NS24A-01
Measurement of Entrapped Biogenic Gas Bubbles in Northern Peat Soils: Application of Resistivity and X-ray Computed Tomography.
Peatlands are the largest natural source per annum of CH4 emissions to the atmosphere. CH4 is lost from peatlands via diffusion or active transport through vascular plants, and as bubbles moving to the peatland surface - ebullition. The build up and ebullition of biogenic gas bubbles within northern peatlands is spatially variable and depends on the rate of CH4 production, the transport of dissolved CH4 to bubbles through pore water, and the physical properties of the peat. Recent measurements suggest a threshold bubble volume must be reached to trigger episodic or cyclic ebullition, which is assumed to be dependent on peat type. However, this threshold theory lacks a secure physical basis and therefore cannot be applied to simulate methane ebullition from northern peatlands with any confidence. We develop an approach to examine the structural attributes of the peat that cause and promote the trapping and release of bubbles by combining resistivity and X-ray computed tomography (CT). The spatial and temporal variation in the biogenic gas content of peat cores are identified from resistivity measurements. Areas of high and low entrapped gas content are subsequently correlated with the pore structure of the peat samples, characterised using CT. The CT images of the peat structure are vectorised to allow them to be analysed for metrics which relate to the ability of the peat to trap bubbles: e.g. stem length and width, number of branches, angle of branches. Difficulties applying these approaches within northern peatlands are examined. The low pore water conductivity of poorly decomposed near surface peat can hamper resistivity measurements at the laboratory scale, and electrolytic reactions induce the development of artificial gas bubbles. The similarity in linear attenuations between poorly decomposed Sphagnum and pore water also makes the peat structure indistinguishable from the pore water within standard CT scans. The peat samples must, therefore, first be doped with a solution of lead(II) nitrate which is adsorbed by the peat fibres, making them visible.
NS24A-02
Geophysical Characterization of Controls on Biogenic gas realease in the Red Lake Peatland Complex, Northern Minnesota
Recently there has been an increased interest in northern peatlands with respect to their role in the global carbon balance, as they are a net sink of carbon dioxide in the biomass, and a net source of biogenic methane. Methane can store heat roughly 25 times more efficiently than carbon dioxide, making characterization of releases to the atmosphere through both diffusion and ebullition events critical to understanding the global carbon budget. The spatial and temporal heterogeneity of ebullition events make this characterization difficult, and traditional sampling schemes are inadequate due to poor spatial sampling scales, destruction of peat fabric during coring, and difficulty of working in remote ecosystems. Observations of zones of hydraulic overpressure related to free phase gas accumulation forming below confining layers in the peat suggest that peat stratigraphy a key factor controlling the spatial heterogeneity of biogenic gas ebullition. We used electrical geophysical methods to characterize the peat stratigraphy and hydrogeological framework of the Red Lake Peatland Complex in Northern Minnesota, one of the largest (140 km2) and most studied peatlands in North America. This mid-continent forested bog complex is comprised of three major peat landforms, each of which was surveyed using ground penetrating radar (GPR), electrical resistivity, and induced polarization (IP): (1) a raised, ombrotrophic, wooded crest; (2) a sphagnum lawn down slope of the bog crest; (3) a spring fen water track where water flows across the peat surface around ovoid wooded islands. GPR measurements show clearly the peat thickness as well as horizontally continuous internal reflections that indicate the presence of confining layers that may allow for over pressuring zones due to the trapping of free phase biogenic gasses. These results also form a novel data set of a well studied bog complex, offering new insights into the peat structure and hydrogeologic framework and have implications for general models of peatland carbon cycling.
NS24A-03
Growing Peat Bogs in Computers: why, how, and Model Testing With Geophysical Data
Northern peatlands contain up to a third of the world's soil carbon - 455 Pg - which is the equivalent of about 60 percent of the atmospheric carbon store. For a peatland to increase in thickness and/or extent, more carbon must be taken up via photosynthesis and incorporated into plant tissue than is returned to the atmosphere as decay gases (either directly from the peatland or from rivers draining it). Although northern peatlands have built up over many hundreds and thousands of years during the Holocene, there are concerns that they may have a significant net warming effect on global climate over coming decades, as warmer temperatures lead to increased rates of decay and release of carbon dioxide and methane to the atmosphere. To understand how peatlands will respond to a changing climate, and whether there will be large changes in their carbon budget, requires a modelling approach. A problem of some well-established models is that they treat the peatland as an unchanging entity and do not allow for feedbacks between ecological and hydrological processes. For example, it is known that the amount of methane lost to the atmosphere from a bog is closely correlated with vegetation type, yet these models do not simulate changes in the composition and pattern of vegetation on the bog surface. A new approach to modelling peatland development, using the DigiBog suite of models, is described in which such dynamics and interactions are accounted for. Example model predictions of vegetation patterns and near-surface peat hydraulic properties are presented. Although it is possible to test the new models using observations of current patterns of bog vegetation, these patterns cannot be used to indicate temporal changes in the peatland even though such changes are a key part of the models' output. To study temporal changes, it is necessary to look at depth-related variation in peat physical properties. However, traditionally, such variations have only been investigated via the excavation of peat cores or the placement of spatially-discrete instruments, so it has proved difficult to obtain adequate test data sets. For example, obtaining even 100 measurements of the in-situ permeability of peat may take many weeks and provide a data set that can only give a partial indication of the plausibility of a model's predictions of peatland development. The prospect of obtaining continuous 2- and 3- D data on peat properties for testing peatland models using ground-penetrating radar and electrical survey is discussed and the need for further developments in these wetland geophysical methods emphasised.
NS24A-04
Pool patterning in a northern peatland using near surface geophysics: the role of glacial deposition
The potential role of stratigraphy and lithology on the processes leading to pool formation were examined in Caribou Bog, a 2200-hectare peatland in central Maine. The area is surrounded by esker deposit outcrops from the Katahdin system, that extends 150 km from central Maine to the coast (orientated approximately N-S), and is dominated by sharp-crested eskers with poorly sorted sand, gravel and boulders. A combination of hydrogeophysical techniques were used to examine the correlation between pool location within the bog and subsurface stratigraphy and lithology that included: ground penetrating radar (GPR), electrical resistivity, EM-31, hydrological measurements and direct sampling. Previous studies in the area showed certain correspondence between elevated mineral soil surfaces (interpreted as buried eskers) and pool location. The work presented here expands upon our previous results by including a wider array of measurements to better constrain that correspondence. Hydrological measurements showed spatial correlation between stronger downward hydraulic gradients and proximity to the esker crests. Geophysical data along the pool area consistently showed two buried esker crests with dipping and undulating bedding, collapse structures and presence of boulders in sediment. A conceptual model for pool development that accounts for the initial (e.g. heterogeneous peat growth due to local enhanced decomposition) and final stages (e.g. lateral spreading) of pool development is proposed based on these findings.
NS24A-05
Distributed Temperature Sensing (DTS) using optical fiber probes to constrain heat and fluid transport in the subsurface
Conventional methods of capturing densely-spaced in situ subsurface temperature data are limited by the cost, accuracy, and the complexity of employing a network of point measurement devices. Distributed temperature sensing (DTS) using optical fiber is a new technology for providing in situ temperature measurements that are dense in both space and time. We present an application of DTS to record temperature within the hyporheic zone of a gravel bar on the Willamette River, Oregon, USA. Fiber optic probes were constructed by wrapping approximately 100 m of fiber optic cable around 2 m lengths of PVC pipe. The fiber-optic probes were inserted vertically, in wells, into the gravel bar. By integrating the signal every 1.5 m along the length of the fiber optic cable, we were able to sample temperature at 3.5-cm intervals along the length of the pipe and provide continuous time series samples at 5 minute intervals for a total of 7 days. The data will be presented as movies that show propagation of diurnal temperature variations into the subsurface as well as lateral variations in temperature that we tentatively attribute to lateral fluid flow through the gravel bar. The use of DTS probes to measure changes in temperature as a proxy for fluid flow has potential application in a wide variety of fields, including marine geology as well as hydrology.
NS24A-06
Surface Mining and Reclamation Effects on Flood Response of Forested Watersheds in the Central Appalachian Plateau
Surface mining of coal and subsequent reclamation represent the dominant change in land use in the Central Appalachian Plateau (CAP) region of the eastern United States. Much work has been done to quantify the hydrologic impacts of surface mining at the plot scale (10 to 100 ha), but the effects at broader scales (100 to 1000 km2) have not been explored adequately, in part due to the lack of accurate land cover data. Reclaimed mines often have infiltration rates an order of magnitude smaller than forested control areas. However, broad-scale classification of reclaimed sites is difficult as the standing vegetation makes them indistinguishable from alternate land uses such as agriculture or forest. In this work we use a land cover data set that accurately maps surface mines (active and reclaimed) in a 187 km2 watershed within the CAP. These land cover data, as well as plot level data from surface mined areas within the watershed, are used with the Hydrologic Simulation Program- Fortran to simulate changes in flood response due to changes in the proportion of watershed area affected by mining and subsequent reclamation. Results indicate that although reclaimed surface mines may meet all legal requirements for reclamation, increased surface mining tends to push the flood response towards what would be expected for increased urbanization rather than what would be expected for simple changes in vegetative cover. The cause is attributed to the massive soil compaction from heavy machinery used in the surface mine reclamation process.
NS24A-07
Detection of Hydrologic Response at the River Basin Scale Caused by Land Use Change
The 187.5 km2 Georges Creek watershed, located on the Appalachian Plateau in western Maryland (USA), has experienced significant land use change due to surface mining of bituminous coal. We estimate that over 17% of the Georges Creek watershed is being actively surface-mined or was mined and reclaimed previously. The adjacent Savage River watershed (127.2 km2) is completely unaffected by surface mining. Both watersheds have long (>60 year) streamflow records maintained by USGS that were analyzed as part of this project, using Savage River as a control. Temporal analysis of the moments of the flood frequency distributions using a moving-window technique indicated that climatic variability affected both watersheds equally. Normalizing annual maximum flows by antecedent streamflow and causative precipitation allowed trends in the Georges Creek watershed flooding response to become more evident. An analysis of sixteen contemporary warm season storm events based on hourly streamflow and NEXRAD Stage III derived precipitation data provided clear evidence of differences in watershed response to rainfall. Georges Creek events (normalized by basin area and precipitation) are, on average, characterized by slightly greater (7%) peak runoff and shorter (3 hr) centroid lags than Savage River, even though the opposite was expected considering relative basin areas. These differences in stormflow response are most likely attributable to differences in current land use in the basins, particularly the large area of reclaimed minelands in Georges Creek. Interestingly, we found that Georges Creek events produce, on average, only 2/3 of the stormflow volume as Savage River, apparently due to infiltration of water into abandoned deep mine workings and an associated trans-basin drainage system that dates to the early 20th century. Long-term trend analysis at the river basin scale using empirical hydrologic methods is thus complicated by climatic variability and the legacy of deep mining in this system.