Andrew S. Reeve

Simulating vertical flow in large peatlands.

Ground-water flow simulations are used to evaluate the importance of three parameters on vertical flow in peatlands: regional slope, permeability of the mineral soil underlying the peat, and peatland topography. Our results indicate that the extent of vertical ground-water flow in peatlands is primarily controlled by mineral soil permeability. Local ground-water flow cells that form under small water-table mounds at bog domes can drive peat pore water into permeable mineral soil. However, when peat forms over low permeability mineral soil, vertical movement of peat pore water becomes negligible and lateral flow of water in the upper portion of the peat column dominates the peatland hydrology. This effect is not due to the low permeability of humified peat layers, as is commonly assumed in many peatland studies. Field data from the Hudson Bay Lowland (Canada) and the Glacial Lake Agassiz peatlands of northern Minnesota confirm the validity of these models. These results are relevant in settings other than peatlands where relief is low and small topographic mounds exist.

Investigating peatland stratigraphy and hydrogeology using integrated electrical geophysics

Hydrology has been suggested as the mechanism controlling vegetation and related surficial pore-water chemistry in large peatlands. Peatland hydrology influences the carbon dynamics within these large carbon reservoirs and will influence their response to global warming. A geophysical survey was completed in Caribou Bog, a large peatland in Maine, to evaluate peatland stratigraphy and hydrology. Geophysical measurements were integrated with direct measurements of peat stratigraphy from probing, fluid chemistry, and vegetation patterns in the peatland. Consistent with previous field studies, ground-penetrating radar (GPR) was an excellent method for delineating peatland stratigraphy. Prominent reflectors from the peat-lake sediment and lake sediment-mineral soil contacts were precisely recorded up to 8 m deep. Two-dimensional resistivity and induced polarization imaging were used to investigate stratigraphy beneath the mineral soil, beyond the range of GPR. We observe that the peat is chargeable, and that IP imaging is an alternative method for defining peat thickness. The chargeability of peat is attributed to the high surface-charge density on partially decomposed organic matter. The electrical conductivity imaging resolved glaciomarine sediment thickness (a confining layer) and its variability across the basin. Comparison of the bulk conductivity images with peatland vegetation revealed a correlation between confining layer thickness and dominant vegetation type, suggesting that stratigraphy exerts a control on hydrogeology and vegetation distribution within this peatland. Terrain conductivity measured with a Geonics EM31 meter correlated with confining glaciomarine sediment thickness and was an effective method for estimating variability in glaciomarine sediment thickness over approximately 18 km 2 . Our understanding of the hydrogeology, stratigraphy, and controls on vegetation growth in this peatland was much enhanced from the geophysical study.

Simulating dispersive mixing in large peatlands

Numerical simulations indicate that mechanical dispersive mixing can be the dominant mass transport mechanism in large peatlands. Dispersive mixing driven by lateral flow can drive solute fluxes from the mineral soil upward to the peat surface and thereby explain observed patterns of bog and fen in large peatlands. Longitudinal and transverse dispersivities of only 0.5 and 0.05 m, respectively, were sufficient to supply solutes to the peat surface in the absence of upward ground-water flow. Incorporation of hydrodynamic dispersion in peatland systems explains apparent contradictions in solute migration in peatlands, allowing the simultaneous downward flux of labile carbon (i.e. root exudates) produced at the peat surface and upward migration of inorganic solutes from the underlying mineral soil. Previous models of peatland hydrogeochemistry that rely on advection alone as the dominant process for solute transport may therefore be inadequate to explain fully the hydrology, geochemistry, and evolution of large peatlands.

Dynamics of methane ebullition from a peat monolith revealed from a dynamic flux chamber system

Methane (CH4) ebullition in northern peatlands is poorly quantified in part due to its high spatiotemporal variability. In this study, a dynamic flux chamber (DFC) system was used to continuously measure CH4 fluxes from a monolith of near-surface Sphagnum peat at the laboratory scale to understand the complex behavior of CH4 ebullition. Coincident transmission ground penetrating radar measurements of gas content were also acquired at three depths within the monolith. A graphical method was developed to separate diffusion, steady ebullition, and episodic ebullition fluxes from the total CH4 flux recorded and to identify the timing and CH4 content of individual ebullition events. The results show that the application of the DFC had minimal disturbance on air-peat CH4 exchange and estimated ebullition fluxes were not sensitive to the uncertainties associated with the graphical model. Steady and episodic ebullition fluxes were estimated to be averagely 36 ± 24% and 38 ± 24% of the total fluxes over the study period, respectively. The coupling between episodic CH4 ebullition and gas content within the three layers supports the existence of a threshold gas content regulating CH4 ebullition. However, the threshold at which active ebullition commenced varied between peat layers with a larger threshold (0.14 m3 m−3) observed in the deeper layers, suggesting that the peat physical structure controls gas bubble dynamics in peat. Temperature variation (23°C to 27°C) was likely only responsible for small episodic ebullition events from the upper peat layer, while large ebullition events from the deeper layers were most likely triggered by drops in atmospheric pressure.

The Influence of Permeable Mineral Lenses on Peatland Hydrology

Cross-sectional computer models were created that incorporated different high permeability zones to explore the potential role of eskers and similar geologic units associated with peatlands on the hydrology of these systems. These computer simulations indicate that small isolated lenses of high permeability material will locally distort the flow field, shifting flow patterns and creating new discharge and/ or recharge zones. The simulation with the greatest hydraulic connectivity between the peat dome and peatland lagg displays widespread and continuous-with-depth downward flow beneath the bog dome, consistent with field observations. These simulations were compared with field data collected from a Maine (USA) peatland, and they suggest an esker identified within this peatland distorts the hydraulic gradients and resulting flow patterns within the peat. Similar subsurface features may be important in other peatland systems.

Climatic drivers for multidecadal shifts in solute transport and methane production zones within a large peat basin

Northern peatlands are an important source for greenhouse gases but their capacity to produce methane remains uncertain under changing climatic conditions. We therefore analyzed a 43-year time series of pore-water chemistry to determine if long-term shifts in precipitation altered the vertical transport of solutes within a large peat basin in northern Minnesota. These data suggest that rates of methane production can be finely tuned to multi-decadal shifts in precipitation that drive the vertical penetration of labile carbon substrates within the Glacial Lake Agassiz Peatlands. Tritium and cation profiles demonstrate that only the upper meter of these peat deposits was flushed by downwardly moving recharge from 1965 through 1983 during a Transitional Dry-to-Moist Period. However, a shift to a moister climate after 1984 drove surface waters much deeper, largely flushing the pore waters of all bogs and fens to depths of 2 m. Labile carbon compounds were transported downward from the rhizosphere to the basal peat at this time producing a substantial enrichment of methane in Δ14C with respect to the solid-phase peat from 1991 to 2008. These data indicate that labile carbon substrates can fuel deep production zones of methanogenesis that more than doubled in thickness across this large peat basin after 1984. Moreover, the entire peat profile apparently has the capacity to produce methane from labile carbon substrates depending on climate-driven modes of solute transport. Future changes in precipitation may therefore play a central role in determining the source strength of peatlands in the global methane cycle.

Wetland Connections: Linking University Researchers and High School Teachers to Advance Science Education and Wetland Conservation

Wetland Connections is a geosciences and biology field-based project connecting students to real-world problem solving. The broad goal was to link University faculty and students with high school teachers and students to conduct field research on Maines wooded wetlands. Three university faculty, each from a different discipline, shared skills with teachers from three high schools as they investigated shrub and graminoid peatlands, mineral-soil forested wetlands, and vernal pools. Building the infrastructure for a long-term, wetland monitoring program required an integrated systems approach. Students collected environmental data to determine how wetland hydrology and water chemistry related to vegetation composition and structure, wildlife habitat, and geochemical gradients. Data were collected on nine wetlands with 178 students, teachers, community members, and University faculty and students directly participating. Data and information on each of the wetland types is being assembled on an interactive web site. Teachers reported the project helped them meet educational standards, inspired students to pursue further studies in the sciences, and provided baseline data on local, previously unstudied wetlands.

Automated inverse computer modeling of borehole flow data in heterogeneous aquifers

A computer model has been developed to simulate borehole flow in heterogeneous aquifers where the vertical distribution of permeability may vary significantly. In crystalline fractured aquifers, flow into or out of a borehole occurs at discrete locations of fracture intersection. Under these circumstances, flow simulations are defined by independent variables of transmissivity and far-field heads for each flow contributing fracture intersecting the borehole. The computer program, ADUCK (A Downhole Underwater Computational Kit), was developed to automatically calibrate model simulations to collected flowmeter data providing an inverse solution to fracture transmissivity and far-field head. ADUCK has been tested in variable borehole flow scenarios, and converges to reasonable solutions in each scenario. The computer program has been created using open-source software to make the ADUCK model widely available to anyone who could benefit from its utility.

Free phase gas processes in a northern peatland inferred from autonomous field‐scale resistivity imaging

The mechanisms that control free phase gas (FPG) dynamics within peatlands, and therefore estimates of past, present, and future gas fluxes to the atmosphere remain unclear. Electrical resistivity imaging (ERI) is capable of autonomously collecting three-dimensional data on the centimeter to tens of meter scale and thus provides a unique opportunity to observe FPG dynamics in situ. We collected 127 3-D ERI data sets as well as water level, soil temperature, atmospheric pressure, and limited methane flux data at a site in a northern peatland over the period July–August 2013 to improve the understanding of mechanisms controlling gas releases at a hitherto uncaptured field scale. Our results show the ability of ERI to image the spatial distribution of gas accumulation and infer dynamics of gas migration through the peat column at high (i.e., hourly) temporal resolution. Furthermore, the method provides insights into the role of certain mechanisms previously associated with the triggering of FPG releases such as drops in atmospheric pressure. During these events, buoyancy-driven gas release primarily occurs in shallow peat as proposed by the “shallow peat model.” Releases from the deeper peat are impeded by confining layers, and we observed a large loss of FPG in deep peat that may likely represent a rupture event, where accumulated FPG escaped the confining layer as suggested by the “deep peat model.” Negative linear correlations between water table elevation and resistivity result from hydrostatic pressure regulating bubble volume, although these variations did not appear to trigger FPG transfer or release.

Improving Understanding of Peatland Hydrogeology Using Electrical Geophysics

A geophysical survey was completed in Caribou Bog, a large peatland in Maine, to evaluate peatland stratigraphy and hydrology. Geophysical measurements were integrated with direct measurements of peat stratigraphy from probing and with measurements of fluid chemistry. Consistent with previous field studies, GPR was an excellent method for delineating peatland stratigraphy. Prominent reflectors from the peat-lake sediment and lake sediment-mineral soil contacts were precisely recorded up to 8 m deep. However, GPR provided no information below the mineral soil contact. 2D resistivity and induced polarization (IP) imaging was used to further investigate the stratigraphy of this peat basin. We observe that the peat is chargeable and that IP imaging is an alternative method for defining peat thickness. This chargeability is attributed to the high surface charge density on partially decomposed organic matter. The conductivity imaging resolved glaciomarine sediment thickness and its variability across the basin. Terrain conductivity measured with a Geonics EM31 correlated with glaciomarine sediment thickness and was effective in characterizing variability in layer thickness over approximately 18 km. The electrical imaging indicates that variations in glaciomarine sediment thickness may exert a key control on the hydrogeology and vegetation distribution within this peatland. Introduction Geophysical methods can assist understanding of peatland stratigraphy and hydrogeology. The ground penetrating radar (GPR) method has been most extensively used. Depending on the electrical conductivity of the peat, GPR can penetrate up to 10 m in peatlands, with a resolution of 10-15 cm (Lowe, 1985; Theimer et al., 1994). The method is effective as moisture content changes occur at important interfaces, causing measurable GPR reflections. Numerous studies illustrate the potential of the method for identifying the base of a peatland (Warner et al., 1990; Pelletier et al., 1991; Poole et al., 1997). Peat can contain up to about 95% water, with water content varying with degree of decomposition and the plant types that make up the peat (Hobbs, 1986). Moisture content typically drops to 30-40 % in the mineral soil, resulting in large-amplitude reflections at the peat-mineral soil contact (Theimer et al., 1994). Significant reflectors within peat have also been identified and associated with local changes in moisture content (Theimer et al., 1994). Reflections from boundaries between different types of peat and the interface between peat and organic-rich lake sediment are also identifiable (Hanninen, 1992). As the dielectric constant of peat is well known (50-70 depending on peat type, Theimer et al., 1994), reliable estimates of the depth to reflectors within and at the base of peat are obtainable. Dc resistivity, EM and induced polarization (IP) methods may assist peatland studies, particularly for examining the relation between mineral soil stratigraphy and properties of peat. Unlike GPR, these methods are not typically limited to studies above the mineral soil. Bulk conductivity (σb) measured in the dc resistivity method depends upon fluid conductivity (σw), moisture content (θ) and 2 surface conduction (σs). As peat is predominantly water, σb is particularly dependent on σw. The shallow pore waters in boreal peatlands are typically relatively dilute. The electrical conductivity of peat pore water usually increases with depth because the mineral soil underlying the peat is a source of inorganic solutes. As decomposed plant material within peat has a high surface charge, surface conduction is probably also significant in controlling bulk conductivity. The IP method measures the magnitude of polarization of a material or, put simply, its ability to store charge. In nonmetallic mineral soil, polarization results from diffusion controlled polarization processes at the interface between the grain surfaces and the pore solution. The IP response thus depends on surface chemistry, which is controlled by charge density, surface area and fluid chemistry. The large surface area and cation exchange capacity (CEC) associated with clay minerals enhances the magnitude of polarization in clay-disseminated sediments and rocks (Vinegar and Waxman, 1984). The application of IP to the study of peatlands has never, to our knowledge, been reported and typical values of M in peat are not documented. The high surface charge density associated with poorly decomposed organic material results in a high CEC (Hobbs, 1986). As charge density is a major control on IP response, we expect the IP response of peat to be significantly different from the mineral soil. In this paper, we report preliminary results of a field study to investigate the utility of electrical methods (resistivity imaging, IP imaging, GPR and terrain conductivity mapping) for understanding the stratigraphy and hydrogeologic setting of a large peatland in Maine. The full findings of the study will be presented in a later paper. Caribou Bog Caribou Bog is a large peatland extending 17 km south to north, from Bangor to Hudson, Maine (Figure 1). Peat thickness reaches 13 m, with as much as 5 m of underlying organic-rich lake sediment. Nine monitoring well clusters are installed across the southern complex. Hydraulic head measurements at these wells indicate that groundwater generally flows to the northwest, with a slight water table mound present near W3 and W4. Vertical groundwater flow changes seasonally; groundwater flows upward from the mineral soil in the spring and downward in the summer. The location of geophysical profiling lines within the study area is shown in Figure 2. Line 0 was established as the reference profile along the long axis of the basin. Six transects were established at monitoring well locations or at the approximate midpoint between them. Samples of peat and mineral soil were obtained at four locations (Figure 2). Loosely compacted, peat forms the major unit within the basin, reaching 5.6 m thick at C2. It is underlain by organic-rich lake sediment, reaching 1.5 m thick at C2. A distinct silt band separates the peat and lake sediment at C1, C2 and C3. The mineral basement is a glaciomarine silt-clay, the Presumpscot formation. Penetration tests were performed along an 80 m section of Line 0 to determine small scale variability in peat thickness. A pointed steel rod was pushed into the peat until refusal, to identify the glaciomarine sediment contact. Fluid conductivity (σw) in W1-W6 was measured at a depth of about 1 m, at one or two intermediate depths, and at the peat-mineral soil interface. Additional measurements in the upper 1.5 m were made at points along Line 0.. Conductivity within the peat varies from 40-77 μS/cm. Such low σw are indicative of surficial waters in northern raised peatlands that contain peat with little soluble material and are hydrologically isolated from external solute-rich sources of water (Siegel and Glaser, 1987). At Well 3 σw increases from 50-70 μS/cm in the peat to 150 μS/cm in the lake sediment, indicative of solute exchange from the mineral basement.