Machine learning algorithms for spatial data. Case studies: Environmental pollution, natural hazards, renewable resources

This paper presents general problems and approaches for the spatial data analysis using machine learning algorithms. Machine learning is a very powerful approach to adaptive data analysis, modelling and visualisation. The key feature of the machine learning algorithms is that they learn from empirical data and can be used in cases when the modelled environmental phenomena are hidden, nonlinear, noisy and highly variable in space and in time. Most of the machines learning algorithms are universal and adaptive modelling tools developed to solve basic problems of learning from data: classification/pattern recognition, regression/mapping and probability density modelling. In the present report some of the widely used machine learning algorithms, namely artificial neural networks (ANN) of different architectures and Support Vector Machines (SVM), are adapted to the problems of the analysis and modelling of geo-spatial data. Machine learning algorithms have an important advantage over traditional models of spatial statistics when problems are considered in a high dimensional geo-feature spaces, when the dimension of space exceeds 5. Such features are usually generated, for example, from digital elevation models, remote sensing images, etc. An important extension of models concerns considering of real space constrains like geomorphology, networks, and other natural structures. Recent developments in semi-supervised learning can improve modelling of environmental phenomena taking into account on geo-manifolds. An important part of the study deals with the analysis of relevant variables and models' inputs. This problem is approached by using different feature selection/feature extraction nonlinear tools. To demonstrate the application of machine learning algorithms several interesting case studies are considered: digital soil mapping using SVM, automatic mapping of soil and water system pollution using ANN; natural hazards risk analysis (avalanches, landslides), assessments of renewable resources (wind fields) with SVM and ANN models, etc. The dimensionality of spaces considered varies from 2 to more than 30. Figures 1, 2, 3 demonstrate some results of the studies and their outputs. Finally, the results of environmental mapping are discussed and compared with traditional models of geostatistics.

Case studies and mathematical models of ecological speciation. 1. Cichlids in a crater lake.

A recent study of a pair of sympatric species of cichlids in Lake Apoyo in Nicaragua is viewed as providing probably one of the most convincing examples of sympatric speciation to date. Here, we describe and study a stochastic, individual-based, explicit genetic model tailored for this cichlid system. Our results show that relatively rapid (<20,000 generations) colonization of a new ecological niche and (sympatric or parapatric) speciation via local adaptation and divergence in habitat and mating preferences are theoretically plausible if: (i) the number of loci underlying the traits controlling local adaptation, and habitat and mating preferences is small; (ii) the strength of selection for local adaptation is intermediate; (iii) the carrying capacity of the population is intermediate; and (iv) the effects of the loci influencing nonrandom mating are strong. We discuss patterns and timescales of ecological speciation identified by our model, and we highlight important parameters and features that need to be studied empirically to provide information that can be used to improve the biological realism and power of mathematical models of ecological speciation.

Determining Pu-241 in environmental samples: case studies in alpine soils

A procedure was developed for determining Pu-241 activity in environmental samples. This beta emitter isotope of plutonium was measured by ultra low level liquid scintillation, after several separation and purification steps that involved the use of a highly selective extraction chromatographic resin (Eichrom-TEVA). Due to the lack of reference material for Pu-241, the method was nevertheless validated using four IAEA reference sediments with information values for Pu-241. Next, the method was used to determine the Pu-241 activity in alpine soils of Switzerland and France. The Pu-241/Pu-239,Pu-240 and Pu-238/Pu-239,Pu-240 activity ratios confirmed that Pu contamination in the tested alpine soils originated mainly from global fallout from nuclear weapon tests conducted in the fifties and sixties. Estimation of the date of the contamination, using the Pu-241/Am-241 age-dating method, further confirmed this origin. However, the Pu-241/Am-241 dating method was limited to samples where Pu-Am fractionation was insignificant. If any, the contribution of the Chernobyl accident is negligible.

Machine learning for geospatial data : algorithms, software tools and case studies

Résumé Cette thèse est consacrée à l'analyse, la modélisation et la visualisation de données environnementales à référence spatiale à l'aide d'algorithmes d'apprentissage automatique (Machine Learning). L'apprentissage automatique peut être considéré au sens large comme une sous-catégorie de l'intelligence artificielle qui concerne particulièrement le développement de techniques et d'algorithmes permettant à une machine d'apprendre à partir de données. Dans cette thèse, les algorithmes d'apprentissage automatique sont adaptés pour être appliqués à des données environnementales et à la prédiction spatiale. Pourquoi l'apprentissage automatique ? Parce que la majorité des algorithmes d'apprentissage automatiques sont universels, adaptatifs, non-linéaires, robustes et efficaces pour la modélisation. Ils peuvent résoudre des problèmes de classification, de régression et de modélisation de densité de probabilités dans des espaces à haute dimension, composés de variables informatives spatialisées (« géo-features ») en plus des coordonnées géographiques. De plus, ils sont idéaux pour être implémentés en tant qu'outils d'aide à la décision pour des questions environnementales allant de la reconnaissance de pattern à la modélisation et la prédiction en passant par la cartographie automatique. Leur efficacité est comparable au modèles géostatistiques dans l'espace des coordonnées géographiques, mais ils sont indispensables pour des données à hautes dimensions incluant des géo-features. Les algorithmes d'apprentissage automatique les plus importants et les plus populaires sont présentés théoriquement et implémentés sous forme de logiciels pour les sciences environnementales. Les principaux algorithmes décrits sont le Perceptron multicouches (MultiLayer Perceptron, MLP) - l'algorithme le plus connu dans l'intelligence artificielle, le réseau de neurones de régression généralisée (General Regression Neural Networks, GRNN), le réseau de neurones probabiliste (Probabilistic Neural Networks, PNN), les cartes auto-organisées (SelfOrganized Maps, SOM), les modèles à mixture Gaussiennes (Gaussian Mixture Models, GMM), les réseaux à fonctions de base radiales (Radial Basis Functions Networks, RBF) et les réseaux à mixture de densité (Mixture Density Networks, MDN). Cette gamme d'algorithmes permet de couvrir des tâches variées telle que la classification, la régression ou l'estimation de densité de probabilité. L'analyse exploratoire des données (Exploratory Data Analysis, EDA) est le premier pas de toute analyse de données. Dans cette thèse les concepts d'analyse exploratoire de données spatiales (Exploratory Spatial Data Analysis, ESDA) sont traités selon l'approche traditionnelle de la géostatistique avec la variographie expérimentale et selon les principes de l'apprentissage automatique. La variographie expérimentale, qui étudie les relations entre pairs de points, est un outil de base pour l'analyse géostatistique de corrélations spatiales anisotropiques qui permet de détecter la présence de patterns spatiaux descriptible par une statistique. L'approche de l'apprentissage automatique pour l'ESDA est présentée à travers l'application de la méthode des k plus proches voisins qui est très simple et possède d'excellentes qualités d'interprétation et de visualisation. Une part importante de la thèse traite de sujets d'actualité comme la cartographie automatique de données spatiales. Le réseau de neurones de régression généralisée est proposé pour résoudre cette tâche efficacement. Les performances du GRNN sont démontrées par des données de Comparaison d'Interpolation Spatiale (SIC) de 2004 pour lesquelles le GRNN bat significativement toutes les autres méthodes, particulièrement lors de situations d'urgence. La thèse est composée de quatre chapitres : théorie, applications, outils logiciels et des exemples guidés. Une partie importante du travail consiste en une collection de logiciels : Machine Learning Office. Cette collection de logiciels a été développée durant les 15 dernières années et a été utilisée pour l'enseignement de nombreux cours, dont des workshops internationaux en Chine, France, Italie, Irlande et Suisse ainsi que dans des projets de recherche fondamentaux et appliqués. Les cas d'études considérés couvrent un vaste spectre de problèmes géoenvironnementaux réels à basse et haute dimensionnalité, tels que la pollution de l'air, du sol et de l'eau par des produits radioactifs et des métaux lourds, la classification de types de sols et d'unités hydrogéologiques, la cartographie des incertitudes pour l'aide à la décision et l'estimation de risques naturels (glissements de terrain, avalanches). Des outils complémentaires pour l'analyse exploratoire des données et la visualisation ont également été développés en prenant soin de créer une interface conviviale et facile à l'utilisation. Machine Learning for geospatial data: algorithms, software tools and case studies Abstract The thesis is devoted to the analysis, modeling and visualisation of spatial environmental data using machine learning algorithms. In a broad sense machine learning can be considered as a subfield of artificial intelligence. It mainly concerns with the development of techniques and algorithms that allow computers to learn from data. In this thesis machine learning algorithms are adapted to learn from spatial environmental data and to make spatial predictions. Why machine learning? In few words most of machine learning algorithms are universal, adaptive, nonlinear, robust and efficient modeling tools. They can find solutions for the classification, regression, and probability density modeling problems in high-dimensional geo-feature spaces, composed of geographical space and additional relevant spatially referenced features. They are well-suited to be implemented as predictive engines in decision support systems, for the purposes of environmental data mining including pattern recognition, modeling and predictions as well as automatic data mapping. They have competitive efficiency to the geostatistical models in low dimensional geographical spaces but are indispensable in high-dimensional geo-feature spaces. The most important and popular machine learning algorithms and models interesting for geo- and environmental sciences are presented in details: from theoretical description of the concepts to the software implementation. The main algorithms and models considered are the following: multi-layer perceptron (a workhorse of machine learning), general regression neural networks, probabilistic neural networks, self-organising (Kohonen) maps, Gaussian mixture models, radial basis functions networks, mixture density networks. This set of models covers machine learning tasks such as classification, regression, and density estimation. Exploratory data analysis (EDA) is initial and very important part of data analysis. In this thesis the concepts of exploratory spatial data analysis (ESDA) is considered using both traditional geostatistical approach such as_experimental variography and machine learning. Experimental variography is a basic tool for geostatistical analysis of anisotropic spatial correlations which helps to understand the presence of spatial patterns, at least described by two-point statistics. A machine learning approach for ESDA is presented by applying the k-nearest neighbors (k-NN) method which is simple and has very good interpretation and visualization properties. Important part of the thesis deals with a hot topic of nowadays, namely, an automatic mapping of geospatial data. General regression neural networks (GRNN) is proposed as efficient model to solve this task. Performance of the GRNN model is demonstrated on Spatial Interpolation Comparison (SIC) 2004 data where GRNN model significantly outperformed all other approaches, especially in case of emergency conditions. The thesis consists of four chapters and has the following structure: theory, applications, software tools, and how-to-do-it examples. An important part of the work is a collection of software tools - Machine Learning Office. Machine Learning Office tools were developed during last 15 years and was used both for many teaching courses, including international workshops in China, France, Italy, Ireland, Switzerland and for realizing fundamental and applied research projects. Case studies considered cover wide spectrum of the real-life low and high-dimensional geo- and environmental problems, such as air, soil and water pollution by radionuclides and heavy metals, soil types and hydro-geological units classification, decision-oriented mapping with uncertainties, natural hazards (landslides, avalanches) assessments and susceptibility mapping. Complementary tools useful for the exploratory data analysis and visualisation were developed as well. The software is user friendly and easy to use.

Dating human skeletal remains using 90Sr and 210Pb: case studies. [Dating human skeletal remains using Sr-90 and Pb-210: Case studies]

In legal medicine, the post mortem interval (PMI) of interest covers the last 50 years. When only human skeletal remains are found, determining the PMI currently relies mostly on the experience of the forensic anthropologist, with few techniques available to help. Recently, several radiometric methods have been proposed to reveal PMI. For instance, (14)C and (90)Sr bomb pulse dating covers the last 60 years and give reliable PMI when teeth or bones are available. (232)Th series dating has also been proposed but requires a large amount of bones. In addition, (210)Pb dating is promising but is submitted to diagenesis and individual habits like smoking that must be handled carefully. Here we determine PMI on 29 cases of forensic interest using (90)Sr bomb pulse. In 12 cases, (210)Pb dating was added to narrow the PMI interval. In addition, anthropological investigations were carried out on 15 cases to confront anthropological expertise to the radiometric method. Results show that 10 of the 29 cases can be discarded as having no forensic interest (PMI>50 years) based only on the (90)Sr bomb pulse dating. For 10 other cases, the additional (210)Pb dating restricts the PMI uncertainty to a few years. In 15 cases, anthropological investigations corroborate the radiometric PMI. This study also shows that diagenesis and inter-individual difference in radionuclide uptake represent the main sources of uncertainty in the PMI determination using radiometric methods.

Flow-R, a model for susceptibility mapping of debris flows and other gravitational hazards at a regional scale: some case studies

Every year, debris flows cause huge damage in mountainous areas. Due to population pressure in hazardous zones, the socio-economic impact is much higher than in the past. Therefore, the development of indicative susceptibility hazard maps is of primary importance, particularly in developing countries. However, the complexity of the phenomenon and the variability of local controlling factors limit the use of processbased models for a first assessment. A debris flow model has been developed for regional susceptibility assessments using digital elevation model (DEM) with a GIS-based approach.. The automatic identification of source areas and the estimation of debris flow spreading, based on GIS tools, provide a substantial basis for a preliminary susceptibility assessment at a regional scale. One of the main advantages of this model is its workability. In fact, everything is open to the user, from the data choice to the selection of the algorithms and their parameters. The Flow-R model was tested in three different contexts: two in Switzerland and one in Pakistan, for indicative susceptibility hazard mapping. It was shown that the quality of the DEM is the most important parameter to obtain reliable results for propagation, but also to identify the potential debris flows sources.

Assessing sedimentary evolution by means of Sr-isotope ratios : 3 case studies on the Caribbean plate : (Cretaceous: Nicoya Peninsula, Costa Rica, Tertiary: Hess Rise, and La Désirade, Guadeloupe, France)

The understanding of sedimentary evolution is intimately related to the knowledge of the exact ages of the sediments. When working on carbonate sediments, age dating is commonly based on paleontological observations and established biozonations, which may prove to be relatively imprecise. Dating by means of strontium isotope ratios in marine bioclasts is the probably best method in order to precisely date carbonate successions, provided that the sample reflects original marine geochemical characteristics. This requires a precise study of the samples including its petrography, SEM and cathodoluminescence observations, stable carbon and oxygen isotope geochemistry and finally the strontium isotope measurement itself. On the Nicoya Peninsula (Northwestern Costa Rica) sediments from the Piedras Blancas Formation, Nambi Formation and Quebrada Pavas Formation were dated by the means of strontium isotope ratios measured in Upper Cretaceous Inoceramus shell fragments. Results have shown average 87Sr/86Sr values of 0.707654 (middle late Campanian) for the Piedras Blancas Formation, 0.707322 (Turonian-Coniacian) for the Nambi Formation and 0.707721 (late Campanian-Maastrichtian) for the Quebrada Pavas Formation. Abundant detrital components in the studied formations constitute a difficulty to strontium isotope dating. In fact, the fossil bearing sediments can easily contaminate the target fossil with strontium mobilized form basalts during diagenesis and thus the obtained strontium isotope ratios may be influenced significantly and so will the obtained ages. The new and more precise age assignments allow for more precision in the chronostratigraphic chart of the sedimentary and tectonic evolution of the Nicoya Peninsula, providing a better insight on the evolution of this region. Meteor Cruise M81 dredged shallow water carbonates from the Hess Rise and Hess Escarpment during March 2010. Several of these shallow water carbonates contain abundant Larger Foraminifera that indicates an Eocene-Oligocene age. In this study the strontium isotope values ranging from 0.707847 to 0.708238 can be interpreted as a Rupelian to Chattian age of these sediments. These platform sediments are placed on seamounts, now located at depths reaching 1600 m. Observation of sedimentologic characteristics of these sediments has helped to resolve apparent discrepancies between fossil and strontium isotope ages. Hence, it is possible to show that the subsidence was active during early Miocene times. On La Désirade (Guadeloupe France), the Neogene to Quaternary carbonate cover has been dated by microfossils and some U/Th-ages. Disagreements subsisted in the paleontological ages of the formations. Strontium isotope ratios ranging from 0.709047 to 0.709076 showed the Limestone Table of La Désirade to range from an Early Pliocene to Late Pliocene/early Pleistocene age. A very late Miocene age (87Sr/86Sr =0.709013) can be determined to the Detrital Offshore Limestone. The flat volcanic basement had to be eroded by wave-action during a long-term stable relative sea-level. Sediments of the Table Limestone on La Désirade show both low-stand and high-stand facies that encroach on the igneous basement, implying deposition during a major phase of subsidence creating accommodation space. Subsidence is followed by tectonic uplift documented by fringing reefs and beach rocks that young from the top of the Table Limestone (180 m) towards the present coastline. Strontium isotope ratios from two different fringing reefs (0.707172 and 0.709145) and from a beach rock (0.709163) allow tentative dating, (125ky, ~ 400ky, 945ky) and indicate an uplift rate of about 5cm/ky for this time period of La Désirade Island. The documented subsidence and uplift history calls for a new model of tectonic evolution of the area.

Constraints on the upper Permian to upper Triassic marine carbon isotope curve : case studies from the Tethys

Secular variations of the seawater carbon isotopic composition provide evidence for paleoceanographic and paleoclimatic changes and may serve for chemiostratigraphic correlations. The present study aimed to improve the current knowledge on the Upper Permian and Triassic segment of the Phanerozoic marine carbon isotope curve, whose Triassic part was poorly constrained by previous studies. Profiles of inorganic carbon isotopes are provided for sections from Himalaya (Salt Range, Kashmir, Spiti and Nepal), Oman and North Dobrogea (Romania) on the basis of whole-rock carbonate analysis. The data acquired, together with a literature compilation confirmed that most of the Upper Permian is characterized by high δ13C values (averaging +40/00) but failed to detect a positive excursion as suggested by recent compilations. In the light of these observations, the large drop in δ13C values associated with the end-Permian mass extinction appears to be driven by a sudden transfer of previously stocked 13C depleted carbon, rather than by the overturn of a Late Permian stratified ocean. The Triassic data-set outlines significant secular variations. The best documented is a carbon isotope positive excursion just across the Lower-Middle Triassic boundary, globally developed since it was detected in various paleogeographic settings. It is interpreted to reflect variations in surface ocean chemistry, possibly related to increased primary productivity, at times when the biotic recovery after the end-Permian mass-extinction began to accelerate significantly and when a sharp rise in seawater δ34S values occurred globally. Strontium isotope data obtained from well preserved biogenic phosphates allow a refinement of the Middle Triassic segment of the seawater strontium isotope curve and show a major inflexion point of the seawater strontium isotope curve also near the Lower Triassic - Middle Triassic boundary. These facts suggest that the transition from the Early to the Middle Triassic was a time of revolutionary global change which represented an important step in the evolution of Mesozoic marine environments. A tentative carbon isotope curve for the Upper Permian to Upper Triassic time interval is proposed. Its major features are: ? high but constant δ13C values during the Late Permian ? a sharp drop in δ13C values in the latest Permian ? subsequent recovery of δ13C values ? a short-lived positive excursion across the Early-Middle Triassic boundary ? a gradual rise in δ13C values starting in the Late Ladinian or in the Early Carnian It is foreseen that these fluctuations of the carbon isotope curve may serve as chronostratigraphic markers and further assist in the correlation of Permian and Triassic carbonate deposits.

Young infants' triangular communication with their parents in the context of maternal postpartum psychosis: Four case studies

With increasing data on the dynamics of normative couples as they transition to parenthood and become a triad, the need for greater understanding of the impact of parental psychopathology on this transition has become clear. The goal of the current article is to begin exploring this area that has received little attention to date, by describing case examples from a study of clinical families as they transitioned to parenthood. Four representative cases were selected from a pool of 13 mother-father-baby triads, for whom the mother had been hospitalized conjointly with her infant due to a psychotic episode during the postpartum period. The families were observed as part of a clinical consultation that included a semistructured play paradigm known as the Lausanne Trilogue Play (LTP; E. Fivaz-Depeursinge, & A. Corboz-Warnery, 1999). Interactions were scored using standardized measures as well as clinical impressions. All families from the clinical sample were noted to struggle and frequently failed to achieve the goals of play. The impact on the infants in terms of their developing sense of self as well as their defensive strategies in this context are discussed, with clinical implications explored.

Determining 241Pu in environmental samples: case studies in alpine soils.

A procedure was developed for determining 241Pu activity in environmental samples. This beta emitter isotope of plutonium was measured by ultra low level liquid scintillation, after several separation and purification steps that involved the use of a highly selective extraction chromatographic resin (Eichrom-TEVA). Due to the lack of reference material for 241Pu, the method was nevertheless validated using four IAEA reference sediments with information values for 241Pu. Next, the method was used to determine the 241Pu activity in alpine soils of Switzerland and France. The 241Pu/239,240Pu and 238Pu/239,240Pu activity ratios confirmed that Pu contamination in the tested alpine soils originated mainly from global fallout from nuclear weapon tests conducted in the fifties and sixties. Estimation of the date of the contamination, using the 241Pu/241Am age-dating method, further confirmed this origin. However, the 241Pu/241Am dating method was limited to samples where Pu-Am fractionation was insignificant. If any, the contribution of the Chernobyl accident is negligible.

Workshop: Capacity building for national public health advocacy: case studies from national associations

The first statement of the EUPHA on the Future of Public Health in Europe refers to the need for going 'to policymakers, politicians and practitioners in all sectors of society and advise them on how to promote public health throughout society'. WHO-EURO Director General Marc Danzon, quoted in the second EUPHA statement on the responsibility of policy makers indicates that 'learning is not systematically applied in health policy development in our continent'. Statement 3 calls for the integration of public health into the political agenda in all sectors. The first EUPHA president, Louise Gunning-Schepers, quoted in Statement 10 called on EUPHA to become 'a powerful advocate of the public health community'. In addition to the above, the EU is now actively seeking ways to build capacity to implement its health strategy. Learning and building the capacity to achieve our aims The aims and objectives to promote the public's health as reflected in EUPHA's 10 statements are also mirrored in the national public health associations. However, many of EUPHA's national associations have little or limited experience in promoting public health policy at the national level. To assist in the learning of advocacy for public health policies, case studies presenting experiences of national public health organizations in lobbying for national public health policy will be presented and discussed. In addition to sharing experiences, the presentations will identify successful approaches to public health advocacy as well as lessons learned from unsuccessful attempts.