Submitted by Biblioteca de Pós-Graduação em Geoquímica BGQ (bgq@ndc.uff.br) on 2017-07-03T17:28:54Z No. of bitstreams: 1 MARTINS, G. S..pdf: 4719726 bytes, checksum: 2256ee6ea720c11998227a048aa70881 (MD5) / Made available in DSpace on 2017-07-03T17:28:54Z (GMT). No. of bitstreams: 1 MARTINS, G. S..pdf: 4719726 bytes, checksum: 2256ee6ea720c11998227a048aa70881 (MD5) / Universidade Federal Fluminense. Instituto de Química. Programa de Pós-Graduação em Geoquímica, Niterói, RJ / O estudo em questão apresenta o uso do Carbono graf ítico (BC) como indicador de queimadas em dois enfoques: ocorrência de paleoincêndios em um r egistro sedimentar do Lago do Saci (PA), com o intuito de evidenciar eventos de distúrbios relacio nados a queima de sistemas florestais associados a ocorrência de paleoclimas secos na parte sudeste da Amazônia ao longo do Holoceno e como indicador de queimadas em registros sedimentares de mudança do u so da terra, no município de Alta Floresta (MT). O perfil da concentração de carbono grafítico (BC) de terminado através do método térmico (CTO-375) refere-se ao testemunho SACI 1, que abrange 9200 an os cal AP e a quatro registros sedimentares de barragens artificiais do norte do Mato Grosso .refe rente aos últimos 100 anos. No registro sedimentar do Lago do Saci foi possível observar quatro fases dis tintas de sedimentação que permite distinguir difer entes fases paleoclimáticas. No Holoceno inferior, observ a-se uma transição de um ambiente úmido para um mais seco, caracterizado por fortes eventos de enxu rradas. No Holoceno Médio estabeleceu-se um ambient e deposicional de baixa energia hidrodinâmica, com al tos valores de carbono grafítico suportando a existência de um clima seco. A transição para o Hol oceno tardio foi marcado por um clima úmido como sugerem o aumento de carbono orgânico total e deriv ados de clorofila assim como reduções nos valores d e concentração de carbono grafítico. Durante o Holoc eno Tardio, o aumento da concentração de BC é observado por volta de 1850 anos cal AP, decorrente s de atividades antrópicas na região ou alteração d e biomassa. Nos registros de mudança do uso da terra, os dados de carbono grafítico auxiliam na determinação da dimensão dos incêndios em escala lo cal ou regional como dado complementar a contagem microscópica de partículas de carvão, bem como proc essos de co-deposição entre o mercúrio e BC, como mostra a correlação significativa e positiva entre os dois parâmetros (p < 0,05; N=59 e r = 0,55). A a nalise das partículas de carvão e Hg sugerem a remobilizaç ão o metal por ação de queimadas, pois os testemunh os LCEN050 e WSW150 apresentaram as maiores concentraç ões de partículas de carvão 10,8x10 6 e 6,96x10 6 partículas/g respectivamente e os menores valores m édios na concentração de Hg (0,06 e 0,07ppm respectivamente). Considerando que em setores mais afastados ao centro municipal e de maior conservaçã o de biomassa, observam-se os maiores valores de acum ulação de carbono grafítico, 0,023% (LCEN050) e 0,078% (WNW150) de BC, sugere-se que o transporte d estas partículas pela circulação atmosférica de longo percurso. Em escala temporal, o processo de o cupação territorial é observado no testemunho WNW150 (única barragem natural entre as quatro estu dadas). O processo de colonização do município, ocorrido na década de 70, é marcado pelo aumento do s fluxos de todos os parâmetros a partir de 1967. O fluxo de partículas de carvão aumenta de 0.028 para 0,500x10 10 partículas/m2/ano e os de Carbono grafítico de 0,405 para 2,28 gBC/m 2 /ano entre 1967 e 1984. / his study presents the use of Black Carbon as indi cator of fire occurrences on two approaches: Occurrences of paleofires in a sediment ary record of Saci lake (PA), to related events of burning forest, associated with the occur rence of dry paleoclimates, in the southeastern part of the Amazon during the Holocene ; and as an indicator of fire records of land change use in Alta Floresta (MT).The concentra tion of Black Carbon (BC)for the core SACI 1, with a basal age of 9200 cal years BP, and four core in the northern Mato Grosso that covers the last 100 years, was determined by the th ermal method (CTO-375).In the sedimentary record of Saci lake was identified four units of sedimentation which correspond to different paleoclimatic phases . In the Early Holocene the transition from humid to a drier climate was observed, characterized by the occurren ce of run-off events. In the Mid-Holocene was established a depositional environment of low h ydrodynamic energy, with high values of BC supporting the existence of a dry climate. The M id-late Holocene transition was marked by a humid climate as suggest the increase in TOC v alues and chlorophyll derivates as well as the reduction in the BC concentration . In the Late Holocene occurs the increase of BC concentrations in 1850 cal yrs BP, due the antropog enic activities in the region or biomass change around the lake. In the land change use reco rds, the BC profile work with a complementary proxy associated with the charcoal pa rticles, could provide the extent of fires in regional or local scale, well as the co-deposition process of mercury and BC , as shown insignificant and positive correlation between the two parameters(p <0.05, N = 59; r = 0.55). The analysis of the coal particles and Hg suggest the r emobilization of the metal by the burning process used for croplands installation, for the cores LCEN 050 and WSW150 had the highest concentrations of carbon particles 10.8x10 6 and 6.96x10 6 particles/g respectively and the lowest average concentration of Hg (0.06and 0.07 ppm respectively) . In the more distant collect point of the town center with larger biomass conservation, there are the highest accumulation of graphitic carbon, 0.023% (LCEN050) and 0.078% (WNW150) BC, it is sugg ested suggest the transport of these particles by the atmospheric circulation. In the te mporal scale, the process of land occupation is observed in the core WNW150. The process of coloniz ation of the city, occurred in the 70's, is marked by increased flows of all parameters since 1967. Th e flow of carbon particles increases from 0.028x10 10 to 0.500x10 10 partícles/m 2 /yr and graphitic carbon of 0.405 to 2.28 gBC/m 2 /yr between 1967 and 1984

Mittels umweltpolitischer Programme von weltweit einzigartigem Ausmaß versucht China sowohl den menschlichen Druck auf die Landbedeckung zu mindern als auch eine nachhaltige Landnutzung zu fördern. Als Hotspot globaler Umweltveränderungen wurden insbesondere für die Innere Mongolei erhebliche Investitionen zur Renaturierung von Landoberflächen von der chinesischen Zentralregierung getätigt. Die Fernerkundung ist ein effektiver Ansatz für ein flächendeckendes Monitoring des Landnutzungswandels. Jedoch erschweren die limitierte Datenverfügbarkeit und das Fehlen verfügbarer Veränderungsanalysemethoden die Anwendung fernerkundlicher Techniken zum Monitoring von Landnutzung und ihrer Veränderungen. Um den Landnutzungswandel in der Inneren Mongolei verlässlich zu kartieren, wurden daher in einem ersten Schritt Möglichkeiten und Grenzen von Zeitreihen räumlich grob aufgelöster Fernerkundungsdaten für das Monitoring von Langzeitveränderungen der Landbedeckung untersucht. Im zweiten Schritt wurde ein Ansatz zur Erfassung von jährlichen Veränderungen zwischen mehreren Landnutzungsklassen entwickelt und angewandt. Die Ergebnisse zeigen, dass die chinesische Landnutzungspolitik seit dem Jahr 2000 wirksam zum Erhalt und zur Regenerierung von Waldökosystemen in der Inneren Mongolei beiträgt. Abnehmende Entwaldung und ein Zuwachs von Waldflächen sind insbesondere in jenen Regionen zu finden, in welchen die landnutzungspolitischen Maßnahmen umgesetzt wurden. Die Konvertierung von Ackerland zu Grasland wurde zumeist innerhalb anfälliger, klimatisch und topographisch ungeeigneter Gebiete beobachtet. Die vorliegende Dissertation veranschaulicht sowohl den Einfluss politischer Maßnahmen und zugrunde liegender sozio-ökonomischer Treiber auf die Landoberfläche als auch die Bedeutung von grob aufgelösten Fernerkundungsdaten und Zeitreihenanalysen für das Monitoring des Landnutzungswandels in großräumigen Gebieten. / Monitoring land use and land cover change (LULCC) support better interpretation about how land surfaces are impacted by human decisions. The overall aim of this thesis is to gain a better understanding about LULCC in Inner Mongolia using remote sensing under consideration of China’s land use policies. With the largest scale land restoration programs in the world, China aims to reduce human pressure on lands and promote sustainable land use. As a hot-spot of environmental change, Inner Mongolia received the heaviest investment from the central government for land restoration. Yet the effectiveness and consequences of China’s land use policies in Inner Mongolia remain unclear. Remote sensing is an effective tool for monitoring land use and land cover change across broad scales, yet data limitations and a lack of available change detection methods hampers the capacity of researchers to apply remote sensing techniques for LULCC monitoring. To reliably map LULCC in Inner Mongolia, the opportunities and limitations of using coarse resolution imagery time series for monitoring long-term land changes was first examined. Second, an approach detecting annual changes between multiple land categories was developed and applied in Inner Mongolia. Results indicate that China’s land use policies effectively preserved and recovered forest ecosystems in Inner Mongolia after the year 2000. The decreasing trends of deforestation and forest gain are obvious in the regions that implement China’s land use policies, which reflect the positive influence of the policy. Cropland retirement was mostly found in ecologically fragile areas where climate and topographic conditions are unsuitable for cultivation. This thesis reveals how political factors and other underlying social-economic drivers impact a country''s land surface, and highlights the values of using coarse resolution imagery and time series analysis for LULCC monitoring across large areas.

Fererkundliches monitoring von Landnutzungswandel ist eine Grundvoraussetzung um negative Auswirkungen globaler Umweltveränderungen zu reduzieren. Verfügbare Methoden unterliegen jedoch räumlichen und zeitlichen Einschränkungen. Die Öffnung des Landsat Archivs, bessere Datenqualität, fortgeschrittene Algorithmen und Rechenkapazitäten erlauben verbessertes Verständnis von Muster-Prozess Zusammenhängen, falls diese Einschränkungen überwunden werden. Das Ziel dieser Dissertation war es, Methoden zu entwickeln und anzuwenden, die eine bessere Nutzung des Landsat Archivs ermöglichen, um Landnutzungswandel in den Karpaten seit 1985 zu quantifizieren. Das sekundäre Ziel war es zu untersuchen, wie der Zusammenbruch des Sozialismus und der Beitritt zur Europäischen Union den regional Landnutzungswandel beeinträchtigt hat. Dafür wurde zunächst ein zeitreihenbasiertes Verfahren genutzt, um mittels verbesserter zeitlicher Auflösung ein besseres Verständnis von Waldstörungsdynamiken während der Umstrukturierung von Forstbesitzverhältnissen zu erlangen. Anschliessend wurden Compositing Algorithmen entwickelt, und mittels dieser die regionalen forst- und landwirtschaftlichen Veränderungen quantifiziert. Die Ergebnisse zeigten, dass das Ende des Sozialismus zu einer drastischen Reduzierung der Waldstörungen geführt hat und gleichzeitig weitflächig Ackerland aufgegeben wurde. Insgesamt nahm die Waldfläche leicht zu aber übermäßige Störungen erfolgten in verschiedenen Gebieten, verursacht durch den kombinierten Effekt forstwirtschaftlicher Vermächtnisse, natürlicher Störungen sowie Waldbewirtschaftung. Nach dem EU-Beitritt nahmen Störungsdynamiken in den Karpaten im Vergleich zu den ersten Übergangsjahren wieder zu, was auf neuorganiserte Forstwirtschaft mit Zugang zu EU Märkten hindeuten kann. Aufgegebene Ackerflächen wurden in den letzten Jahren vermehrt wiederbewirtschaftet, wahrscheinlich bedingt durch die Agrarpolitik der EU sowie einen zunehmenden Einfluss globaler Märkte. / Remote sensing based monitoring of land change is a prerequisite to reduce the negative impacts of global environmental change. However, available monitoring methods suffer from spatial and/or temporal limitations. The opening of the Landsat archive, advancements in data quality, processing algorithms and capabilities, can improve pattern-process understanding if these limitations are overcome. The overall aim of this dissertation was first, to develop and apply methods that better utilize the rich Landsat record and to map and quantify land change in the Carpathian ecoregion since 1985. The secondary objective is to investigate how the collapse of socialism and accession to the European Union (EU) affected regional land change. First, a trajectory based change detection approach was used to investigate how increased observation frequency helps understanding how forest ownership changes affected forest disturbance dynamics. Second, compositing algorithms are developed to facilitate mapping and change detection over large areas. This allowed assessing changes in forest cover and agriculture. Results showed that overall the collapse of socialism led to drastic declines in forest disturbances and simultaneously to widespread cropland abandonment. Forest cover overall expanded but excessive harvesting prevailed in certain areas, due to combined effects of land use legacies, natural disturbances and forest management. Following the EU accession, disturbance levels increased compared to the transition years, potentially relating to a re-established forestry sector with access to EU timber markets. Abandoned cropland was recultivated throughout the Carpathians during the most recent years, likely influenced by the EU Common Agricultural Policy and increasingly by global markets. This dissertation exemplifies the value of the Landsat archive for land change research which can improve our understanding of land change globally and thus help mitigate its negative impacts.

Land use changes have become a major contributor to the anthropogenic global change. The ongoing dispersion and concentration of the human species, being at their orders unprecedented, have indisputably altered Earth’s surface and atmosphere. The effects are so salient and irreversible that a new geological epoch, following the interglacial Holocene, has been announced: the Anthropocene. While its onset is by some scholars dated back to the Neolithic revolution, it is commonly referred to the late 18th century. The rapid development since the industrial revolution and its implications gave rise to an increasing awareness of the extensive anthropogenic land change and led to an urgent need for sustainable strategies for land use and land management. By preserving of landscape and settlement patterns at discrete points in time, archival geospatial data sources such as remote sensing imagery and historical geotopographic maps, in particular, could give evidence of the dynamic land use change during this crucial period. In this context, this thesis set out to explore the potentials of retrospective geoinformation for monitoring, communicating, modeling and eventually understanding the complex and gradually evolving processes of land cover and land use change. Currently, large amounts of geospatial data sources such as archival maps are being worldwide made online accessible by libraries and national mapping agencies. Despite their abundance and relevance, the usage of historical land use and land cover information in research is still often hindered by the laborious visual interpretation, limiting the temporal and spatial coverage of studies. Thus, the core of the thesis is dedicated to the computational acquisition of geoinformation from archival map sources by means of digital image analysis. Based on a comprehensive review of literature as well as the data and proposed algorithms, two major challenges for long-term retrospective information acquisition and change detection were identified: first, the diversity of geographical entity representations over space and time, and second, the uncertainty inherent to both the data source itself and its utilization for land change detection. To address the former challenge, image segmentation is considered a global non-linear optimization problem. The segmentation methods and parameters are adjusted using a metaheuristic, evolutionary approach. For preserving adaptability in high level image analysis, a hybrid model- and data-driven strategy, combining a knowledge-based and a neural net classifier, is recommended. To address the second challenge, a probabilistic object- and field-based change detection approach for modeling the positional, thematic, and temporal uncertainty adherent to both data and processing, is developed. Experimental results indicate the suitability of the methodology in support of land change monitoring. In conclusion, potentials of application and directions for further research are given.

Land use changes have become a major contributor to the anthropogenic global change. The ongoing dispersion and concentration of the human species, being at their orders unprecedented, have indisputably altered Earth’s surface and atmosphere. The effects are so salient and irreversible that a new geological epoch, following the interglacial Holocene, has been announced: the Anthropocene. While its onset is by some scholars dated back to the Neolithic revolution, it is commonly referred to the late 18th century. The rapid development since the industrial revolution and its implications gave rise to an increasing awareness of the extensive anthropogenic land change and led to an urgent need for sustainable strategies for land use and land management. By preserving of landscape and settlement patterns at discrete points in time, archival geospatial data sources such as remote sensing imagery and historical geotopographic maps, in particular, could give evidence of the dynamic land use change during this crucial period. In this context, this thesis set out to explore the potentials of retrospective geoinformation for monitoring, communicating, modeling and eventually understanding the complex and gradually evolving processes of land cover and land use change. Currently, large amounts of geospatial data sources such as archival maps are being worldwide made online accessible by libraries and national mapping agencies. Despite their abundance and relevance, the usage of historical land use and land cover information in research is still often hindered by the laborious visual interpretation, limiting the temporal and spatial coverage of studies. Thus, the core of the thesis is dedicated to the computational acquisition of geoinformation from archival map sources by means of digital image analysis. Based on a comprehensive review of literature as well as the data and proposed algorithms, two major challenges for long-term retrospective information acquisition and change detection were identified: first, the diversity of geographical entity representations over space and time, and second, the uncertainty inherent to both the data source itself and its utilization for land change detection. To address the former challenge, image segmentation is considered a global non-linear optimization problem. The segmentation methods and parameters are adjusted using a metaheuristic, evolutionary approach. For preserving adaptability in high level image analysis, a hybrid model- and data-driven strategy, combining a knowledge-based and a neural net classifier, is recommended. To address the second challenge, a probabilistic object- and field-based change detection approach for modeling the positional, thematic, and temporal uncertainty adherent to both data and processing, is developed. Experimental results indicate the suitability of the methodology in support of land change monitoring. In conclusion, potentials of application and directions for further research are given.