Ce travail met en jeu plusieurs systèmes dynamiques sur des tores en dimension finie, pour lesquels on sait établir des théorèmes limites, qui permettent de préciser leur comportement stochastique. On généralise d'abord le théorème limite local usuel sur un sous-shift de type fini, en ajoutant un terme de perturbation, en reprenant la preuve classique, par des techniques d'opérateurs. On en déduit un théorème limite local pour les sommes de « Riesz-Raïkov unitaires étendues », et des observables höldériennes. Pour cela, on reprend une méthode employée par Bernard Petit, en utilisant des codages symboliques, et le théorème limite local avec perturbation. Puis, on présente plusieurs situations de composées d'automorphismes hyperboliques du tore en dimension deux pour lesquelles on sait établir un théorème limite central quelque soit le choix de la composée. En particulier, on aborde le cas des matrices à coefficients entiers positifs.

<p>We study the power series ring R= K[[x₁,x₂,x₃,...]]on countably infinitely many variables, over a field K, and two particular K-subalgebras of it: the ring S, which is isomorphic to an inverse limit of the polynomial rings in finitely many variables over K, and the ring R', which is the largest graded subalgebra of R.</p><p>Of particular interest are the homogeneous, finitely generated ideals in R', among them the <i>generic ideals</i>. The definition of S as an inverse limit yields a set of <i>truncation homomorphisms</i> from S to K[x₁,...,x_n] which restrict to R'. We have that the truncation of a generic I in R' is a generic ideal in K[x₁,...,x_n]. It is shown in <b>Initial ideals of Truncated Homogeneous Ideals</b> that the initial ideal of such an ideal converge to the initial ideal of the corresponding ideal in R'. This initial ideal need no longer be finitely generated, but it is always <i>locally finitely generated</i>: this is proved in <b>Gröbner Bases in R'</b>. We show in <b>Reverse lexicographic initial ideals of generic ideals are finitely generated</b> that the initial ideal of a generic ideal in R' is finitely generated. This contrast to the lexicographic term order.</p><p> If I in R' is a homogeneous, locally finitely generated ideal, and if we write the Hilbert series of the truncated algebras K[x₁,...,x_n] module the truncation of I as q_n(t)/(1-t)ⁿ, then we show in <b>Generalized Hilbert Numerators </b>that the q_n's converge to a power series in t which we call the <i>generalized Hilbert numerator</i> of the algebra R'/I.</p><p>In <b>Gröbner bases for non-homogeneous ideals in R'</b> we show that the calculations of Gröbner bases and initial ideals in R' can be done also for some non-homogeneous ideals, namely those which have an <i>associated homogeneous ideal</i> which is locally finitely generated.</p><p>The fact that S is an inverse limit of polynomial rings, which are naturally endowed with the discrete topology, provides S with a topology which makes it into a complete Hausdorff topological ring. The ring R', with the subspace topology, is dense in R, and the latter ring is the Cauchy completion of the former. In <b>Topological properties of R'</b> we show that with respect to this topology, locally finitely generated ideals in R'are <i>closed</i>.</p>

Mangroves and seagrass beds serve as essential fish habitat for many economically- and ecologically-valuable species. Depending on their location, these shallow-water habitats are often characterized by substantial fluctuation in salinity levels, which can represent a source of osmoregulatory stress for associated organisms. In South Florida, one of the most important fish species that utilizes these habitats is the gray snapper (Lutjanus griseus). Although this species constitutes a significant portion of the region?s total recreational fishery harvest, the effects of salinity on its distribution, physiology and behavior remain poorly understood. The main goal of this thesis was then to investigate the ecophysiological basis of habitat selection by the gray snapper. Specific objectives include to: (1) examine patterns of distribution and abundance across gradients in environmental salinity; (2) measure physiological status and responses to controlled salinity challenges and; (3) conduct behavioral trials to examine for salinity preferenda (if any). To begin investigating if salinity could be a primary factor structuring the gray snapper assemblages, I examined empirical data collected from Biscayne Bay to test the null hypothesis that gray snapper abundances were evenly distributed along the full salinity range at which samples have been collected. Using the delta approach, three abundance metrics (frequency of occurrence, concentration and delta density) were used as an index for the distribution and abundance of this species. Results indicated that abundance patterns for the smaller gray snapper were consistent with a strategy of reducing osmoregulatory costs by selecting intermediate salinities. However, corresponding abundance patterns for subadult gray snapper were inconsistent with this strategy of minimizing energetic costs, suggesting that this life stage may be indifferent to the range of salinities at which they were observed. These patterns helped developed further hypotheses regarding the ecophysiology of juvenile and subadult gray snapper, the latter of which was then tested via laboratory experiments. Subsequently, I challenged fish in the laboratory with six different salinity treatments (0, 5, 30, 50, 60 and 70ppt, including control) for 192 consecutive hours and collected blood samples at different time points. Results indicated that physiological stress to salinity changes is unlikely to occur at a salinity range of 5 to 50 ppt. At salinities of 0 and 60 ppt transient significant changes in plasma osmolality and/or blood haematocrit were observed, but were corrected after an initial adjustment period of approximately 96 hours. At the highest salinity treatment (70 ppt), a constant osmolality could not be maintained, resulting in death for all fish within 48 hours of exposure. Overall, these findings demonstrate the strong euryhalinity and extraordinary tolerance of this species to both extreme hypo- and hypersaline environments. Finally, I investigated the salinity preference and effects on swimming behavior of the gray snapper in an automated salinity choice shuttlebox via 48-hr trials. In general, gray snapper tested displayed either one of two distinctively different salinity preferences. Half of gray snappers displayed a salinity preference in the range of 9-15 ppt, whereas the other half displayed a salinity preference in the range of 19-23 ppt. Recorded swimming speeds in all fish tested reflected a significant, but weak negative linear relationship with salinity during both time periods of the day (light and dark); however, gray snapper were usually most active during the dark period across all salinities. Overall, these findings reveal that gray snapper prefer slightly hyperosmotic salinities that may minimize the physiological costs of osmoregulation compared to extreme salinities.

A significant problem with exploring for electrically conductive mineral deposits with airborne electromagnetic (AEM) methods is that many of the most valuable sulphide deposits are too conductive to be detected with conventional systems. High-grade sulphide deposits with bulk electrical conductivities on the order of 100,000 S/m can appear as “perfect conductors” to most EM systems because the decay of secondary fields (the “time constant” of the deposit) generated in the target by the system transmitter takes much longer than the short measuring time of EM systems. Their EM response is essentially undetectable with off-time measurements. One solution is to make measurements during the transmitter on-time when the secondary field of the target produced by magnetic flux exclusion is large. The difficulty is that the secondary field must be measured in the presence of a primary field which is orders of magnitude larger. The goal of this thesis is to advance the methodology of making AEM measurements during transmitter on-time by analysing experimental data from three different AEM systems. The first system analysed is a very large separation, two helicopter system where geometry is measured using GPS sensors. In order to calculate the primary field at the receiver with sufficient accuracy, the very large (nominally 400 m) separation requires geometry to be known to better than 1 m. Using the measured geometry to estimate and remove the primary field, I show that a very conductive target can be detected at depths of 200m using the total secondary field. I then used fluxgate magnetometers to correct for receiver rotation which allowed the component of the secondary field to be determined. The second system I examined was a large separation fixed-wing AEM system. Using a towed receiver bird with a smaller (˜ 135m) separation, the geometry must be known much more accurately. In the absence of direct measurement of this geometry, I used a least-squares prediction approach using measurements of aircraft manoeuvres which allowed primary field contamination to be estimated. Subtracting this estimate from the recorded signal increased the maximum time constant observed in a field survey for conductive targets by a factor of seven. Finally, a study of a nominally rigid helicopter EM system employing a bucking coil to cancel primary field showed that system geometry (specifically, the position of the receiver coil relative to the transmitter and bucking coils) must be known to better than 0.01 mm to detect deep targets. Again, direct measurements of system geometry were not available. A least-squares prediction filter using helicopter manoeuvre and system pitch and roll measurements was applied, but was not able to estimate primary field well enough to provide an accurate secondary on-time response. Direct measurements of relative motion of the system components might solve this problem.

We study the power series ring R= K[[x1,x2,x3,...]]on countably infinitely many variables, over a field K, and two particular K-subalgebras of it: the ring S, which is isomorphic to an inverse limit of the polynomial rings in finitely many variables over K, and the ring R', which is the largest graded subalgebra of R. Of particular interest are the homogeneous, finitely generated ideals in R', among them the generic ideals. The definition of S as an inverse limit yields a set of truncation homomorphisms from S to K[x1,...,xn] which restrict to R'. We have that the truncation of a generic I in R' is a generic ideal in K[x1,...,xn]. It is shown in <b>Initial ideals of Truncated Homogeneous Ideals</b> that the initial ideal of such an ideal converge to the initial ideal of the corresponding ideal in R'. This initial ideal need no longer be finitely generated, but it is always locally finitely generated: this is proved in <b>Gröbner Bases in R'</b>. We show in <b>Reverse lexicographic initial ideals of generic ideals are finitely generated</b> that the initial ideal of a generic ideal in R' is finitely generated. This contrast to the lexicographic term order. If I in R' is a homogeneous, locally finitely generated ideal, and if we write the Hilbert series of the truncated algebras K[x1,...,xn] module the truncation of I as qn(t)/(1-t)n, then we show in <b>Generalized Hilbert Numerators </b>that the qn's converge to a power series in t which we call the generalized Hilbert numerator of the algebra R'/I. In <b>Gröbner bases for non-homogeneous ideals in R'</b> we show that the calculations of Gröbner bases and initial ideals in R' can be done also for some non-homogeneous ideals, namely those which have an associated homogeneous ideal which is locally finitely generated. The fact that S is an inverse limit of polynomial rings, which are naturally endowed with the discrete topology, provides S with a topology which makes it into a complete Hausdorff topological ring. The ring R', with the subspace topology, is dense in R, and the latter ring is the Cauchy completion of the former. In <b>Topological properties of R'</b> we show that with respect to this topology, locally finitely generated ideals in R'are closed.

Although research has previously connected comic books and national ideology, there has yet to be a study examining the role of villains in this relationship. By analyzing representations of evil and villainy in the long-running series Captain America and understanding them in light of the model of the circuit of culture, the transforming imaginary space of the American nation can be traced.

A significant problem with exploring for electrically conductive mineral deposits with airborne electromagnetic (AEM) methods is that many of the most valuable sulphide deposits are too conductive to be detected with conventional systems. High-grade sulphide deposits with bulk electrical conductivities on the order of 100,000 S/m can appear as “perfect conductors” to most EM systems because the decay of secondary fields (the “time constant” of the deposit) generated in the target by the system transmitter takes much longer than the short measuring time of EM systems. Their EM response is essentially undetectable with off-time measurements. One solution is to make measurements during the transmitter on-time when the secondary field of the target produced by magnetic flux exclusion is large. The difficulty is that the secondary field must be measured in the presence of a primary field which is orders of magnitude larger. The goal of this thesis is to advance the methodology of making AEM measurements during transmitter on-time by analysing experimental data from three different AEM systems. The first system analysed is a very large separation, two helicopter system where geometry is measured using GPS sensors. In order to calculate the primary field at the receiver with sufficient accuracy, the very large (nominally 400 m) separation requires geometry to be known to better than 1 m. Using the measured geometry to estimate and remove the primary field, I show that a very conductive target can be detected at depths of 200m using the total secondary field. I then used fluxgate magnetometers to correct for receiver rotation which allowed the component of the secondary field to be determined. The second system I examined was a large separation fixed-wing AEM system. Using a towed receiver bird with a smaller (˜ 135m) separation, the geometry must be known much more accurately. In the absence of direct measurement of this geometry, I used a least-squares prediction approach using measurements of aircraft manoeuvres which allowed primary field contamination to be estimated. Subtracting this estimate from the recorded signal increased the maximum time constant observed in a field survey for conductive targets by a factor of seven. Finally, a study of a nominally rigid helicopter EM system employing a bucking coil to cancel primary field showed that system geometry (specifically, the position of the receiver coil relative to the transmitter and bucking coils) must be known to better than 0.01 mm to detect deep targets. Again, direct measurements of system geometry were not available. A least-squares prediction filter using helicopter manoeuvre and system pitch and roll measurements was applied, but was not able to estimate primary field well enough to provide an accurate secondary on-time response. Direct measurements of relative motion of the system components might solve this problem.

No description available.

We collected the data for decomposition of spruce litter to determine the limit values of mass loss and to find both chemical and climate factors that influence limit values. Our data contained 28 sequences of spruce which mainly in Sweden and a small part in other places. We choose mean annual temperature (MAT) and mean annual precipitation (MAP) as climate factors and water solubles, lignin, N, P, K, Ca, Mg and Mn as chemical factors. Then we got the estimated limit values by performing a nonlinear model with mass loss and time spots, and found out the influential factors by using another linear mixed model. At the end we knew that linear mixed model is a proper and efficient approach for determining the factors, P and MAP are the significant factors and Species is a good random effect to explain the variance within groups.

According to the empirical findings from evolution of liquidity, this dissertation constructs an optimal order submission strategy model within which a mixture of market and limit orders can be submitted by both informed and uninformed traders. In the Stacklberg Game Model, informed traders with short-lived private information are regarded as leaders, and uniformed traders with learning behaviors are referred as followers. Our theoretical findings conclude as follows: Firstly, the order strategies of all traders can be characterized as coming under one of seven regimes, pure market buy orders, a combination of market and limit buy orders, pure limit buy orders, a combination of limit buy and limit sell orders, pure limit sell orders, a combination of market and limit sell orders, and pure market sell orders. Traders will select their optimal trading strategy according to the regime within which their liquidation value falls. Parlour (1998) is a special case of this study. Secondly, an increase (reduction) in liquidation value will result in a non-linear increase in the optimal proportion of market order submissions by buyers (sellers). Thirdly, the probability of submitting limit orders for uniformed traders increases when information traders get large profit from the private information. The extreme case is uniformed traders only submit limit orders. This result is consistent with Foucault (1999). Fourthly, the price interval will be much wider when limit orders are submitted by uniformed traders than by informed traders. The reasons are that uniformed traders have no private information and that they are high risk aversion. Finally, numerical illustrations confirm the reliability of this model.