Bayesian updating revisited

Bayesian updating methods provide an alternate philosophy to the characterization of the input variables of a stochastic mathematical model. Here, a priori values of statistical parameters are assumed on subjective grounds or by analysis of a data base from a geologically similar area. As measurements become available during site investigations, updated estimates of parameters characterizing spatial variability are generated. However, in solving the traditional updating equations, an updated covariance matrix may be generated that is not positive-definite, particularly when observed data errors are small. In addition, measurements may indicate that initial estimates of the statistical parameters are poor. The traditional procedure does not have a facility to revise the parameter estimates before the update is carried out. alternatively, Bayesian updating can be viewed as a linear inverse problem that minimizes a weighted combination of solution simplicity and data misfit. Depending on the weight given to the a priori information, a different solution is generated. A Bayesian updating procedure for log-conductivity interpolation that uses a singular value decomposition (SVD) is presented. An efficient and stable algorithm is outlined that computes the updated log-conductivity field and the a posteriori covariance of the estimated values (estimation errors). In addition, an information density matrix is constructed that indicates how well predicted data match observations. Analysis of this matrix indicates the relative importance of the observed data. The SVD updating procedure is used to interpolate the log-conductivity fields of a series of hypothetical aquifers to demonstrate pitfalls and possibilities of the method.     

Use of an Annelid Pollution Index for monitoring a moderately polluted littoral zone

After two sampling surveys, in 1974–1975 and 1984, and with the use of an Annelid Pollution Index, the existence of a degraded Cystoseira stricta community has been established. The community studied was present on superficial rocky substrates on the French Mediterranean coast in unpolluted or moderately polluted zones although, for the latter, pollution reduction measures had been put into effect. The most polluted station in 1984 was close to the effluent of a coastal village sewage treatment plant.     

Coastal Massachusetts pond development: edaphic,climatic, and sea level impacts since deglaciation

Kettle ponds in the Cape Cod National Seashore in southeastern Massachusetts differ in their evolution due to depth of the original ice block, the clay content of outwash in their drainage basins, and their siting in relation to geomorphic changes caused by sea-level rise, barrier beach formation, and saltmarsh development. Stratigraphic records of microfossil, carbon isotope, and sediment changes also document late-glacial and Holocene climatic changes.The ponds are separated into 3 groups, each of which follow different development scenarios. Group I ponds date from the late-glacial. They formed in clay-rich outwash, have perched aquifers and continuous lake sediment deposition. The earliest pollen and macrofossil assemblages in Group I pond sediments suggest tundra and spruce-willow parklands before 12 000 yr B.P., boreal forest between 12 000 and 10 500 yr B.P., bog/heath initiation and expansion during the Younger Dryas between 11 000 and 10 000 yr B.P., northern conifer forest between 10 500 and 9500 yr B.P., and establishment of the Cape oak and pitch pine barrens vegetation after 9500 yr B.P. Sedimentation rate changes suggest lowered freshwater levels between 9000 and 5000 yr B.P. caused by decreased precipitation on the Atlantic Coastal Plain. Lake sediment deposition began in the middle Holocene in Group II ponds which formed in clay-poor outwash. These ponds date from about 6000-5000 yr B.P. In these ponds sediment deposition began as sea level rose and the freshwater lens intersected the dry basins. The basal radiocarbon dates of these ponds and stable carbon isotope analyses of the pond sediments suggest a sea-level curve for Cape Cod Bay. Holocene topographic changes in upland and the landscape surrounding the ponds is reconstructed for this coastal area.Group III ponds in the late Holocene landscape of the Provincelands dunes originated as interdunal bogs about 1000 yr B.P. and became ponds more recently as water-levels increased. Peat formation in the Provincelands reflects climatic changes evident on both sides of the Atlantic region.This is the 8th in a series of papers published in this special AMQUA issue. These papers were presented at the 1994 meeting of the American Quaternary Association held 19–22 June, 1994, at the University of Minnesota, Minneapolis, Minnesota, USA. Dr Linda C. K. Shane served as guest editor for these papers.     

A rapid-response 2-D drag anemometer for atmospheric turbulence measurements

A rapid response drag anemometer for measuring streamwise and lateral components of horizontal windspeed is described. Theory of operation, design and calibration are discussed with emphasis on the electronic preconditioning of signals and problems associated with using a mechanically resonant system as a sensor. Field comparisons showed half-hourly means and standard deviations of the streamwise component to be within 8% and 5% of respective values obtained from a 3-dimensional sonic anemometer. The lateral component from the drag anemometer was significantly more noisy than that from the 3-D sonic due to induced oscillations arising from vortex shedding. After mechanical and electronic filtering, half-hourly standard deviation comparisons agreed to within 6% for this component. Friction velocities obtained from the drag anemometer in combination with a 1-D sonic, agreed with measurements from the 3-D sonic anemometer to within 4% over a measured range of 0.05 to 1.2 m s^-1     

Geochemical characteristics of Cocos Plate seamount lavas

A wide compositional continuum of basalts has been erupted from near-ridge seamounts constructed on the Cocos Plate between the Clipperton and Orozco Francture Zones. They range from highly evolved to moderately primitive (3.0–7.8% MgO), LREE-enriched alkali basalts, to moderately evolved to near-primary (5.2–9.5% MgO) tholeiites indistinguishable from N-type MORB. The data set of 159 quench glass analyses exhibits a remarkably consistent variation in both major and trace element composition that is keyed to variations in (La/Sm). Modeling of potential liquid lines of descent at pressures ranging from 1 bar to 8 kbar shows that this covariation is partially due to systematic differences in liquid lines of descent, where the alkaline lavas have undergone substantially more high pressure clinopyroxene fractionation and substantially less low pressure plagioclase fractionation than the tholeiites. In addition, systematic variation in the composition of the more primitive glasses indicates that they were derived from mixing of discrete enriched and depleted melts in the heterogenous seamount mantle source at pressures of 8–10 kbar and greater, and that clinopyroxene may be a residual phase during partial melting. These results show that porous media flow in the seamount mantle source is minor and that melt transport is accomplished primarily through cracking and diking. This study supports suggestions that the general homogeneity of basalt along the EPR is due to mixing in sub-axial magma chambers and mush zones, with additional mixing during partial mantle melting and melt segregation.     

Lamprophyric lavas in the Colima graben,SW Mexico

The Colima graben, located in SW Mexico, is one of three grabens which intersect about 50 km SSW of Guadalajara, forming a triple junction. The 90 km long, 20–60 km wide Colima graben represents a N-S rift of the E-W trending Mexican Volcanic Belt. Since the Early Pliocene, the Colima graben has served as a locus for the eruption of alkaline lavas, the most recent of which are basanites and minettes erupted from Late Pleistocene cinder cones (Luhr and Carmichael 1981). In this paper, we report on older alkaline lavas which crop out in the graben's walls. These rocks include phlogopite- and hornblende-bearing lamprophyres, a phlogopite-kalsilite-ankaratrite, and high-K andesites. These lavas crop out throughout the Colima graben area, and are intimately associated with calc-alkaline lavas in the field. Compared to these, the alkaline rocks are strikingly enriched in the incompatible elements, particularly Ba, Sr, P, and the LREE. Unlike the younger Late Pleistocene alkaline cinder cone lavas, most of the graben wall lamprophyres and the high-K andesites represent magmas that appear to have undergone significant evolution since their generation, including fractionation, crustal contamination, and possible magma mixing. Least-squares modeling indicates that the cinder cone minettes represent reasonable parental magmas for the graben lamprophyres. The occurrence of these alkaline lavas in an active calc-alkaline volcanic arc is unusual, and we suggest that they are a manifestation of the rifting processes which produced the Colima graben.