An increasing trend in both the number of disasters and affected people has been observed, especially during the second half of the 20th century. The physical, economic and social impact that natural hazards have had on a global scale has prompted an increasing interest of governments, international institutions and the academia. This has immensely contributed to improve the knowledge on the subject and has helped multiply the number of initiatives to reduce the negative consequences of natural hazards on people. The scale on which studies supporting disaster risk reduction (DRR) actions are performed is a critical parameter. Given that disasters are recognized to be place-dependent, studying the geographical distribution of disasters on a local scale is essential to make DRR practical and feasible for local authorities, organizations and civilians. However, studying disasters on the local scale is still a challenge due to the constraints posed by scarce data availability. Social vulnerability in many disaster-prone areas is however a pressing issue that needs to be swiftly addressed despite of the many limitations of data for such studies. This thesis explored methodological alternatives to study the geographical distribution of natural disasters and their potential causes in disaster-prone and data-scarce areas. The Samala River catchment in Guatemala was selected as a case study, which is representative of areas with high social vulnerability and data scarcity.  Exploratory methods to derive critical disaster information in such areas were constructed using the geographical and social data available for the study area. The hindrances posed by the available data were evaluated and the use of non-traditional datasets such as nightlights imagery to complement the available data were explored as a way of overcoming the observed limitations. The exploratory methods developed in this thesis aim at (a) deriving information on natural disasters under data-scarce circumstances, (b) exploring the correlation between the spatial distribution of natural disasters and the physical context in order to look for causalities, (c) using open data to study the social context as a potential cause of disasters in data-scarce areas, and (d) mapping vulnerabilities to support actions for disaster risk reduction. Although the available data for the case study was limited in quantity and quality and many sources of uncertainty exist in the proposed methods, this thesis argues that the potential contribution to the development of DRR on a local scale is more important than the identified drawbacks. The use of non-traditional data such as remotely sensed imagery made it possible to derive information on the occurrences of disasters and, in particular, causal relationships between location of disasters and their physical and social context. / El número de desastres y personas afectadas por esos desastres en el mundo han mostrado una tendencia creciente, especialmente en la segunda mitad del siglo veinte. El impacto físico, económico y social que las amenazas naturales han causado a nivel global ha causado que gobiernos, instituciones internacionales y la academia se interesen cada vez más en los desastres causados por esas amenazas. Este interés ha contribuido a mejorar el conocimiento existente sobre desastres y ha contribuido a multiplicar las iniciativas orientadas a reducir sus efectos negativos en las personas. La escala en la cual las iniciativas para la reducción del riesgo de desastres (RRD) se llevan a cabo es un parámetro crítico para su materialización. Hoy en día se reconoce la estrecha relación que existe entre los desastres y los lugares donde éstos se registran. Por esta razón, estudiar la distribución de los desastres en una escala local es esencial para que la RRD sea práctica y factible para autoridades y organizaciones locales, y también para la sociedad civil. Sin embargo, estudiar los desastres en una escala local es aún un problema por resolver debido a las restricciones impuestas por la escasa disponibilidad de datos de alta resolución. A pesar de las dificultades y limitaciones identificadas, la vulnerabilidad social en las regiones propensas a desastres es un problema importante que necesita ser atendido con prontitud. La presente tesis exploró alternativas metodológicas para estudiar la distribución geográfica de los desastres naturales y sus causas potenciales, particularmente en áreas propensas a desastres y en condiciones de información limitada. La cuenca del Río Samalá fue seleccionada como caso de estudio debido a que es un área representativa de áreas propensa a desastres con alta vulnerabilidad social y además escasez de datos. El trabajo de investigación propone métodos exploratorios para extraer información crítica sobre desastres utilizando la información geográfica y social que esté disponible, evaluando los obstáculos impuestos por la reducida disponibilidad de datos. La información existente fue complementada con el uso de fuentes de información no tradicional, e.g. imágenes satelitales de luces nocturnas, como una manera de superar las limitaciones identificadas. Los métodos desarrollados en este trabajo de tesis tuvieron como objetivos (a) obtener información sobre desastres naturales en condiciones de escasez de datos, (b) explorar la correlación entre la distribución espacial de los desastres naturales y su contexto físico para identificar causalidades, (c) utilizar información de libre acceso para estudiar el contexto social de los desastres como causa potencial de los desastres en áreas con escasez de datos, y (d) mapear vulnerabilidades para sustentar acciones para la RRD. Este trabajo de tesis sostiene que la contribución potencial de los métodos propuestos al desarrollo de la RRD en la escala social es más importante que las incertidumbres que implican y las limitaciones creadas por la reducida calidad y cantidad de información para el caso de estudio. El uso de fuentes de información no tradicionales tales como imágenes satelitales hizo posible incrementar la información sobre las incidencias de desastres y, en particular, buscar relación de dependencia entre los lugares particulares en los que los desastres fueron registrados y su contexto físico y social.

The languages spoken in developing countries are diverse and most are currently under-resourced from an automatic speech recognition (ASR) perspective. In South Africa alone, 10 of the 11 official languages belong to this category. Given the potential for future applications of speech-based information systems such as spoken dialog system (SDSs) in these countries, the design of minimal ASR audio corpora is an important research area. Specifically, current ASR systems utilise acoustic models to represent acoustic variability, and effective ASR corpus design aims to optimise the amount of relevant variation within training data while minimising the size of the corpus. Therefore an investigation of the effect that different amounts and types of training data have on these models is needed. With this dissertation specific consideration is given to the data sufficiency principals that apply to the training of acoustic models. The investigation of this task lead to the following main achievements: 1) We define a new stability measurement protocol that provides the capability to view the variability of ASR training data. 2) This protocol allows for the investigation of the effect that various acoustic model complexities and ASR normalisation techniques have on ASR training data requirements. Specific trends with regard to the data requirements for different phone categories and how these are affected by various modelling strategies are observed. 3) Based on this analysis acoustic distances between phones are estimated across language borders, paving the way for further research in cross-language data sharing. Finally the knowledge obtained from these experiments is applied to perform a data sufficiency analysis of a new speech recognition corpus of South African languages: The Lwazi ASR corpus. The findings correlate well with initial phone recognition results and yield insight into the sufficient number of speakers required for the development of minimal telephone ASR corpora. / Thesis (M. Ing. (Computer and Electronical Engineering))--North-West University, Potchefstroom Campus, 2009.

In the world of human language technology, resource–scarce languages (RSLs) suffer from the problem of little available electronic data and linguistic expertise. The Lwazi project in South Africa is a large–scale endeavour to collect and apply such resources for all eleven of the official South African languages. One of the deliverables of the project is more natural text–to–speech (TTS) voices. Naturalness is primarily determined by prosody and it is shown that many aspects of prosodic modelling is, in turn, dependent on part–of–speech (POS) information. Solving the POS problem is, therefore, a prudent first step towards meeting the goal of natural TTS voices. In a resource–scarce environment, obtaining and applying the POS information are not trivial. Firstly, an automatic tagger is required to tag the text to be synthesised with POS categories, but state–of–the–art POS taggers are data–driven and thus require large amounts of labelled training data. Secondly, the subsequent processes in TTS that are used to apply the POS information towards prosodic modelling are resource–intensive themselves: some require non–trivial linguistic knowledge; others require labelled data as well. The first problem asks the question of which available POS tagging algorithm will be the most accurate on little training data. This research sets out to answer the question by reviewing the most popular supervised data–driven algorithms. Since literature to date consists mostly of isolated papers discussing one algorithm, the aim of the review is to consolidate the research into a single point of reference. A subsequent experimental investigation compares the tagging algorithms on small training data sets of English and Afrikaans, and it is shown that the hidden Markov model (HMM) tagger outperforms the rest when using both a comprehensive and a reduced POS tagset. Regarding the second problem, the question arises whether it is perhaps possible to circumvent the traditional approaches to prosodic modelling by learning the latter directly from the speech data using POS information. In other words, does the addition of POS features to the HTS context labels improve the naturalness of a TTS voice? Towards answering this question, HTS voices are trained from English and Afrikaans prosodically rich speech. The voices are compared with and without POS features incorporated into the HTS context labels, analytically and perceptually. For the analytical experiments, measures of prosody to quantify the comparisons are explored. It is then also noted whether the results of the perceptual experiments correlate with their analytical counterparts. It is found that, when a minimal feature set is used for the HTS context labels, the addition of POS tags does improve the naturalness of the voice. However, the same effect can be accomplished by including segmental counting and positional information instead of the POS tags. / Thesis (M.Sc. Engineering Sciences (Electrical and Electronic Engineering))--North-West University, Potchefstroom Campus, 2011.

The languages spoken in developing countries are diverse and most are currently under-resourced from an automatic speech recognition (ASR) perspective. In South Africa alone, 10 of the 11 official languages belong to this category. Given the potential for future applications of speech-based information systems such as spoken dialog system (SDSs) in these countries, the design of minimal ASR audio corpora is an important research area. Specifically, current ASR systems utilise acoustic models to represent acoustic variability, and effective ASR corpus design aims to optimise the amount of relevant variation within training data while minimising the size of the corpus. Therefore an investigation of the effect that different amounts and types of training data have on these models is needed. With this dissertation specific consideration is given to the data sufficiency principals that apply to the training of acoustic models. The investigation of this task lead to the following main achievements: 1) We define a new stability measurement protocol that provides the capability to view the variability of ASR training data. 2) This protocol allows for the investigation of the effect that various acoustic model complexities and ASR normalisation techniques have on ASR training data requirements. Specific trends with regard to the data requirements for different phone categories and how these are affected by various modelling strategies are observed. 3) Based on this analysis acoustic distances between phones are estimated across language borders, paving the way for further research in cross-language data sharing. Finally the knowledge obtained from these experiments is applied to perform a data sufficiency analysis of a new speech recognition corpus of South African languages: The Lwazi ASR corpus. The findings correlate well with initial phone recognition results and yield insight into the sufficient number of speakers required for the development of minimal telephone ASR corpora. / Thesis (M. Ing. (Computer and Electronical Engineering))--North-West University, Potchefstroom Campus, 2009.

In the world of human language technology, resource–scarce languages (RSLs) suffer from the problem of little available electronic data and linguistic expertise. The Lwazi project in South Africa is a large–scale endeavour to collect and apply such resources for all eleven of the official South African languages. One of the deliverables of the project is more natural text–to–speech (TTS) voices. Naturalness is primarily determined by prosody and it is shown that many aspects of prosodic modelling is, in turn, dependent on part–of–speech (POS) information. Solving the POS problem is, therefore, a prudent first step towards meeting the goal of natural TTS voices. In a resource–scarce environment, obtaining and applying the POS information are not trivial. Firstly, an automatic tagger is required to tag the text to be synthesised with POS categories, but state–of–the–art POS taggers are data–driven and thus require large amounts of labelled training data. Secondly, the subsequent processes in TTS that are used to apply the POS information towards prosodic modelling are resource–intensive themselves: some require non–trivial linguistic knowledge; others require labelled data as well. The first problem asks the question of which available POS tagging algorithm will be the most accurate on little training data. This research sets out to answer the question by reviewing the most popular supervised data–driven algorithms. Since literature to date consists mostly of isolated papers discussing one algorithm, the aim of the review is to consolidate the research into a single point of reference. A subsequent experimental investigation compares the tagging algorithms on small training data sets of English and Afrikaans, and it is shown that the hidden Markov model (HMM) tagger outperforms the rest when using both a comprehensive and a reduced POS tagset. Regarding the second problem, the question arises whether it is perhaps possible to circumvent the traditional approaches to prosodic modelling by learning the latter directly from the speech data using POS information. In other words, does the addition of POS features to the HTS context labels improve the naturalness of a TTS voice? Towards answering this question, HTS voices are trained from English and Afrikaans prosodically rich speech. The voices are compared with and without POS features incorporated into the HTS context labels, analytically and perceptually. For the analytical experiments, measures of prosody to quantify the comparisons are explored. It is then also noted whether the results of the perceptual experiments correlate with their analytical counterparts. It is found that, when a minimal feature set is used for the HTS context labels, the addition of POS tags does improve the naturalness of the voice. However, the same effect can be accomplished by including segmental counting and positional information instead of the POS tags. / Thesis (M.Sc. Engineering Sciences (Electrical and Electronic Engineering))--North-West University, Potchefstroom Campus, 2011.

Our generation has experienced one of the most dramatic changes in how society communicates. Today, we have online information on almost any imaginable topic. However, most of this information is available in only a few dozen languages. In this thesis, I explore the use of parallel texts to enable cross-language information retrieval (CLIR) for languages with scarce resources. To build the parallel text I use the Bible. I evaluate different variables and their impact on the resulting CLIR system, specifically: (1) the CLIR results when using different amounts of parallel text; (2) the role of paraphrasing on the quality of the CLIR output; (3) the impact on accuracy when translating the query versus translating the collection of documents; and finally (4) how the results are affected by the use of different dialects. The results show that all these variables have a direct impact on the quality of the CLIR system.

This study aims to point out which factors should be contemplated in the planning of water resources management and then analyse the 2000/2003 basin plan for the Paulista section of the Paraíba do Sul river basin. The first step to register and organize these factors was to research the existing literature regarding to this subject. The premise to guide this study was: water is a scarce resource and a limiting for sustainable development, and it is imperative to utilize this resource in a coherent and planned way. The proposal focused also on the analysis of publications concerning the geographic aspects of the Paraíba do Sul river basin as well as the quality aspects, such as demands, uses and water offers. The studied factors are associated to the geographic aspects of the current borders, as well as their geographic localizations. Finally, according to the understanding, it was pointed out and discussed which should be the factors and why they should contemplated in the planning. / O objetivo deste trabalho é verificar, na gestão de recursos hídricos, quais são os fatores a serem considerados no seu planejamento e, a partir deles, analisar o do Plano de Bacia 2000/2003 da UGRHI 02 da área paulista da Bacia Hidrográfica do Rio Paraíba do Sul. Partindo-se da pesquisa dos vários aspectos abordados na literatura como itens a serem considerados no planejamento, fez-se o levantamento e a sistematização dos fatores. Considera-se a água um recurso escasso e um limitante para o crescimento sustentável. Assim, é de grande importância o seu uso de forma disciplinada. Evidencia-se a análise de textos relacionados ao aspecto geográfico da bacia hidrográfica do Rio Paraíba do Sul, à qualidade, à demanda, aos usos e à oferta de água. Os fatores são associados ao aspecto geográfico das fronteiras a montante e a jusante. Finalmente, foram localizados e discutidos quais são e porque determinados fatores devem ser considerados na gestão e planejamento dos recursos hídricos.

Wetlands are dynamic ecosystems and important sources of natural resources that provide a large array of ecosystem services. Unfortunately, most wetlands are threatened by human and natural stressors, such as damming, irrigation, water abstraction, climate change and variability that compromise the sustainability of the whole system. The Inner Niger Delta (IND), Mali, West Africa, is one of the biggest floodplains in the world, has a vast natural resource that attracts many people to live in and around the delta. The IND is considered a hub of human activities that include agriculture, fishing, transport, and tourism and plays an important role in promoting sustainable development for food security, water management, and the environment. As for most wetlands in the world, the very existence of the IND is at stake with the ever-increasing number of dams and irrigation schemes that are built to feed the growing population in the region. Given the fragility of the system and the multiplicity of water uses in the IND, the current knowledge of the flood dynamics and its relation to ecosystem services and the productivity of economic activity is insufficient. There is no operational hydrodynamic model of the IND, and the Malian authorities rely on simplified models and empirical relations for water resources management in the area. This thesis contributes to a better water resources management of the IND by a) developing the first 2D hydrodynamic model based on a triangular adaptative mesh of the IND that performs well despite the poor quality of available topographic/bathymetric data b) developing an innovative way of accounting for the strong hysteresis phenomenon in the IND in the hydrodynamic modeling that allowed a better reproduction of the hydraulic connectivity between important lakes and the main river and c) developing the first non-stationary relationship between the water levels at a reference station and the flooded area in the IND. The first part of the thesis deals with the challenge of developing a hydrodynamic model using only two low-resolution satellite-derived Digital Elevation Models: the Shuttle Radar Topography Mission (SRTM), which has a 30m horizontal resolution, and the Multi-Error-Removed Improved-Terrain (MERIT). Given the low vertical accuracy of global DEMs, another DEM was derived using the waterline method, by combining water extent map from satellite images and local water level information. Channel depths were approximated using the hydraulic geometric relationship methods, while the friction coefficient was derived from the global land-use class classification (GLCC) data. The river network was extracted from the water extent map corresponding to the lowest water level. Six different hydrodynamic models were created by varying the DEM and downstream boundary conditions. Each of the models was calibrated for discharge and water levels. Bayesian Model Averaging (BMA) was finally used to combine the outputs of all six hydrodynamic models into one robust simulation. In the second part, the effect of hysteresis at the downstream boundary condition of the hydrodynamic model was examined. Existing hydrodynamic models of the IND use a static stage-discharge relationship as a downstream boundary condition during both the rise and recession of the flood, leading to potential inaccuracies in the simulation of the flood extent. This paper explores the improvement in the simulation of the flood and connectivity dynamics resulting from the use of a looped rating curve at the downstream boundary of a hydrodynamic model of the IND. The hysteresis effect is integrated into the rating curve using two methods, one based on dimensionless discharges and levels (DLRC) and the other based on the modified Jones formula (MJRC). Results show that the hysteresis effect is better represented using DLRC and that simulations using any of the modified rating curves improves the accuracy of floodplain extent simulations in the areas close to the downstream station, as well as the timing of the connectivity of the river system to one important lake in the IND. The improvement in water level simulation decreases steadily with distance from the downstream boundary of the modeled area. The third part of the thesis deals with the development of an improved relation between inundation extent and water levels in the IND. Accurate knowledge of the flooded extent considered crucial for the proper management of natural resources in the IND. Several authors have developed empirical relationships between water levels at key stations in the IND and the flooded extent in an attempt to provide simple tools to link hydraulic parameters to the performance socio-economic activities in the IND. However, simulations from a hydrodynamic model of the IND showed that the relationship between water levels and the inundation extents varies greatly from year to year, and cannot be adequately captured by static formulas. First, it is demonstrated in this paper that if the maximum water level area is known in advance, accurate relationships between water levels and inundation extents can be derived. In the second part of the paper, stepwise regression is used to develop a function that can forecast maximum water levels at Akka using observed streamflow and precipitation upstream of the Delta. The combination of the two results allows a realtime estimation of the inundated area in the IND using observed water levels, precipitation, and streamflow.

Kandidatuppsats i Företagsekonomi C, Ekonomihögskolan vid Linnéuniversitetet i Växjö, Ekonomistyrning, 2FE13E, VT 2014. Författare: Robin Arnefjord, Erik Davidsson & Alexander Persson Handledare/Examinator: Petter Boye Medbedömare: Jan Alpenberg Titel: Capellagården –Hur samarbeten och nätverk kan bidra till ökade intäkter   Bakgrund: Capellagården är en fristående internatskola på södra Öland som likt många andra skolor av detta slag brottas med knappa resurser. Detta har lett till att Capellagården eftersökt alternativa sätt att effektivisera användningen av tillgängliga resurser samt genom nya samarbeten och nätverk hitta nya intäkter. Syfte: Syftet är att beskriva hur samarbeten och nätverk kan öka intäkternaför en skola som Capellagården. Detta görs i tre steg där vi inleder med att beskriva dagens situation för att sedan studera liknande skolor och slutligen identifiera utvecklingsmöjligheter. Metod: I studien så har vi använt oss av ett abduktivt angreppssätt samt en kvalitativ forskningsstrategi. Empirin har huvudsakligen hämtats från intervjuer med företrädare för tre hantverksskolor. Slutsats: Vi kan konstatera att Capellagården kan dra lärdom av andra skolors sätt att arbeta med samarbeten och nätverk för att öka intäkterna.Exempelvis genom att användning av benchmarking. Samarbetena leder också till synergieffekter i form av kompetens- samt resursdelning. Ett ökat kommunalt samarbete samt en utveckling av arbetet med studenternas nätverk hade stärkt skolan. / Bachelor Thesis Business Administration III, School of Economics at Linnaeus University of Växjö. Management Accounting, 2FE13E, Spring 2014. Authors: Robin Arnefjord, Erik Davidsson & Alexander Persson Examinator/Advisor: Petter Boye Second Examiner: Jan Alpenberg Title: Capellagården – How co-operation and networks can contribute to increased revenues   Background: Capellagården is a stand-alone boarding school located in the southern part of Öland and like many other schools they struggle with scarce resources. This has led to that Capellagården have sought alternative ways to make the usage of available resources more effective as well as through new co-operations and networks find new revenue. Purpose: The purpose is to describe how co-operation and networks can help with the development of a school like Capellagården and increase revenues. This is done in three steps where we start with a description of the present-day situation then study similar schools and finally identify potential development opportunities. Method: We have used an abductive approach and a qualitative research strategy in this study. The empirical material has primarily been collected through interviews with relevant parties of three Craft schools. Conclusion: We can conclude that Capellagården can learn from the way other schools work with co-operation and networks to increase revenues. For example through the usage of benchmarking. Co-operations will also lead to synergy effects in the shape of sharing competence and resources. An increase in the municipal co-operation and a development of the students network would have strengthened the school.

<p>Flooding is one of the most devastating and frequently occurring natural phenomena in the world. Due to the adverse impacts of floods on the life and property of humans, it is crucial to investigate the best flood modeling approaches for delineation of floodplain areas. Conventionally, different hydrodynamic models are used to identify the floodplain areas. However, the high computational cost, and the dependency of these models on detailed input datasets limit their application for large scale floodplain mapping in data-scarce regions. Recently, a new floodplain mapping method based on a hydrogeomorphic feature, named Height Above Nearest Drainage (<i>HAND</i>), has been proposed as a successful alternative for fast and efficient floodplain mapping at the large scale. The overall goal of this study is to improve the performance of <i>HAND</i>-based method by overcoming its current limitations. The main focus will be on extending the application of the <i>HAND</i>-based method to data-scarce environments. To achieve this goal, regionalization techniques are integrated with the floodplain models at the regional and continental scales. Considering these facts, four research objective are established to (1) Develop a regression model to create 100-year floodplain maps at a regional scale (2) Develop a classification framework for creating 100-year floodplain maps for the Contiguous United States (3) Develop a new version of the <i>HAND</i>-based method for creating probabilistic 100-year floodplain maps, and (4) Propose a general regionalization framework for transferring information from data-rich basins to data-scarce environments. </p> <p> </p> <p>In the first objective, the state of North Carolina is selected as the study area, and a regression model is developed to regionalize the available 100-year Flood Insurance Rate Maps (FIRMs) to the data-scarce regions. The regression model is an exponential equation with three independent variables including the average slope, the average elevation, and the main stream slope of the watershed. The results show that the estimated floodplains are within the expected range of accuracy of C>0.6 and F>0.9 for majority of watersheds located in the mid-altitude regions, but it overpredicts and underpredicts in the flat and mountainous regions respectively. </p> <p> </p> <p>The second objective of this research extends the spatial application of the <i>HAND</i>-based method to the entire United States by proposing a new classification framework. The proposed framework classifies the watersheds into three groups by using seven watershed characteristics related to the topography, climate and land use. The validation results show that the average error of floodplain maps is around 14% which demonstrate the reliability and robustness of the proposed framework for continental floodplain mapping. In addition to the acceptable accuracy, the proposed framework creates the floodplain maps for any watershed within the United States. </p> <p> </p> <p>The <i>HAND</i>-based method is a deterministic modeling approach to floodplain mapping. In the third objective, the probabilistic version of this method is proposed. Using a probabilistic approach to floodplain mapping provides more informative maps. In this study, a flat watershed in the state of Kansas is selected as the case study, and the performance of four probabilistic functions for floodplain mapping is compared. The results show that a linear function with one parameter and a gamma function with two parameters are the best options for this study area. It is also shown that the proposed probabilistic approach can reduce the overpredictions and underpredictions made by the deterministic <i>HAND</i>-based approach. </p> <p> </p> <p>In the fourth objective, a new regionalization framework for transferring the calibrated environmental models to data-scarce regions is proposed. This framework aims to improve the current similarity-based regionalization methods by reducing the subjectivity that exists in the selection of basin descriptors. Using this framework for the probabilistic <i>HAND</i>-based method in the third objective, the floodplains are regionalized for a large set of watersheds in the Central United States. The results show that “vertical component of centroid (or latitude)” is the dominant descriptor of spatial variabilities in the probabilistic floodplain maps. This is an interesting finding which shows how a systematic approach can help to explore the hidden descriptors for regionalization. It is demonstrated that using common methods, such as correlation coefficient calculation, or stepwise regression analysis, will not reveal the critical role of latitude on the spatial variability of floodplains.</p>