Physical volcanology, sedimentology, stratigraphy and petrochemistry of the Berry Creek metavolcanics: an Archean calc-alkaline complex, Lake of the Woods, Ontario

The steeply dipping, isoclinally folded early Precambrian (Archean) Berry Creek Metavolcanic Complex comprises primary to resedimented pyroclastic, epiclastic and autoclastic deposits. Tephra erupted from central volcanic edifices was dumped by mass flow mechanisms into peripheral volcanosedimentary depressions. Sedimentation has been essentially contemporaneous with eruption and transport of tephra. The monolithic to heterolithic tuffaceous horizons are interpreted as subaerial to subaqueous pumice and ash flows, secondary debris flows, lahars, slump deposits and turbidites. Monolithic debris flows, derived from crumble breccia and dcme talus, formed during downslope collapse and subsequent gravity flowage. Heterolithic tuff, lahars and lava flow morphologies suggest at least temporary emergence of the edifice. Local collapse may have accompanied pyroclastic volcanism. The tephra, produced by hydromagmatic to magmatic eruptions, were rapidly transported, by primary and secondary mechanisms, to a shallow littoral to deep water subaqueous fan developed upon the subjacent mafic metavolcanic platform. Deposition resulted from traction, traction carpet, and suspension sedimentation from laminar to turbulent flows. Facies mapping revealed proximal (channel to overbank) to distal facies epiclastics (greywackes, argillite) intercalated with proximal vent to medial fan facies crystal rich ash flows, debris flows, bedded tuff and shallow water to deep water lava flows. Framework and matrix support debris flows exhibit a variety of subaqueous sedimentary structures, e.g., coarse tail grading, double grading, inverse to normal grading, graded stratified pebbly horizons, erosional channels. Pelitic to psammitic AE turbidites also contain primary stru~tures, e.g., flames, load casts, dewatering pipes. Despite low to intermediate pressure greenschist to amphibolite grade metamorphism and variably penetrative deformation, relicts of pumice fragments and shards were recognized as recrystallized quartzofeldspathic pseudomorphs. The mafic to felsic metavolcanics and metasediments contain blasts of hornblende, actinolite, garnet, pistacitic epidote, staurolite, albitic plagioclase, and rarely andalusite and cordierite. The mafic metavolcanics (Adams River Bay, Black River, Kenu Lake, Lobstick Bay, Snake Bay) display _holeiitic trends with komatiitic affinities. Chemical variations are consistent with high level fractionation of olivine, plagioclase, amphibole, and later magnetite from a parental komatiite. The intermediate to felsic (64-74% Si02) metavolcanics generally exhibit calc-alkaline trends. The compositional discontinuity, defined by major and trace element diversity, can be explained by a mechanism involving two different magma sources. Application of fractionation series models are inconsistent with the observed data. The tholeiitic basalts and basaltic andesites are probably derived by low pressure fractionation of a depleted (high degree of partial melting) mantle source. The depleted (low Y, Zr) calc-alkaline metavolcanics may be produced by partial melting of a geochemically evolved source, e.g., tonalitetrondhjemite, garnet amphibolite or hydrous basalt.

Evolutionary synthesis of stochastic gene network models using feature-based search spaces

A feature-based fitness function is applied in a genetic programming system to synthesize stochastic gene regulatory network models whose behaviour is defined by a time course of protein expression levels. Typically, when targeting time series data, the fitness function is based on a sum-of-errors involving the values of the fluctuating signal. While this approach is successful in many instances, its performance can deteriorate in the presence of noise. This thesis explores a fitness measure determined from a set of statistical features characterizing the time series' sequence of values, rather than the actual values themselves. Through a series of experiments involving symbolic regression with added noise and gene regulatory network models based on the stochastic 'if-calculus, it is shown to successfully target oscillating and non-oscillating signals. This practical and versatile fitness function offers an alternate approach, worthy of consideration for use in algorithms that evaluate noisy or stochastic behaviour.