[1] High-frequency measurements of water vapor (q) and carbon dioxide (c) concentrations were collected over the course of a transition from dry to wet surface conditions in an agricultural setting on the eastern shore of Virginia. Daytime correlation coefficients between q and c were close to À1 during the dry conditions but became degraded during the wet conditions. An application of wavelet analysis to the highfrequency time series showed that the degraded q-c correlations on the wet day were mainly caused by the influence of large-scale eddies, which introduced positively correlated q-c components to the half-hour time series. Consistent differences in q-c correlation were also observed for smaller eddy scales, which are more indicative of the surface-atmosphere exchange. Correlations between q and c for this range of eddy scales were likewise closer to À1 for dry conditions, when transfer efficiencies of both scalars exhibited greater similarity. These correlations are influenced by the nonidentical source-sink distributions of the water vapor and carbon dioxide fluxes and the relative magnitude of their constituent fluxes. A new method is introduced to estimate the components of the water vapor flux (transpiration and direct evaporation) and carbon dioxide flux (photosynthesis and respiration) by applying flux variance similarity assumptions separately to the stomatal and the nonstomatal exchange and by considering q-c correlation. Water use efficiency for the vegetation, and how it varies with respect to vapor pressure deficit, is the only input needed for this approach that uses standard eddy covariance measurements. Reasonable estimates yielded by this technique when applied to the contrasting wet and dry days demonstrate its potential for flux partitioning.Citation: Scanlon, T. M., and P. Sahu (2008), On the correlation structure of water vapor and carbon dioxide in the atmospheric surface layer: A basis for flux partitioning, Water Resour. Res., 44, W10418,

Understanding flow-system dynamics of underground coal mine complexes is essential to designing in-situ remediation. Such complexes can be studied by applying time series analysis to the precipitation series and well and mine discharge hydrographs. The Corning mine complex discharges acid mine drainage into Sunday Creek in Ohio, USA. Analysis of the Corning discharge shows that the aquifer has a short (9 days) response time and has little capacity for long-term storage. A time lag of 3-4 days occurs between precipitation and discharge, which corresponds to pressure pulse propagation displacing stored mine water, rather than actual advective flow of water. A gradual decline in hydraulic head is observed from the unsaturated recharge area to the fully flooded discharge area. All well hydrographs show a similar seasonal trend, with mine water levels peaking in early summer, and reaching their lowest levels in early fall, consistent with seasonal recharge patterns. A relatively isolated and fully saturated portion of the mine is sensitive to diurnal barometric pressure, as is typical of confined aquifers, with well water level declining as pressure increases. Other portions of the mine behave as unconfined aquifers, insensitive to barometric pressures. The results also demonstrate the important spatial heterogeneity of the aquifer and indicate that the mine does not behave as a single pool.

No abstract

The Corning mine complex, a suite of abandoned partially flooded room-and-pillar mines in Perry County, Ohio, contributes nearly 100% of the annual acidity load to upper Sunday Creek. Discharge, which issues from a single hole, averages 73 liters/s (2.6 cfs); acidity load averages 590 kg/day; and metal (Fe, Al and Mn) loads average 260, 2 and 13 kg/day respectively. The discharge is a high priority for remediation, but is not well suited for treatment by passive systems. Source-control strategies require knowing recharge sources, flow paths, underground pool interconnections, and mine residence times. This paper describes the development of both conceptual and quantitative models of the mine system, based on mine and soil water budgets, an equivalent-porous-medium numerical model, a barometric efficiency model, and a chemical mixing model. The models were based on monthly water sampling, continuous mine-pool and discharge hydrographs, borehole logs, meteorological data and mine maps. The recharge rate is 20 cm/yr (13% of precipitation). Stream capture contributes 13% of the mine's annual recharge, with diffuse recharge accounting for 87%. During intense rainfall events, however, 50% of recharge can occur by stream capture. Mine storage varies seasonally, depending on recharge, which in turn depends on not just precipitation but also evapotranspirative demand and soilmoisture storage. Consequently, mine storage and discharge are highest in the late spring and lowest in the late summer and early fall. The mine aquifer is a gently dipping grid of rooms and tunnels that collectively provide hydraulic resistance to flow, and it does not form a single hydrostatic pool. Assuming that mines are fully interconnected and that water is well mixed yields a residence time of 5.1 years. However, barometric pressure response shows that the eastern 40% of the mine is separate and partially confined, with exceptionally poor water quality. The eastern portion accounts for only 10% of the flow, but contributes 50% of the chemical load. Excluding the weakly-connected eastern 40%, residence time is 3.9 years. Barometric responses of heads in various parts of the mine show unconfined behavior, confined behavior, or "displacement" behavior, in which heads increase with barometric pressure.