Storage, migration, and eruption of magma at Kilauea volcano, Hawaii, 1971–1972

The magmatic plumbing system of Kilauea Volcano consists of a broad region of magma generation in the upper mantle, a steeply inclined zone through which magma rises to an intravolcano reservoir located about 2 to 6 km beneath the summit of the volcano, and a network of conduits that carry magma from this reservoir to sites of eruption within the caldera and along east and southwest rift zones. The functioning of most parts of this system was illustrated by activity during 1971 and 1972. When a 29-month-long eruption at Mauna Ulu on the east rift zone began to wane in 1971, the summit region of the volcano began to inflate rapidly; apparently, blockage of the feeder conduit to Mauna Ulu diverted a continuing supply of mantle-derived magma to prolonged storage in the summit reservoir. Rapid inflation of the summit area persisted at a nearly constant rate from June 1971 to February 1972, when a conduit to Mauna Ulu was reopened. The cadence of inflation was twice interrupted briefly, first by a 10-hour eruption in Kilauea Caldera on 14 August, and later by an eruption that began in the caldera and migrated 12 km down the southwest rift zone between 24 and 29 September. The 14 August and 24–29 September eruptions added about 10⁷ m³ and 8 × 10⁶ m³, respectively, of new lava to the surface of Kilauea. These volumes, combined with the volume increase represented by inflation of the volcanic edifice itself, account for an approximately 6 × 10⁶ m³/month rate of growth between June 1971 and January 1972, essentially the same rate at which mantle-derived magma was supplied to Kilauea between 1952 and the end of the Mauna Ulu eruption in 1971.The August and September 1971 lavas are tholeiitic basalts of similar major-element chemical composition. The compositions can be reproduced by mixing various proportions of chemically distinct variants of lava that erupted during the preceding activity at Mauna Ulu. Thus, part of the magma rising from the mantle to feed the Mauna Ulu eruption may have been stored within the summit reservoir from 4 to 20 months before it was erupted in the summit caldera and along the southwest rift zone in August and September.The September 1971 activity was only the fourth eruption on the southwest rift zone during Kilauea's 200 years of recorded history, in contrast to more than 20 eruptions on the east rift zone. Order-of-magnitude differences in topographic and geophysical expression indicate greatly disparate eruption rates for far more than historic time and thus suggest a considerably larger dike swarm within the east rift zone than within the southwest rift zone. Characteristics of the historic eruptions on the southwest rift zone suggest that magma may be fed directly from active lava lakes in Kilauea Caldera or from shallow cupolas at the top of the summit magma reservoir, through fissures that propagate down rift from the caldera itself at the onset of eruption. Moreover, emplacement of this magma into the southwest rift zone may be possible only when compressive stress across the rift is reduced by some unknown critical amount owing either to seaward displacement of the terrane south-southeast of the rift zone or to a deflated condition of Mauna Loa Volcano adjacent to the northwest, or both. The former condition arises when the forceful emplacement of dikes into the east rift zone wedges the south flank of Kilauea seaward. Such controls on the potential for eruption along the southwest rift zone may be related to the topographic and geophysical constrasts between the two rift zones.     

Optimized system to improve pumping rate stability during aquifer tests

Aquifer hydraulic properties are commonly estimated using aquifer tests, which are based on an assumption of a uniform and constant pumping rate. Substantial uncertainties in the flow rate across the borehole-formation interface can be induced by dynamic head losses, caused by rapid changes in borehole water levels early in an aquifer test. A system is presented that substantially reduces these sources of uncertainty by explicitly accounting for dynamic head losses. The system which employs commonly available components (including a datalogger, pressure transducers, a variable-speed pump motor, a flow controller, and flowmeters), is inexpensive, highly mobile, and easily set up. It optimizes the flow rate at the borehole-formation interface, making it suitable for any type of aquifer test, including constant, step, or ramped withdrawal and injection, as well as sinusoidal. The system was demonstrated for both withdrawal and injection tests in three aquifers at the Savannah River Site. No modifications to the control system were required, although a small number of characteristics of the pumping and monitoring system were added to the operating program. The pumping system provided a statistically significant, constant flow rate with time. The range in pumping variability (95% confidence interval) was from +/- 2.58 x 10(-4) L/sec to +/- 9.07 x 10(-4) L/sec, across a wide range in field and aquifer conditions.     

A Modified Radial Diagram Approach for Evaluating Natural Attenuation Trends for Chlorinated Solvents and Inorganic Redox Indicators

Selection of monitored natural attenuation as a ground water remedy requires that sound scientific documentation clearly illustrating the effectiveness of this remedial alternative be presented to regulatory agencies and concerned citizens. An innovative radial diagram approach is applied to illustrate natural attenuation trends for total benzene, toluene, ethylbenzene, and xylenes (BTEX) and chlorinated ethenes at a former fire training area at Pittsburgh Air Force Base, New York. A BTEX-CAH (chlorinated aliphatic hydrocarbons) radial diagram map shows that concentrations of site contaminants are generally decreasing along the primary flowpath downgradient from the source area. This radial diagram map also suggests that there is a spatial correlation between decreasing CAH parent compound concentrations and increasing or stable daughter product concentrations. This provides secondary evidence of intrinsic biodegradation of TCE downgradient from the source area. A SEQUENCE-Redox™ map suggests that there is a spatial correlation between trends in electron acceptor and metabolic byproduct concentrations, and the decline in total BTEX concentrations downgradient from the source area. This correlation provides secondary evidence for the intrinsic biodegradation of total BTEX in the aquifer. This study demonstrates that radial diagram visual aids can provide a clear and efficient approach for documenting natural attenuation lines of evidence, as an alternative or a complement to using multiple contour maps, tabulated data, or log-linear plots.     

Estrogenic Activity and Volume Fraction of Waste Water Origin in Monitoring Wells Along the Santa Cruz River, Arizona

The fate of estrogenic activity in waste water effluent was examined during surface transport and incidental recharge along the Santa Cruz River in Pima County, Arizona. Based on measurement of boron isotopes, the fractional contribution of reclaimed water in surface waters and ground water wells proximate to the river was determined for a contemporary sample set. Estrogenic activity decreased by −60% over the 25 mi length of the river below effluent discharge points in Tucson. In ground water samples obtained from monitoring wells that are proximate to the Santa Cruz River, both dissolved organic carbon ( p = 0.0003) and estrogenic activity ( p = 3 × 10⁻⁶) were highly correlated to fractional waste water content. Results indicate that proximate ground water quality is sensitive to incidental recharge of reclaimed water in the Santa Cruz River bed. In a few locations, little attenuation of estrogenic activity was apparent during percolation of effluent in the river channel to well withdrawal points.     

Biogeochemical transformation of Fe minerals in a petroleum-contaminated aquifer

The Bemidji aquifer in Minnesota, USA is a well-studied site of subsurface petroleum contamination. The site contains an anoxic groundwater plume where soluble petroleum constituents serve as an energy source for a region of methanogenesis near the source and bacterial Fe(III) reduction further down gradient. Methanogenesis apparently begins when bioavailable Fe(III) is exhausted within the sediment. Past studies indicate that Geobacter species and Geothrix fermentens-like organisms are the primary dissimilatory Fe-reducing bacteria at this site. The Fe mineralogy of the pristine aquifer sediments and samples from the methanogenic (source) and Fe(III) reducing zones were characterized in this study to identify microbiologic changes to Fe valence and mineral distribution, and to identify whether new biogenic mineral phases had formed. Methods applied included X-ray diffraction; X-ray fluorescence (XRF); and chemical extraction; optical, transmission, and scanning electron microscopy; and Mössbauer spectroscopy.All of the sediments were low in total Fe content (≈ 1%) and exhibited complex Fe-mineralogy. The bulk pristine sediment and its sand, silt, and clay-sized fractions were studied in detail. The pristine sediments contained Fe(II) and Fe(III) mineral phases. Ferrous iron represented approximately 50% of Fe_TOT. The relative Fe(II) concentration increased in the sand fraction, and its primary mineralogic residence was clinochlore with minor concentrations found as a ferroan calcite grain cement in carbonate lithic fragments. Fe(III) existed in silicates (epidote, clinochlore, muscovite) and Fe(III) oxides of detrital and authigenic origin. The detrital Fe(III) oxides included hematite and goethite in the form of mm-sized nodular concretions and smaller-sized dispersed crystallites, and euhedral magnetite grains. Authigenic Fe(III) oxides increased in concentration with decreasing particle size through the silt and clay fraction. Chemical extraction and Mössbauer analysis indicated that this was a ferrihydrite like-phase. Quantitative mineralogic and Fe(II/III) ratio comparisons between the pristine and contaminated sediments were not possible because of textural differences. However, comparisons between the texturally-similar source (where bioavailable Fe(III) had been exhausted) and Fe(III) reducing zone sediments (where bioavailable Fe(III) remained) indicated that dispersed detrital, crystalline Fe(III) oxides and a portion of the authigenic, poorly crystalline Fe(III) oxide fraction had been depleted from the source zone sediment by microbiologic activity. Little or no effect of microbiologic activity was observed on silicate Fe(III). The presence of residual “ferrihydrite” in the most bioreduced, anoxic plume sediment (source) implied that a portion of the authigenic Fe(III) oxides were biologically inaccessible in weathered, lithic fragment interiors. Little evidence was found for the modern biogenesis of authigenic ferrous-containing mineral phases, perhaps with the exception of thin siderite or ferroan calcite surface precipitates on carbonate lithic fragments within source zone sediments.     

The sedimentary record of drifting ice (early Wisconsin Sunnybrook deposit) in an ancestral ice-dammed Lake Ontario, Canada

Outcrops of pebbly mud (diamict) at Scarborough in Southern Ontario, Canada (the so-called Sunnybrook ‘Till’) are associated with the earliest incursion of the Laurentide Ice Sheet (LIS) into mid-continent North America some 45,000 years ago. The Sunnybrook is a blanket-like deposit containing deepwater ostracodes and occurs conformably within a thick (100 m) succession of deltaic and glaciolacustrine facies that record water depth changes in a large proglacial lake. Contextual evidence (associated facies, sedimentary structures, deposit geometry and landforms) indicates a low energy depositional setting in an ice-dammed ancestral Lake Ontario in which scouring by floating ice masses was an important process. U-shaped, iceberg-cut scours (with lateral berms) up to 7 m deep, occur on the upper surface of the Sunnybrook and are underlain by ‘sub-scour’ structures that extend several meters below the scour base. Ice-rafted concentrations of clasts (‘clast layers’), grooved surfaces formed by floating ice glissading over a muddy lake floor (‘soft sediment striations’) and melanges of sand and mud mixed by grounding ice keels (‘ice keel turbates’) are present and are all well known from modern cold environments. The wider significance of this depositional model is that the LIS margin lay east of Scarborough and did not overrun Southern Ontario. This finding is in agreement with recent data from the Erie Basin of Canada, Ohio, and Indiana where deposits formerly correlated with the Sunnybrook (and thus implying an extensive early Wisconsin ice sheet) are now regarded as Illinoian. A speculative hypothesis is proposed that relates deposition of the Sunnybrook and two younger deposits of similar sedimentology, to surge-like instabilities of the southern LIS margin.     

Process-based snowmelt modeling: does it require more input data than temperature-index modeling?

Modeling snow hydrology for cold regions remains a problematic aspect of many hydro-environmental models. Temperature-index methods are commonly used and are routinely justified under the auspices that process-based models require too many input data. To test this claim, we used a physical, process-based model to simulate snowmelt at four locations across the conterminous US using energy components estimated from measured daily maximum and minimum temperature, i.e. using only the same data required for temperature-index models. The results showed good agreement between observed and predicted snow water equivalents, average R²>0.9. We duplicated the simulations using a simple temperature-index model best fitted to the data and results were poorer, R²<0.8. At one site we applied the process-based model without substantial parameter estimation, and there were no significant (=0.05) differences between these results and those obtained using temperature-estimated parameters, despite relatively poorly predicted specific energy budget components (R²<0.8). These results encourage the use of mechanistic snowmelt modeling approaches in hydrological models, especially in distributed hydrological models for which landscape snow distribution may be controlled by spatially distributed components of the environmental energy budget.     

Observations of planview flux patterns within convective structures of the marine atmospheric surface layer

Air/sea flux variability on horizontal scales from 50 m to several km results, in part, from the presence of coherent convective structures within the atmospheric boundary layer. The horizontal distribution of fluxes within these convective updrafts and downdrafts is, therefore, central to studies of air/sea interaction and remote sensing of sea surface wind and wave fields. This study derives these flux patterns from observations of the Marine Atmospheric Surface Layer (MASL).Research aircraft flights through the MASL provide an optimal means for sampling large numbers of the above-mentioned coherent structures. The NCAR Electra flew numerous legs through the MASL at a height of 50 m during the 1987 stratocumulus phase of Project FIRE (First ISSCP (International Satellite Cloud Climatology Program) Regional Experiment).In situ measurements from these legs serve as the dataset for this paper. The data are processed in such a way as to retain only the turbulence fluctuations. Conditional sampling, based on the vertical velocity field, results in the isolation of convective updrafts and downdrafts. Compositing of the data for these two classes of convective drafts results in horizontal planviews of the vertical fluxes of buoyancy, absolute humidity, along-meanwind component of momentum, and vertical velocity. To ensure dynamical similarity, these horizontal planviews are oriented in a coordinate system aligned with the mean wind.