Kernel-based Mapping of Meteorological Fields in Complex Orography

Due to the advances in sensor networks and remote sensing technologies, the acquisition and storage rates of meteorological and climatological data increases every day and ask for novel and efficient processing algorithms. A fundamental problem of data analysis and modeling is the spatial prediction of meteorological variables in complex orography, which serves among others to extended climatological analyses, for the assimilation of data into numerical weather prediction models, for preparing inputs to hydrological models and for real time monitoring and short-term forecasting of weather.In this thesis, a new framework for spatial estimation is proposed by taking advantage of a class of algorithms emerging from the statistical learning theory. Nonparametric kernel-based methods for nonlinear data classification, regression and target detection, known as support vector machines (SVM), are adapted for mapping of meteorological variables in complex orography.With the advent of high resolution digital elevation models, the field of spatial prediction met new horizons. In fact, by exploiting image processing tools along with physical heuristics, an incredible number of terrain features which account for the topographic conditions at multiple spatial scales can be extracted. Such features are highly relevant for the mapping of meteorological variables because they control a considerable part of the spatial variability of meteorological fields in the complex Alpine orography. For instance, patterns of orographic rainfall, wind speed and cold air pools are known to be correlated with particular terrain forms, e.g. convex/concave surfaces and upwind sides of mountain slopes.Kernel-based methods are employed to learn the nonlinear statistical dependence which links the multidimensional space of geographical and topographic explanatory variables to the variable of interest, that is the wind speed as measured at the weather stations or the occurrence of orographic rainfall patterns as extracted from sequences of radar images. Compared to low dimensional models integrating only the geographical coordinates, the proposed framework opens a way to regionalize meteorological variables which are multidimensional in nature and rarely show spatial auto-correlation in the original space making the use of classical geostatistics tangled.The challenges which are explored during the thesis are manifolds. First, the complexity of models is optimized to impose appropriate smoothness properties and reduce the impact of noisy measurements. Secondly, a multiple kernel extension of SVM is considered to select the multiscale features which explain most of the spatial variability of wind speed. Then, SVM target detection methods are implemented to describe the orographic conditions which cause persistent and stationary rainfall patterns. Finally, the optimal splitting of the data is studied to estimate realistic performances and confidence intervals characterizing the uncertainty of predictions.The resulting maps of average wind speeds find applications within renewable resources assessment and opens a route to decrease the temporal scale of analysis to meet hydrological requirements. Furthermore, the maps depicting the susceptibility to orographic rainfall enhancement can be used to improve current radar-based quantitative precipitation estimation and forecasting systems and to generate stochastic ensembles of precipitation fields conditioned upon the orography.

Comparing ICA-based and single-trial topographic ERP analyses.

Single-trial analysis of human electroencephalography (EEG) has been recently proposed for better understanding the contribution of individual subjects to a group-analysis effect as well as for investigating single-subject mechanisms. Independent Component Analysis (ICA) has been repeatedly applied to concatenated single-trial responses and at a single-subject level in order to extract those components that resemble activities of interest. More recently we have proposed a single-trial method based on topographic maps that determines which voltage configurations are reliably observed at the event-related potential (ERP) level taking advantage of repetitions across trials. Here, we investigated the correspondence between the maps obtained by ICA versus the topographies that we obtained by the single-trial clustering algorithm that best explained the variance of the ERP. To do this, we used exemplar data provided from the EEGLAB website that are based on a dataset from a visual target detection task. We show there to be robust correspondence both at the level of the activation time courses and at the level of voltage configurations of a subset of relevant maps. We additionally show the estimated inverse solution (based on low-resolution electromagnetic tomography) of two corresponding maps occurring at approximately 300 ms post-stimulus onset, as estimated by the two aforementioned approaches. The spatial distribution of the estimated sources significantly correlated and had in common a right parietal activation within Brodmann's Area (BA) 40. Despite their differences in terms of theoretical bases, the consistency between the results of these two approaches shows that their underlying assumptions are indeed compatible.

Characterization of rapeseed (Brassica napus) oils by bulk C, O, H, and fatty acid C stable isotope analyses

Rapeseed (Brassica napus) oils differing in cultivar, sites of growth, and harvest year were characterized by fatty acid concentrations and carbon, hydrogen, and oxygen stable isotope analyses of bulk oils (delta(13)C(bulk), delta(2)H(bulk), delta(18)O(bulk) values) and individual fatty acids (delta(13)C(FA)). The delta(13)C(bulk), delta(2)H(bulk), and delta(18)O(bulk) values were determined by continuous flow combustion and high-temperature conversion elemental analyzer isotope ratio mass spectrometry (EA/IRMS, TC-EA/IRMS). The delta(13)C(FA) values were determined using gas chromatography-combustion isotope ratio mass spectrometry (GC/C/IRMS). For comparison, other C(3) vegetable oils rich in linolenic acid (flax and false flax oils) and rich in linoleic acid (poppy, sunflower, and safflower oils) were submitted to the same chemical and isotopic analyses. The bulk and molecular delta(13)C values were typical for C(3) plants. The delta(13)C value of palmitic acid (delta(13)C(16:0)) and n-3 alpha-linolenic acid (delta(13)C(18:3n-3)) differed (p < 0.001) between rape, flax, and poppy oils. Also within species, significant differences of delta(13)C(FA) were observed (p < 0.01). The hydrogen and oxygen isotope compositions of rape oil differed between cultivars (p < 0.05). Major differences in the individual delta(13)C(FA) values were found. A plant-specific carbon isotope fractionation occurs during the biosynthesis of the fatty acids and particularly during desaturation of C(18) acids in rape and flax. Bulk oil and specific fatty acid stable isotope analysis might be useful in tracing dietary lipids differing in their origin.

Generation and nonlinear evolution of shore-oblique/transverse sand bars

The coupling between topography, waves and currents in the surf zone may selforganize to produce the formation of shore-transverse or shore-oblique sand bars on an otherwise alongshore uniform beach. In the absence of shore-parallel bars, this has been shown by previous studies of linear stability analysis, but is now extended to the finite-amplitude regime. To this end, a nonlinear model coupling wave transformation and breaking, a shallow-water equations solver, sediment transport and bed updating is developed. The sediment flux consists of a stirring factor multiplied by the depthaveraged current plus a downslope correction. It is found that the cross-shore profile of the ratio of stirring factor to water depth together with the wave incidence angle primarily determine the shape and the type of bars, either transverse or oblique to the shore. In the latter case, they can open an acute angle against the current (upcurrent oriented) or with the current (down-current oriented). At the initial stages of development, both the intensity of the instability which is responsible for the formation of the bars and the damping due to downslope transport grow at a similar rate with bar amplitude, the former being somewhat stronger. As bars keep on growing, their finite-amplitude shape either enhances downslope transport or weakens the instability mechanism so that an equilibrium between both opposing tendencies occurs, leading to a final saturated amplitude. The overall shape of the saturated bars in plan view is similar to that of the small-amplitude ones. However, the final spacings may be up to a factor of 2 larger and final celerities can also be about a factor of 2 smaller or larger. In the case of alongshore migrating bars, the asymmetry of the longshore sections, the lee being steeper than the stoss, is well reproduced. Complex dynamics with merging and splitting of individual bars sometimes occur. Finally, in the case of shore-normal incidence the rip currents in the troughs between the bars are jet-like while the onshore return flow is wider and weaker as is observed in nature.

Parametric Approach to Blind Deconvolution of Nonlinear Channels

A parametric procedure for the blind inversion of nonlinear channels is proposed, based on a recent method of blind source separation in nonlinear mixtures. Experiments show that the proposed algorithms perform efficiently, even in the presence of hard distortion. The method, based on the minimization of the output mutual information, needs the knowledge of log-derivative of input distribution (the so-called score function). Each algorithm consists of three adaptive blocks: one devoted to adaptive estimation of the score function, and two other blocks estimating the inverses of the linear and nonlinear parts of the channel, (quasi-)optimally adapted using the estimated score functions. This paper is mainly concerned by the nonlinear part, for which we propose two parametric models, the first based on a polynomial model and the second on a neural network, while [14, 15] proposed non-parametric approaches.

A Fast Gradient Approximation for Nonlinear Blind Signal Processing

When dealing with nonlinear blind processing algorithms (deconvolution or post-nonlinear source separation), complex mathematical estimations must be done giving as a result very slow algorithms. This is the case, for example, in speech processing, spike signals deconvolution or microarray data analysis. In this paper, we propose a simple method to reduce computational time for the inversion of Wiener systems or the separation of post-nonlinear mixtures, by using a linear approximation in a minimum mutual information algorithm. Simulation results demonstrate that linear spline interpolation is fast and accurate, obtaining very good results (similar to those obtained without approximation) while computational time is dramatically decreased. On the other hand, cubic spline interpolation also obtains similar good results, but due to its intrinsic complexity, the global algorithm is much more slow and hence not useful for our purpose.

Post-Nonlinear Mixtures and Beyond

Although sources in general nonlinear mixturm arc not separable iising only statistical independence, a special and realistic case of nonlinear mixtnres, the post nonlinear (PNL) mixture is separable choosing a suited separating system. Then, a natural approach is based on the estimation of tho separating Bystem parameters by minimizing an indcpendence criterion, like estimated mwce mutual information. This class of methods requires higher (than 2) order statistics, and cannot separate Gaarsian sources. However, use of [weak) prior, like source temporal correlation or nonstationarity, leads to other source separation Jgw rithms, which are able to separate Gaussian sourra, and can even, for a few of them, works with second-order statistics. Recently, modeling time correlated s011rces by Markov models, we propose vcry efficient algorithms hmed on minimization of the conditional mutual information. Currently, using the prior of temporally correlated sources, we investigate the fesihility of inverting PNL mixtures with non-bijectiw non-liacarities, like quadratic functions. In this paper, we review the main ICA and BSS results for riunlinear mixtures, present PNL models and algorithms, and finish with advanced resutts using temporally correlated snu~sm

Energy relaxation in nonlinear one-dimensional lattices

We study energy relaxation in thermalized one-dimensional nonlinear arrays of the Fermi-Pasta-Ulam type. The ends of the thermalized systems are placed in contact with a zero-temperature reservoir via damping forces. Harmonic arrays relax by sequential phonon decay into the cold reservoir, the lower-frequency modes relaxing first. The relaxation pathway for purely anharmonic arrays involves the degradation of higher-energy nonlinear modes into lower-energy ones. The lowest-energy modes are absorbed by the cold reservoir, but a small amount of energy is persistently left behind in the array in the form of almost stationary low-frequency localized modes. Arrays with interactions that contain both a harmonic and an anharmonic contribution exhibit behavior that involves the interplay of phonon modes and breather modes. At long times relaxation is extremely slow due to the spontaneous appearance and persistence of energetic high-frequency stationary breathers. Breather behavior is further ascertained by explicitly injecting a localized excitation into the thermalized arrays and observing the relaxation behavior.

A protocol for a systematic review on the impact of unpublished studies and studies published in the gray literature in meta-analyses.

BACKGROUND: Meta-analyses are particularly vulnerable to the effects of publication bias. Despite methodologists' best efforts to locate all evidence for a given topic the most comprehensive searches are likely to miss unpublished studies and studies that are published in the gray literature only. If the results of the missing studies differ systematically from the published ones, a meta-analysis will be biased with an inaccurate assessment of the intervention's effects.As part of the OPEN project (http://www.open-project.eu) we will conduct a systematic review with the following objectives:â-ª To assess the impact of studies that are not published or published in the gray literature on pooled effect estimates in meta-analyses (quantitative measure).â-ª To assess whether the inclusion of unpublished studies or studies published in the gray literature leads to different conclusions in meta-analyses (qualitative measure). METHODS/DESIGN: Inclusion criteria: Methodological research projects of a cohort of meta-analyses which compare the effect of the inclusion or exclusion of unpublished studies or studies published in the gray literature.Literature search: To identify relevant research projects we will conduct electronic searches in Medline, Embase and The Cochrane Library; check reference lists; and contact experts.Outcomes: 1) The extent to which the effect estimate in a meta-analyses changes with the inclusion or exclusion of studies that were not published or published in the gray literature; and 2) the extent to which the inclusion of unpublished studies impacts the meta-analyses' conclusions.Data collection: Information will be collected on the area of health care; the number of meta-analyses included in the methodological research project; the number of studies included in the meta-analyses; the number of study participants; the number and type of unpublished studies; studies published in the gray literature and published studies; the sources used to retrieve studies that are unpublished, published in the gray literature, or commercially published; and the validity of the methodological research project.Data synthesis: Data synthesis will involve descriptive and statistical summaries of the findings of the included methodological research projects. DISCUSSION: Results are expected to be publicly available in the middle of 2013.

Reference values and factors associated with renal resistive index in a family-based population study.

Increased renal resistive index (RRI) has been recently associated with target organ damage and cardiovascular or renal outcomes in patients with hypertension and diabetes mellitus. However, reference values in the general population and information on familial aggregation are largely lacking. We determined the distribution of RRI, associated factors, and heritability in a population-based study. Families of European ancestry were randomly selected in 3 Swiss cities. Anthropometric parameters and cardiovascular risk factors were assessed. A renal Doppler ultrasound was performed, and RRI was measured in 3 segmental arteries of both kidneys. We used multilevel linear regression analysis to explore the factors associated with RRI, adjusting for center and family relationships. Sex-specific reference values for RRI were generated according to age. Heritability was estimated by variance components using the ASSOC program (SAGE software). Four hundred women (mean age±SD, 44.9±16.7 years) and 326 men (42.1±16.8 years) with normal renal ultrasound had mean RRI of 0.64±0.05 and 0.62±0.05, respectively (P<0.001). In multivariable analyses, RRI was positively associated with female sex, age, systolic blood pressure, and body mass index. We observed an inverse correlation with diastolic blood pressure and heart rate. Age had a nonlinear association with RRI. We found no independent association of RRI with diabetes mellitus, hypertension treatment, smoking, cholesterol levels, or estimated glomerular filtration rate. The adjusted heritability estimate was 42±8% (P<0.001). In a population-based sample with normal renal ultrasound, RRI normal values depend on sex, age, blood pressure, heart rate, and body mass index. The significant heritability of RRI suggests that genes influence this phenotype.

Authentication of vegetable oils by bulk and molecular carbon isotope analyses with emphasis on olive oil and pumpkin seed oil

The authenticity of vegetable oils consumed in Slovenia and Croatia was investigated by carbon isotope analysis of the individual fatty acids by the use of gas chromatography-combustion-isotope ratio mass spectrometry (GC/C/IRMS), and through carbon isotope analysis of the bulk oil. The fatty acids from samples of olive, pumpkin, sunflower, maize, rape, soybean, and sesame oils were separated by alkaline hydrolysis and derivatized to methyl esters for chemical characterization by capillary gas chromatography/mass spectrometry (GC/MS) prior to isotopic analysis. Enrichment in heavy carbon isotope (C-13) of th, bulk oil and of the individual fatty acids are related to (1) a thermally induced degradation during processing (deodorization, steam washing, or bleaching), (2) hydrolytic rancidity (lipolysis) and oxidative rancidity of the vegetable oils during storage, and (3) the potential blend with refined oil or other vegetable oils. The impurity or admixture of different oils may be assessed from the delta C-13(16:0) VS. delta C-13(18:1) covariations. The fatty acid compositions of Slovenian and Croatian olive oils are compared with those from the most important Mediterranean producer countries (Spain, Italy, Greece, and France).

Subgroup And Quality Of Life Analyses Of The Phase Iii Clinical Trial Of Novottf-100a Versus Best Standard Chemotherapy For Recurrent Glioblastoma

BACKGROUND: NovoTTF is a portable device delivering low-intensity, intermediate-frequency, electric fields using noninvasive, disposable scalp electrodes. These fields physically interfere with cell division. Preliminary studies in recurrent and newly diagnosed glioblastoma (GBM) have shown promising results. A phase III study in recurrent GBM has recently been concluded. METHODS: Adults (KPS ≥ 70%) with recurrent GBM (any recurrence) were randomized (stratified by surgery and center) to either NovoTTF administered continuously (20-24 hours/day, 7 days/week) or the best available chemotherapy (best physician choice [BPC]). Primary endpoint was overall survival (OS); 6-month progression-free survival (PFS6), 1-year survival, and QOL were secondary endpoints. RESULTS: Two hundred thirty-seven patients were randomized (28 centers in the United States and Europe) to either NovoTTF alone (120 patients) or BPC (117 patients). Patient characteristics were balanced, median age was 54 years (range, 23-80 years), median KPS was 80% (range, 50-100). One quarter had surgery for recurrence, and over half were at their second or more recurrence. A survival advantage for the device group was seen in patients treated according to protocol (median OS, 7.8 months vs. 6.1 months; n = 185; p = 0.01). Moreover, subgroup analysis in patients with better prognostic baseline characteristics (KPS ≥ 80%; age ≤ 60; 1st-3rd recurrence) demonstrated a robust survival benefit for NovoTTF patients compared to matched BPC patients (median OS, 8.8 months vs. 6.6 months; n = 110; p < 0.01). In this group, 1-year survival was 35% vs. 20% and PFS6 was 25.6% vs. 7.7%. Interestingly, in patients who failed bevacizumab prior to the trial, OS was also significantly extended by NovoTTF (4.4 months vs. 3.1 months; n = 23 vs. n = 21; p < 0.02). Quality of life was equivalent or superior in NovoTTF patients. CONCLUSIONS: NovoTTF, a noninvasive, novel cancer treatment modality shows significant therapeutic efficacy with improved quality of life. The impact of NovoTTF was more pronounced when patients with better baseline prognostic factors were treated. A large scale phase III clinical trial in newly diagnosed GBM is currently being conducted.