, researchers have undertaken a number of synthetic efforts. Several meta-analyses have been conducted, each revealing new aspects of soil responses to precipitation changes. We synthesize the findings of 16 meta-analyses focused on the effects of decreased and increased precipitation on 42 soil response variables, covering a wide range of soil processes and examining responses of individual variables as well as more integrative responses of carbon and nitrogen cycles. We found a strong agreement among meta-analyses that decreased and increased precipitation inhibits and promotes belowground carbon and nitrogen cycling, respectively, while microbial communities are relatively resistant to precipitation changes. Much attention has been paid to fluxes and pools in carbon, nitrogen, and phosphorus cycles, such as gas emissions, soil carbon, soil phosphorus, extractable nitrogen ions, and biomass, but the rates of processes underlying these variables are less frequently covered in meta-analytic studies (e.g., rates of mineralization, fixation, and de/nitrification). Shifting scientific attention to these “processes” would, therefore, deepen the current understanding of the effects of precipitation changes on soil and provide new insights. By comparing meta-analyses focused on different variables, we provide here a quantitative and holistic view of soil responses to changes in precipitation.

This thesis is a collection of three research articles to quantify carbon fluxes and isotopic signature changes across global terrestrial ecosystems. Chapter 2, the first article of this thesis, focuses on the importance of an under-estimated methane soil sink for contemporary and future methane budgets in the pan-Arctic region. Methane emissions from organic-rich soils in the Arctic have been extensively studied due to their potential to increase the atmospheric methane burden as permafrost thaws. However, this methane source might have been overestimated without considering high affinity methanotrophs (HAM, methane oxidizing bacteria) recently identified in Arctic mineral soils. From this study, we find that HAM dynamics double the upland methane sink (~5.5 TgCH4yr-1) north of 50°N in simulations from 2000 to 2016 by integrating the dynamics of HAM and methanogens into a biogeochemistry model that includes permafrost soil organic carbon (SOC) dynamics. The increase is equivalent to at least half of the difference in net methane emissions estimated between process-based models and observation-based inversions, and the revised estimates better match site-level and regional observations. The new model projects double wetland methane emissions between 2017-2100 due to more accessible permafrost carbon. However, most of the increase in wetland emissions is offset by a concordant increase in the upland sink, leading to only an 18% increase in net methane emission (from 29 to 35 TgCH4yr-1). The projected net methane emissions may decrease further due to different physiological responses between HAM and methanogens in response to increasing temperature. This article was published in Nature Climate Change in March 2020. In Chapter 3, the second article of this thesis, I develop and validate the first biogeochemistry model to simulate carbon isotopic signatures (δ13C) of methane emitted from global wetlands, and examined the importance of the wetland carbon isotope map for studying the global methane cycle. I incorporated a carbon isotope-enabled module into an extant biogeochemistry model to mechanistically simulate the spatial and temporal variability of global wetland δ13C-CH4. The new model explicitly considers isotopic fractionation during methane production, oxidation, and transport processes. I estimate a mean global wetland δ13C-CH4 of -60.78‰ with its seasonal and inter-annual variability. I find that the new model matches field chamber observations 35% better in terms of root mean square estimates compared to an empirical static wetland δ13C-CH4 map. The model also reasonably reproduces the regional heterogeneity of wetland δ13C-CH4 in Alaska, consistent with vertical profiles of δ13C-CH4 from NOAA aircraft measurements. Furthermore, I show that the latitudinal gradient of atmospheric δ13C-CH4 simulated by a chemical transport model using the new wetland δ13C-CH4 map reproduces the observed latitudinal gradient based on NOAA/INSTAAR global flask-air measurements. I believe this study is the first process-based biogeochemistry model to map the global distribution of wetland δ13C-CH4, which will significantly help atmospheric chemistry transport models partition global methane emissions. This article is in preparation for submission to Nature Geoscience. Chapter 4 of this thesis, the third article, investigates the importance of leaf carbon allocation for seasonal leaf carbon isotopic signature changes and water use efficiency in temperate forests. Temperate deciduous trees remobilize stored carbon early in the growing season to produce new leaves and xylem vessels. The use of remobilized carbon for building leaf tissue dampens the link between environmental stomatal response and inferred intrinsic water use efficiency (iWUE) using leaf carbon isotopic signatures (δ13C). So far, few studies consider carbon allocation processes in interpreting leaf δ13C signals. To understand effects of carbon allocation on δ13C and iWUE estimates, we analyzed and modeled the seasonal leaf δ13C of four temperate deciduous species (Acer saccharum, Liriodendron tulipifera, Sassafras albidum, and Quercus alba) and compared the iWUE estimates from different methods, species, and drought conditions. At the start of the growing season, leaf δ13C values were more enriched, due to remobilized carbon during leaf-out. The bias towards enriched leaf δ13C values explains the higher iWUE from leaf isotopic methods compared with iWUE from leaf gas exchange measurements. I further showed that the discrepancy of iWUE estimates between methods may be species-specific and drought sensitive. The use of δ13C of plant tissues as a proxy for stomatal response to environmental processes, through iWUE, is complicated due to carbon allocation and care must be taken when interpreting estimates to avoid proxy bias. This article is in review for publication in New Phytologist.

Recent field studies have documented a surprisingly strong and consistent methane sink in arctic mineral soils, thought to be due to high-affinity methanotrophy. However, the distinctive physiology of these methanotrophs is poorly represented in mechanistic methane models. We developed a new model, constrained by microcosm experiments, to simulate the activity of high-affinity methanotrophs. The model was tested against soil core-thawing experiments and field-based measurements of methane fluxes and was compared to conventional mechanistic methane models. Our simulations show that high-affinity methanotrophy can be an important component of the net methane flux from arctic mineral soils. Simulations without this process overestimate methane emissions. Furthermore, simulations of methane flux seasonality are improved by dynamic simulation of active microbial biomass. Because a large fraction of the Arctic is characterized by mineral soils, high-affinity methanotrophy will likely have a strong effect on its net methane flux.

Metatranscriptomics has recently been applied to investigate the active biogeochemical processes and elemental cycles, and in situ responses of microbiomes to environmental stimuli and stress factors. De novo assembly of RNA-Sequencing (RNA-Seq) data can reveal a more detailed description of the metabolic interactions amongst the active microbial communities. However, the quality of the assemblies and the depiction of the metabolic network provided by various de novo assemblers have not yet been thoroughly assessed. In this study, we compared 15 de novo metatranscriptomic assemblies for a fracture fluid sample collected from a borehole located at 1.34 km below land surface in a South African gold mine. These assemblies were constructed from total, non-coding, and coding reads using five de novo transcriptomic assemblers (Trans-ABySS, Trinity, Oases, IDBA-tran, and Rockhopper). They were evaluated based on the number of transcripts, transcript length, range of transcript coverage, continuity, percentage of transcripts with confident annotation assignments, as well as taxonomic and functional diversity patterns. The results showed that these parameters varied considerably among the assemblies, with Trans-ABySS and Trinity generating the best assemblies for non-coding and coding RNA reads, respectively, because the high number of transcripts assembled covered a wide expression range, and captured extensively the taxonomic and metabolic gene diversity, respectively. We concluded that the choice of de novo transcriptomic assemblers impacts substantially the taxonomic and functional compositions. Care should be taken to obtain high-quality assemblies for informing the in situ metabolic landscape.

Subsurface lithoautotrophic microbial ecosystems (SLiMEs) under oligotrophic conditions are typically supported by H₂. Methanogens and sulfate reducers, and the respective energy processes, are thought to be the dominant players and have been the research foci. Recent investigations showed that, in some deep, fluid-filled fractures in the Witwatersrand Basin, South Africa, methanogens contribute