Les trois dernières guerres ont laissé dans le sous-sol français des millions de tonnes de munitions non explosées (en anglais, unexploded ordnance ou UXO) qui constituent un risque pour l'homme et l'environnement. La géophysique et particulièrement les méthodes magnétiques offrent des solutions pour localiser et caractériser ces objets qui sont le plus souvent aimantés. Mon travail de thèse consiste d'abord à réaliser une synthèse bibliographique des méthodes de recherche d'UXO en détaillant en particulier la méthode magnétique. Le dispositif de cartographie magnétique multi-capteur conçu par I'EOST et utilisé pour la recherche d'UXO subit un certain nombre de défauts. Ainsi (i) la fabrication d'un nouveau numériseur, (ii) l'acquisition d'un nouveau système de géolocalisation plus complet et (iii) l'utilisation d'un capteur supplémentaire afin de corriger les variations temporelles du champ permettent d'améliorer la qualité de nos cartographies. De même la mesure à différentes altitudes nous permet de gérer au mieux le temps imparti et le niveau de détail que l'on souhaite obtenir. Selon l'approximation habituellement employée en méthodes potentielles, l'anomalie du champ magnétique est négligeable par rapport au champ magnétique régional. En parallèle du travail de terrain décrit précédemment, une étude plus poussée du dipôle m'a permis d'évaluer que cette approximation entraîne une erreur non négligeable pouvant atteindre 16% de l'anomalie. Cette différence se ressent aussi sur les résultats obtenus par inversion. J'ai pu de plus développer un algorithme utilisant la méthode d'Euler et réévaluer la méthode que l'on utilisait déjà auparavant, basée sur le signal analytique. Après comparaison des différentes méthodes d'inversion, cette dernière semble le mieux contraindre la cible. / Last three wars have left millions of tons of unexploded ordnance (UXO) in French subsoil. They put the population and the environment at risk. Geophysics and particularly the magnetic methods present some solutions to characterise and locate those generally magnetised objects. The multi-sensor magnetic cartography device designed at EOST has some faults. Consequently, (i) the production of a new digitizer, (ii) the acquisition of a new geo-spatial positioning system and (iii) the use of an extra sensor to correct the time-related variations of the magnetic field allow us to improve our cartography quality. Likewise, multi-altitude measurement allow us to optimise the time in relation to the wanted cartography definition. According to the approximation generally used in magnetism theory, the anomaly is insignificant compared to the regional magnetic field. At the same time as my previously described field work, a thorough study of the dipole theory, allowed me to evaluate that this approximation results in a non-negligible mistake reaching 16% of the anomaly. The inversion results are sensible to this mistake. Moreover, l've developed an algorithm based on Euler deconvolution and l've re-evaluated the method already used which is based on analytic signal. After I compared some of the inversion methods, the one based on the analytic signal seems to ascertain the better the object characteristics.

Student Number : 9807515F - MSc Dissertation - School of Computational and Applied Mathematics - Faculty of Science / Statistical and algebraic techniques of subspace tracking were tested for filtering the earth’s response from airborne magnetic field gradients in order to discriminate the relatively small response (dipole) of objects on the earth’s surface, such as UXO. Filtering the data was not very effective with these methods but a subspace was found in the data for the magnitude of the magnetic moment of the dipole. This subspace is easily obtained using the singular value decomposition and can be used for an approximate location, without depth estimation, as well as the relative size of the dipole.

No description available.

Remediation of unexploded ordnance (UXO) and prioritization of excavation procedures for archaeological artifacts using electromagnetic (EM) induction are studied in this dissertation. Lowering of the false alarm rates that require excavation and artifact excavation prioritization can reduce the costs associated with unnecessary procedures. Data were taken over 5 areas at the San Jacinto Battleground near Houston, Texas, using an EM-63 metal detection instrument. The areas were selected using the archaeological concepts of cultural and natural formation processes applied to what is thought to be areas that were involved in the 1836 Battle of San Jacinto. Innovative use of a Statistical Point Pattern Analysis (PPA) is employed to identify clustering of EM anomalies. The K-function uses point {x,y} data to look for possible clusters in relation to other points in the data set. The clusters once identified using K-function will be further examined for classification and prioritization using the Weighted K-function. The Weighted K-function uses a third variable such as millivolt values or time decay to aid in segregation and prioritization of anomalies present. Once the anomalies of interest are identified, their locations are determined using the Gi-Statistics Technique. The Gi*-Statistic uses the individual Cartesian{x, y} points as origin locations to establish a range of distances to other cluster points in the data set. The segregation and location of anomalies supplied by this analysis will have several benefits. Prioritization of excavations will narrow down what areas should be excavated first. Anomalies of interest can be located to guide excavation procedures within the areas surveyed. Knowing what anomalies are of greater importance than others will help to lower false alarm rates for UXO remediation or for archaeological artifact selection. Knowing significant anomaly location will reduce the number of excavations which will subsequently save time and money. The procedures and analyses presented here are an interdisciplinary compilation of geophysics, archaeology and statistical analysis brought together for the first time to examine problems associated with UXO remediation as well as archaeological artifact selection at historic battlegrounds using electromagnetic data.

A workflow for the discrimination of multi-peak anomalies due to an unexploded ordnance (UXO) is presented. The effects of changes in the depth and orientation of a subsurface target on its TEM response are explored. Further, the variation of the peak separation, peak widths, maxima to minimum ratios and maxima values through time are modelled. Models of spatial and temporal variations are performed for both one and twotargets. These models are devoid of noise and focus primarily on the 3lb FLBGR bomb. It is found that the decay of the magnitudes of the anomaly maxima are related to the decay of the characteristic polarization curves of the UXO that creates them. The behaviour of the decay of the magnitudes of the maxima values of multi-peak anomalies is determined to be a good indication of target number in the subsurface. Observing these decays, it is possible to distinguish between two multi-peak scenarios, namely: a) a multipeak anomaly produced by a single UXO or two UXOs of the same type; and b) a multipeak anomaly produced by two UXOs of different types.

The dominant liberal peacebuilding critiques tends to focus on ‘states’ and the failure of interventions in rebuilding them. Consequently, a standardised critique has emerged largely because the critics apply a broad brush across a diverse range of contexts, programmes, issues and activities as illustrated by the lack of scrutiny on mine action and emerging contexts such as Somaliland. The liberal peacebuilding critics critique the standardised ‘one size fits all approach’ employed by interveners, yet they take the same approach. I therefore argue for the need to broaden the critique to include other elements and contexts of peacebuilding. I demonstrate that as an intervention mine action has intrinsic peacebuilding potential. However, the way mine action is implemented both globally and in Somaliland reflects the same dominant characteristics of the liberal peacebuilding critique i.e.; it is externally led; uses technical and standardised formulaic approaches; disregards local context thus failing to secure local ownership. Attributes that the critics argue have led to the failure and/or limited success of peacebuilding interventions. I therefore contend with the critics and demonstrate how these attributes have contributed to the challenges of implementing mine action activities thereby limiting mine actions ‘peace-ability’ potential in Somaliland. However, beyond the implementation modalities there are other factors that further contribute to limiting this potential; these include the Sector Actors; the Somaliland context i.e. the historical and political context, and the perception of Somaliland people. Thus in conclusion I argue for a nuanced critique that acknowledges the challenging realities of implementing programmes in challenging post conflict environments.

Unexploded Ordnance (UXO) contamination presents a considerable level of danger in almost all post-conflict environments. Globally there are a vast amount of casualties every year. However, accurate numbers of casualties is hard to obtain both globally and locally. The purpose of this thesis is to examine how UXO affects Human Security in Lao PDR. The methodology used is a theory consuming empirical and heuristic method. The thesis is a case study that does not attempt to generalize but to understand and analyze the relation between UXO and Human Security in the context of Lao PDR. The theory used in the thesis is Human Security based on the concept of the 1994 UNDP Human Development Report. The findings of the thesis make clear the connection between UXO contamination and lack of Human Security in Lao PDR. The direct and indirect consequences of UXO contamination are explored. Finally, the thesis works at a broader societal level where the links to poverty and development are illustrated.

L’objectif de ce travail est d’améliorer la caractérisation des sources à l’origine d’anomalies magnétiques, par le biais de développements théoriques et expérimentaux. Pour des structures invariantes dans une direction horizontale (structures à 2D), l’idée de Nabighian (1972) a été généralisée, ce qui implique une nouvelle façon d’étudier les méthodes potentielles à 2D. Ces développements ont permis de proposer une nouvelle approche théorique et de nouvelles représentations dans le plan complexe des fonctions magnétiques. La deuxième approche de ce travail est expérimentale. Une étude y est faite par le biais de données et de cas synthétiques, afin d’estimer l’apport d’acquisitions magnétiques à différentes altitudes pour caractériser une source. Dans un troisième temps, les développements théoriques et expérimentaux sont mis en œuvre pour deux exemples : des anomalies magnétiques en domaine océanique de type Vine et Mathews et une cartographie pour l’archéologie. / Magnetic anomalies recorded outside bodies provide high quality information relative to buried structures. By using theoretical and experimental developments, this thesis aims to improve the characterization of the source inducing the magnetic anomaly. Firstly, some properties of three dimensional magnetic functions are presented, for example the presence of several maxima of the analytic signal operator (3D) is demonstrated. The Nabighian (1972) equation is generalized, which imply a new process to study potential method in two dimensions. These developments enable a new visualization of the anomaly in the complex field. The second approach is experimental. Synthetic cases are used to estimate the contribution of different configurations of magnetic data acquisitions at different altitudes to characterize the magnetic source. Theoretical and experimental developments are finally applied to two field examples: oceanic magnetic anomalies and archaeological magnetic prospection

The dominant liberal peacebuilding critiques tends to focus on ‘states’ and the failure of interventions in rebuilding them. Consequently, a standardised critique has emerged largely because the critics apply a broad brush across a diverse range of contexts, programmes, issues and activities as illustrated by the lack of scrutiny on mine action and emerging contexts such as Somaliland. The liberal peacebuilding critics critique the standardised ‘one size fits all approach’ employed by interveners, yet they take the same approach. I therefore argue for the need to broaden the critique to include other elements and contexts of peacebuilding. I demonstrate that as an intervention mine action has intrinsic peacebuilding potential. However, the way mine action is implemented both globally and in Somaliland reflects the same dominant characteristics of the liberal peacebuilding critique i.e.; it is externally led; uses technical and standardised formulaic approaches; disregards local context thus failing to secure local ownership. Attributes that the critics argue have led to the failure and/or limited success of peacebuilding interventions. I therefore contend with the critics and demonstrate how these attributes have contributed to the challenges of implementing mine action activities thereby limiting mine actions ‘peace-ability’ potential in Somaliland. However, beyond the implementation modalities there are other factors that further contribute to limiting this potential; these include the Sector Actors; the Somaliland context i.e. the historical and political context, and the perception of Somaliland people. Thus in conclusion I argue for a nuanced critique that acknowledges the challenging realities of implementing programmes in challenging post conflict environments.

The theme of this dissertation is to advance safety hazard mitigation by detecting and characterizing hidden targets of concern. Ground-penetrating radar (GPR) is used to detect and characterize hidden targets that pose safety hazards at Earth's surface, within shallow soil, and within rock. The resulting images detect unexploded ordnance (UXO) and detect fractures that pose collapse hazards in a mine. Detecting and characterizing fractures and voids within rock prior to excavation can enable mitigation of mine collapse hazards. GPR data were acquired on the wall of a pillar in an underground mine. Strong radar reflections in the field data correlate with fractures and a cave exposed on the pillar walls. Pillar wall roughness was included in migration, a wavefield imaging algorithm, to quantitatively locate fractures and voids and map their spatial relationships within the rock. Quantifying the radar reflection amplitudes enabled mapping the distance between fracture walls. Detecting and characterizing UXO and landmines from a safe distance can enable de-mining. Migration was used to improve GPR imaging for unmanned aerial vehicle (UAV) data acquisitions. Existing algorithms were adapted for UAV flight irregularities and surface topography, and a new algorithm was developed that does not depend on the unknown soil wavespeed. Errors associated with wavespeed and raypath assumptions were quantified. The algorithms were tested with real and synthetic datasets. The improved and new algorithms are more successful than previous algorithms. To detect linear targets at all orientations, fully polarized GPR data are needed. Polarity combinations were investigated to optimize the detection of surface and subsurface small targets and linear targets. Scattering caused by topographic roughness is the primary shallow subsurface noise. For subsurface targets, detection is optimized by migration plus a polarity combination that captures all scattered energy. Strong reflection and scattering from the air-ground boundary can hide surface targets. Detection is optimized by removing the strong isotropic surface scattering, imaging targets by their anisotropic scattering. / Doctor of Philosophy / The theme of this dissertation is to advance safety hazard mitigation by detecting and characterizing hidden targets of concern. Ground-penetrating radar (GPR) is used to detect and characterize hidden targets that pose safety hazards at Earth's surface, within shallow soil, and within rock. The resulting images detect unexploded ordnance (UXO)/landmines and detect fractures that pose collapse hazards in a mine. Detecting and characterizing fractures and voids within rock prior to mining can enable mitigation of mine collapse hazards. GPR data were acquired on the wall of a pillar in an underground mine. Strong radar reflections in the field data correlate with fractures and a cave exposed on the pillar walls. Pillar wall roughness was included in migration, a wavefield imaging algorithm, to quantitatively locate fractures and voids and map their spatial relationships within the rock. Quantifying the radar reflection amplitudes enabled mapping the distance between fracture walls. Detecting and characterizing UXO, landmines from a safe distance can enable de-mining. Migration was used to improve GPR imaging for an unmanned aerial vehicle (drone) data acquisition. Existing algorithms were adapted for drone flight irregularities and surface topography, and a new algorithm was developed that does not depend on the unknown soil properties. Errors associated with the algorithms' assumptions were quantified. The algorithms were tested with real and computer-generated datasets. The improved and new algorithms are more successful than previous algorithms. To detect all targets regardless of their orientation, GPR data need to be acquired with antenna pointing in multiple directions (different polarities). Polarity combinations were investigated to optimize the detection of surface and subsurface small targets and linear targets. Scattering caused by topographic roughness is the primary shallow subsurface noise. For subsurface targets, detection is optimized by migration plus a polarity combination that captures all scattered energy. Strong radar reflection from the air-ground boundary can hide surface targets. Detection is optimized by removing the strong ground surface from the data, and imaging targets by differences in their radar scattering.