Refine
Has Fulltext
- yes (23)
Is part of the Bibliography
- yes (23)
Year of publication
Document Type
- Doctoral Thesis (21)
- Journal article (1)
- Master Thesis (1)
Keywords
- Fernerkundung (23) (remove)
Institute
- Institut für Geographie und Geologie (23) (remove)
Sonstige beteiligte Institutionen
- Deutsches Zentrum für Luft & Raumfahrt (DLR) (1)
- Deutsches Zentrum für Luft- und Raumfahrt (DLR) (1)
- Deutsches Zentrum für Luft- und Raumfahrt (DLR), Deutsches Fernerkundungsdatenzengrum (DFD) (1)
- Deutsches Zentrum für Luft- und Raumfahrt e.V. (1)
- Lehrstuhl für Fernerkundung der Uni Würzburg, in Kooperation mit dem Deutschen Fernerkundungsdatenzentrum (DFD) des Deutschen Zentrums für Luft- und Raumfahrt (DLR) (1)
- South African National Biodiversity Institute (SANBI) (1)
Processes of the Earth’s surface occur at different scales of time and intensity. Climate in particular determines the activity and seasonal development of vegetation. These dynamics are predominantly driven by temperature in the humid mid-latitudes and by the availability of water in semi-arid regions. Human activities are a modifying parameter for many ecosystems and can become the prime force in well-developed regions with an intensively managed environment. Accounting for these dynamics, i.e. seasonal dynamics of ecosystems and short- to long-term changes in land-cover composition, requires multiple measurements in time. With respect to the characterization of the Earth surface and its transformation due to global warming and human-induced global change, there is a need for appropriate data and methods to determine the activity of vegetation and the change of land cover. Space-borne remote sensing is capable of monitoring the activity and development of vegetation as well as changes of the land surface. In many instances, satellite images are the only means to comprehensively assess the surface characteristics of large areas. A high temporal frequency of image acquisition, forming a time series of satellite data, can be employed for mapping the development of vegetation in space and time. Time series allow for detecting and assessing changes and multi-year transformation processes of high and low intensity, or even abrupt events such as fire and flooding. The operational processing of satellite data and automated information-extraction techniques are the basis for consistent and continuous long-term product generation. This provides the potential for directly using remote-sensing data and products for analyzing the land surface in relation to global warming and global change, including deforestation and land transformation. This study aims at the development of an advanced approach to time-series generation using data-quality indicators. A second goal focuses on the application of time series for automated land-cover classification and update, using fractional cover estimates to accommodate for the comparatively coarse spatial resolution. Requirements of this study are the robustness and high accuracy of the approaches as well as the full transferability to other regions and datasets. In this respect, the developments of this study form a methodological framework, which can be filled with appropriate modules for a specific sensor and application. In order to attain the first goal, time-series compilation, a stand-alone software application called TiSeG (Time Series Generator) has been developed. TiSeG evaluates the pixel-level quality indicators provided with each MODIS land product. It computes two important data-availability indicators, the number of invalid pixels and the maximum gap length. Both indices are visualized in time and space, indicating the feasibility of temporal interpolation. The level of desired data quality can be modified spatially and temporally to account for distinct environments in a larger study area and for seasonal differences. Pixels regarded as invalid are either masked or interpolated with spatial or temporal techniques.
Vulnerabilitätsabschätzung der erdbebengefährdeten Megacity Istanbul mit Methoden der Fernerkundung
(2008)
Urbane Räume zählen zu den dynamischsten Regionen dieser Erde. Besonders Megacities zeigen bereits heute Trends und Dimensionen der Urbanisierung, deren regionale und globale Folgen noch kaum vorhersehbar, und erst ansatzweise erforscht sind. Die enorme räumliche Konzentration von Menschen, Werten und Infrastruktur auf engem Raum ist für diese urbanen Räume die Grundlage einer hohen Verwundbarkeit (Vulnerabilität). Gerade im Kontext von Naturgefahren potenzieren sich die Risiken, die durch den schnellen strukturellen, sozioökonomischen und ökologischen Wandel entstehen. Das übergeordnete Ziel dieser Dissertation ist daher die Analyse von Potentialen der Fernerkundung zur Abschätzung von Risiko und Vulnerabilität am Beispiel der erdbebengefährdeten Megacity Istanbul. Um die Zielstellung systematisch zu verfolgen, wird ein konzeptioneller, thematischer Leitfaden entwickelt. Dieser besteht aus einer Systematisierung der abstrakten Überbegriffe ‚Risiko’, ‚Vulnerabilität’ und ‚Gefährdung’ in einem Indikatorensystem. Konkrete, messbare Indikatoren für das System ‚urbaner Raum’ erlauben eine quantitative Abschätzung von Einzelaspekten, addieren sich aber auch zu einer ganzheitlichen Perspektive des Risikos. Basierend auf dieser holistischen Idee, erlaubt das Indikatorensystem Potentiale, aber auch Limitierungen der Fernerkundungsdaten und Bildverarbeitungsmethoden für die Abschätzung von Risiko und Vulnerabilität zu identifizieren. Anhand des Leitfadens werden zielgerichtet Methoden zur automatisierten Extraktion räumlicher Informationen aus Fernerkundungsdaten entwickelt. Ein objektorientierter, modularer Klassifikationsansatz ermöglicht eine Landbedeckungsklassifikation höchst aufgelöster Daten im urbanen Raum. Dieses modulare Rahmenwerk zielt auf eine einfache und schnelle Übertragbarkeit auf andere höchst auflösende Sensoren bzw. andere urbane Strukturen. Zur Anpassung der Methoden werden neben IKONOS Daten der Megacity Istanbul und der erdbeben- und tsunamigefährdeten Küstenstadt Padang in Indonesien, Quickbird Daten für die zukünftige Megacity Hyderabad in Indien getestet. Die Resultate zeigen die detaillierte und hochgenaue Erfassung kleinräumiger, heterogener urbaner Objekte mit Genauigkeiten von über 80 %. Auch mittel aufgelöste Landsat Daten werden mit einem objektorientierten modularen Rahmenwerk mit hohen Genauigkeiten klassifiziert, um komplementäre temporale und gesamtstädtische Analysen hinzuzufügen. Damit wird eine aktuelle, flächendeckende und multiskalige Informationsbasis generiert, die als Ausgangsprodukt zur Analyse urbaner Vulnerabilität dient. Basierend auf diesen Informationsebenen werden dem konzeptionellen Leitfaden folgend Indikatoren zur Abschätzung von Vulnerabilität und Risiko extrahiert. Der Fokus ist dabei die Entwicklung von Methoden zur automatisierten, interpreterunabhängigen Ableitung vulnerabilitäts- und gefährdungsrelevanter Indikatoren. Die physische Analyse des kleinräumigen urbanen Raums konzentriert sich dabei auf die Typisierung des Gebäudebestandes mit Parametern wie Dichte, Höhe, Alter, Größe, Form sowie Dachtyp. Indirekt wird zudem mittels dieser Parameter die Bevölkerungsdichteverteilung abgeleitet. Weitere Standortfaktoren ergeben sich aus Lageparametern wie Distanzen zu Hauptverkehrsachsen, Freiflächenanalysen oder der Geländeoberfläche. Schließlich führt die Vulnerabilitätsabschätzung den modellhaften, thematischen Leitfaden mit den abgeleiteten Indikatoren zusammen. Dazu erfolgt eine Normierung der unterschiedlichen abgeleiteten Indikatoren auf einen einheitlichen Vulnerabilitätsindex. Dieser zielt auf eine räumliche und zeitliche Vergleichbarkeit und die Möglichkeit, die vielfältigen Informationsebenen zu kombinieren. Damit wird das Zusammenspiel verschiedenster Indikatoren simuliert und erlaubt daraus Identifizierung und Lokalisierung von Brennpunkten im Desasterfall. Über das fernerkundliche Potential hinaus, werden die Resultate in einer interdisziplinären Methode zu einem synergetischen Mehrwert erhoben. Statt einer quantitativen Abschätzung der physischen Gebäudeparameter, ermöglicht eine Methode des Bauingenieurwesens in Kombination mit der fernerkundlichen Gebäudetypisierung eine Abschätzung der wahrscheinlichen Schadensanfälligkeit von Gebäuden im Falle eines Erdbebens. Exemplarisch wird das Potential der Resultate für Entscheidungsträger anhand eines Erdbebensszenarios aufgezeigt. Risiko und Vulnerabilität lassen sich dadurch räumlich sowohl nach betroffenen Häusern und betroffenen Menschen als auch nach räumlichen Standortfaktoren wie beispielsweise Zugänglichkeit quantifizieren. Dies ermöglicht gezielt präventiv zu agieren oder während und nach einem Desaster gezieltes Krisenmanagement zu betreiben. Im Hinblick auf die zentrale Fragestellung dieser Dissertation lässt sich resümieren, dass die Aktualität sowie die geometrische und thematische Qualität der Resultate aus Fernerkundungsdaten, den Anforderungen des komplexen, kleinräumigen und dynamischen urbanen Raums gerecht werden. Die Resultate führen zu der Erkenntnis, dass das Potential der Fernerkundung zur Abschätzung von Vulnerabilität und Risiko vor allem in der direkten Ableitung physischer Indikatoren sowie der indirekten Ableitung demographischer Parameter liegt.
In den letzten Jahrzehnten ist eine verstärkte Veränderung der Landoberfläche beobachtet worden. Diese Prozesse sind direkten und indirekten anthropogenen Einflüssen zuzuschreiben, wie Deforestation oder Klimawandel. Mit dieser Entwicklung geht der Verlust und die Fragmentation von naturnahen Flächen einher. Für das Fortbestehen von Populationen verschiedenster Organismen in einer derartig geformten Landschaft ist entscheidend, inwieweit die Migration zwischen bestehenden Fragmenten gewährleistet ist. Diese wird von der Eignung der umgebenden Landschaft beeinflusst. Im Kontext einer klimatischen Veränderung und verstärkter anthropogener Landnutzung ist die Analyse der räumlichen Anordnung von Habitatfragmenten und der Qualität der umgebenden Landschaft besonders für die globale Aufrechterhaltung der Biodiversität wichtig. Großräumige Muster der Landschaftsveränderung können mit Hilfe von Satellitendaten analysiert werden, da es nur diese ermöglichen die Landbedeckung flächendeckend, reproduzierbar und auf einer adäquaten räumlichen Auflösung zu kartieren. Besonders zeitlich hochaufgelöste Daten liefern wertvolle Informationen bezüglich der Dynamik der Landbedeckung. Diese Arbeit beschäftigt sich mit der Analyse der Fragmentation in Westafrika und der potentiellen Bedeutung von singulären Fragmenten und deren potentiellen Auswirkungen auf die Biodiversität. Dafür wurden zeitlich hoch- und räumlich mittelaufgelöste Daten des Aufnahmesystems MODIS verwendet, mit denen für das Untersuchungsgebiet Westafrika die Landbedeckung klassifziert wurde. Für die darauf folgenden Analysen der räumlichen Konfiguration der Fragmente wurde der Fokus auf Regenwaldgebiete gelegt. Die Analyse von räumlichen Mustern der Regenwaldfragmente liefert weiterführende qualitative Informationen der individuellen Teilbereiche. Die räumliche Anordnung wurde sowohl mit etablierten Maßen als auch mittels in dieser Arbeit erstellter robuster und übertragbarer Indizes quantifiziert. Es konnte gezeigt werden, dass die Verwendung von aussagekräftigen Indizes, besonders, wenn sie alle benachbarten Fragmente und die Qualität der umgebenden Matrix berücksichtigen, die räumliche Differenzierung von Fragmenten verbessert. Jedoch ist die Anwendung dieser Maße abhängig von den Ansprüchen einer Art. Daher muss die artspezifische Perzeptionen der Landschaft auf der Basis der Indizes implementiert werden, da die Übertragung der Ergebnisse einzelner Indizes auf andere räumliche Auflösungen und andere Regionen nur begrenzt möglich war. Des Weiteren wurden potentielle Einflussfaktoren auf die räumlichen Muster mittels Neutraler Landschaftsmodelle untersucht. Hierbei ergaben sich je nach Region und Index unterschiedliche Ergebnisse, allerdings konnte der Einfluss anthropogen induzierter Veränderungen auf die Landbedeckung postuliert werden. Die große Bedeutung der räumlichen Attribution von Landbedeckungsklassen konnte in dieser Arbeit aufgezeigt werden. Der alleinige Fokus auf die Kartierung von z. B. Waldfragmenten ohne deren räumliche Anordnung zu berücksichtigen, kann zu falschen Schlüssen bezüglich deren ökologischen, hydrologischen und klimatologischen Bedeutung führen.
Glacier outlines during the ‘Little Ice Age’ maximum in Jotunheimen were mapped by using remote sensing techniques (vertical aerial photos and satellite imagery), glacier outlines from the 1980s and 2003, a digital terrain model (DTM), geomorphological maps of individual glaciers, and field-GPS measurements. The related inventory data (surface area, minimum and maximum altitude) and several other variables (e.g. slope, range) were calculated automatically by using a geographical information system. The length of the glacier flowline was mapped manually based on the glacier outlines at the maximum of the ‘Little Ice Age’ and the DTM. The glacier data during the maximum of the ‘Little Ice Age’ were compared with the Norwegian glacier inventory of 2003. Based on the glacier inventories during the maximum of the ‘Little Ice Age’, the 1980s and 2003, a simple parameterization after HAEBERLI & HOELZLE (1995) was performed to estimate unmeasured glacier variables, as e.g. surface velocity or mean net mass balance. Input data were composed of surface glacier area, minimum and maximum elevation, and glacier length. The results of the parameterization were compared with the results of previous parameterizations in the European Alps and the Southern Alps of New Zealand (HAEBERLI & HOELZLE 1995; HOELZLE et al. 2007). A relationship between these results of the inventories and of the parameterization and climate and climate changes was made.
Der Klimawandel und insbesondere die globale Erwärmung gehören aktuell zu den größten Herausforderungen an Politik und Wissenschaft. Steigende CO2-Emissionen sind hierbei maßgeblich für die Klimaerwärmung verantwortlich. Ein regulierender Faktor beim CO2-Austausch mit der Atmosphäre ist die Vegetation, welche als CO2-Senke aber auch als CO2-Quelle fungieren kann. Diese Funktionen können durch Analysen der Landbedeckungsänderung in Kombination mit Modellierungen der Kohlenstoffbilanz quantifiziert werden, was insbesondere von aktuellen und zukünftigen politischen Instrumenten wie CDM (Clean Development Mechanism) oder REDD (Reducing Emissions from Deforestation and Degradation) gefordert wird. Vor allem in Regionen mit starker Landbedeckungsänderung und hoher Bevölkerungsdichte sowie bei geringem Wissen über die Produktivität und CO2-Speicherpotentiale der Vegetation, bedarf es einer Erforschung und Quantifizierung der terrestrischen Kohlenstoffspeicher. Eine Region, für die dies in besonderem Maße zutrifft, ist Westafrika. Jüngste Studien haben gezeigt, dass sich einerseits die Folgen des Klimawandels und Umweltveränderungen sehr stark in Westafrika auswirken werden und andererseits Bevölkerungswachstum eine starke Änderung der Landbedeckung für die Nutzung als agrarische Fläche bewirkt hat. Folglich sind in dieser Region die terrestrischen Kohlenstoffspeicher durch Ausdehnung der Landwirtschaft und Waldrodung besonders gefährdet. Große Flächen agieren anstelle ihrer ursprünglichen Funktion als CO2-Senke bereits als CO2-Quelle. [...]
Die Veränderung der terrestrischen Ökosysteme, ist ein grundlegendes Element des Globalen Wandels. In diesem Kontext unterliegt auch eines der größten Biome der Erde, die tropische und subtropische Savanne, immer stärkeren Veränderungen. Dieses Biom in sozioökonomischer und ökologischer Hinsicht von besonderer Bedeutung. Für einen rasch wachsenden Teil der Weltbevölkerung bildet es die Grundlage für das Betreiben von Weidewirtschaft, Ackerbau und Tourismus. In nationalen und internationalen Forschungsprogrammen zum Globalen Wandel hat die Analyse von Landnutzungs- und Landbedeckungsänderungen in den vergangenen Jahrzehnten zunehmend an Bedeutung gewonnen. Die Landbedeckungsdynamik von Savannenökosystemen ist jedoch noch nicht hinreichend verstanden, so dass diese Ökosysteme in globalen Studien nur ansatzweise berücksichtigt werden können. Besondere Herausforderungen bei der Erfassung der Landbedeckung und ihrer Dynamik liegen im Falle der Savannen in der heterogenen räumlichen Verteilung der Wuchsformen, in den graduellen Übergängen zwischen Landbedeckungsklassen und in der hohen inner- und interannuellen Variabilität der Vegetationsdecke. Vor diesem Hintergrund beschäftigt sich diese Dissertation mit der fernerkundungsbasierten Erfassung und Interpretation der Vegetationsstruktur und der Vegetationsdynamik von Savannen am Beispiel ausgewählter afrikanischer Untersuchungsregionen. Die Vegetationsstruktur wird in dieser Dissertation in Form von Bedeckungsgraden holziger Vegetation, krautiger Vegetation und vegetationsloser Fläche erfasst. Es kommt ein mehrskaliges Verfahren zum Einsatz, in dem höchstaufgelöste IKONOS- und QuickBird-Daten, Landsat-Daten und annuelle MODIS-Zeitreihen ausgewertet werden. Der Ansatz basiert auf der Methodik der Ensemble-Regeressionbäume und stellt eine Erweiterung und Optimierung der Herangehensweise des MODIS-Standardproduktes Vegetation Continuous Fields (VCF) nach Hansen et al. (2002) dar. Beim Vergleich mit unabhängigen Validierungsdaten der nächst höheren Auflösungsebene zeigt sich das Potenzial der vorgestellten Methodik. Die räumliche Übertragbarkeit der Regressionsbäume wird am Beispiel von zwei Vegetationstypen innerhalb der Zentralnamibischen Savanne dargestellt. In diesem Zusammenhang zeigt sich der hohe Stellenwert einer optimalen Auswahl an Trainingsdaten mit einer repräsentativen Abdeckung der Wertespanne aller existierenden Bedeckungsgrade. Die erarbeiteten Resultate unterstreichen, die optimale Eignung der Subpixel-Bedeckungsgrade, gerade zur Beschreibung von Savannenlandschaften. In der Kombination von herkömmlichen, diskreten Landbedeckungs- oder Vegetationskarten mit Informationen zu Bedeckungsgraden wird ein besonderer Mehrwert für weiterführende Analysen gesehen. Die Dynamik der Savannenvegetation wird in dieser Arbeit sowohl auf biannueller als auch auf mehrjähriger Skala charakterisiert. Bei der biannuellen Analyse werden die Veränderungen der holzigen Vegetationsbedeckung zwischen den Jahren 2003/04 und 2006/07 erfasst. Hierfür findet eine zeitliche Übertragung des zuvor vorgestellten Verfahrens zur Ableitung von Bedeckungsanteilen statt. Im Rahmen der biannuellen Untersuchungen können Veränderungsflächen identifiziert werden, ohne Einschränkung auf Übergänge zwischen fest definierten Klassengrenzen. In Ergänzung der biannuellen Analysen werden aus MODIS-EVI- und Niederschlagszeitreihen Maßzahlen abgeleitet, die den Zusammenhang zwischen Niederschlag und Vegetationsentwicklung, die Variabilität und die Trends der Vegetation über einen Zeitraum von acht Jahren beschreiben. Hierbei kommen beispielsweise Korrelationsanalysen zwischen Vegetationsindex- und Niederschlagszeitreihen zum Einsatz. Zudem werden Trendanalysen der Vegetationsindex-Zeitreihen durchgeführt. Die Trends werden einerseits allein aus den Zeitreihen der Vegetationsindizes ermittelt, andererseits wird bei der Berechnung von Restrends (Residual Trends) der Einfluss des Niederschlags berücksichtigt. Neben den Korrelations- und Trendanalysen werden unterschiedliche Variabilitätsmaße der Vegetationsindex-Zeitreihen genutzt, um die mehrjährige Vegetationsdynamik zu beschreiben. Durch die Kombination von Fernerkundungsdaten unterschiedlicher räumlicher und zeitlicher Auflösungen wird in dieser Dissertation die heterogene Vegetationsstruktur und die komplexe Vegetationsdynamik ausgewählter afrikanischer Savannenökosysteme beschreiben.
Mapping Bushfire Distribution and Burn Severity in West Africa Using Remote Sensing Observations
(2010)
Fire has long been considered to be the main ecological factor explaining the origin and maintenance of West African savannas. It has a very high occurrence in these savannas due to high human pressure caused by strong demographic growth and, concomitantly, is used to transform natural savannas into farmland and is also used as a provider of energy. This study was carried out with the support of the BIOTA project funded by the German ministry for Research and Education. The objective of this study is to establish the spatial and temporal distribution of bushfires during a long observation period from 2000 to 2009 as well as to assess fire impact on vegetation through mapping of the burn severity; based on remote sensing and field data collections. Remote sensing was used for this study because of the advantages that it offers in collecting data for long time periods and on different scales. In this case, the Moderate Resolution Imaging Spectroradiometer (MODIS) satellite instrument at 1km resolution is used to assess active fires, and understand the seasonality of fire, its occurrence and its frequency within the vegetation types on a regional scale. Landsat ETM+ imagery at 30 m and field data collections were used to define the characteristics of burn severity related to the biomass loss on a local scale. At a regional scale, the occurrence of fires and rainfall per month correlated very well (R2 = 0.951, r = -0.878, P < 0.01), which shows that the lower the amount of rainfall, the higher the fire occurrence and vice versa. In the dry season, four fire seasons were determined on a regional scale, namely very early fires, which announce the beginning of the fires, early and late fires making up the peak of fire in December/January and very late fires showing the end of the fire season and the beginning of the rainy season. Considerable fire activity was shown to take place in the vegetation zones between the Forest and the Sahel areas. Within these zones, parts of the Sudano-Guinean and the Guinean zones showed a high pixel frequency, i.e. fires occurred in the same place in many years. This high pixel frequency was also found in most protected areas in these zones. As to the kinds of land cover affected by fire, the highest fire occurrence is observed within the Deciduous woodlands and Deciduous shrublands. Concerning the burn severity, which was observed at a local scale, field data correlated closely with the ΔNBR derived from Landsat scenes of Pendjari National Park (R2 = 0.76). The correlation coefficient according to Pearson is r = 0.84 and according to Spearman-Rho, the correlation coefficient is r = 0.86. Very low and low burn severity (with ΔNBR value from 0 to 0.40) affected the vegetation weakly (0-35 percent of biomass loss) whereas moderate and high burn severity greatly affected the vegetation, leading to up to 100 percent of biomass loss, with the ΔNBR value ranging from 0.41 to 0.99. It can be seen from these results that remotely sensed images offer a tool to determine the fire distribution over large regions in savannas and that the Normalised Burn Ratio index can be applied to West Africa savannas. The outcomes of this thesis will hopefully contribute to understanding and, eventually, improving fire regimes in West Africa and their response to climate change and changes in vegetation diversity.
The urban micro climate has been increasingly recognised as an important aspect for urban planning. Therefore, urban planners need reliable information on the micro climatic characteristics of the urban environment. A suitable spatial scale and large spatial coverage are important requirements for such information. This thesis presents a conceptual framework for the use of airborne hyperspectral data to support urban micro climate characterisation, taking into account the information needs of urban planning. The potential of hyperspectral remote sensing in characterising the micro climate is demonstrated and evaluated by applying HyMap airborne hyperspectral and height data to a case study of the German city of Munich. The developed conceptual framework consists of three parts. The first is concerned with the capabilities of airborne hyperspectral remote sensing to map physical urban characteristics. The high spatial resolution of the sensor allows to separate the relatively small urban objects. The high spectral resolution enables the identification of the large range of surface materials that are used in an urban area at up to sub-pixel level. The surface materials are representative for the urban objects of which the urban landscape is composed. These spatial urban characteristics strongly influence the urban micro climate. The second part of the conceptual framework provides an approach to use the hyperspectral surface information for the characterisation of the urban micro climate. This can be achieved by integrating the remote sensing material map into a micro climate model. Also spatial indicators were found to provide useful information on the micro climate for urban planners. They are commonly used in urban planning to describe building blocks and are related to several micro climatic parameters such as temperature and humidity. The third part of the conceptual framework addresses the combination and presentation of the derived indicators and simulation results under consideration of the planning requirements. Building blocks and urban structural types were found to be an adequate means to group and present the derived information for micro climate related questions to urban planners. The conceptual framework was successfully applied to a case study in Munich. Airborne hyperspectral HyMap data has been used to derive a material map at sub-pixel level by multiple endmember linear spectral unmixing. This technique was developed by the German Research Centre for Geosciences (GFZ) for applications in Dresden and Potsdam. A priori information on building locations was used to support the separation between spectrally similar materials used both on building roofs and non-built surfaces. In addition, surface albedo and leaf area index are derived from the HyMap data. The sub-pixel material map supported by object height data is then used to derive spatial indicators, such as imperviousness or building density. To provide a more detailed micro climate characterisation at building block level, the surface materials, albedo, leaf area index (LAI) and object height are used as input for simulations with the micro climate model ENVI-met. Concluding, this thesis demonstrated the potential of hyperspectral remote sensing to support urban micro climate characterisation. A detailed mapping of surface materials at sub-pixel level could be performed. This provides valuable, detailed information on a large range of spatial characteristics relevant to the assessment of the urban micro climate. The developed conceptual framework has been proven to be applicable to the case study, providing a means to characterise the urban micro climate. The remote sensing products and subsequent micro climatic information are presented at a suitable spatial scale and in understandable maps and graphics. The use of well-known spatial indicators and the framework of urban structural types can simplify the communication with urban planners on the findings on the micro climate. Further research is needed primarily on the sensitivity of the micro climate model towards the remote sensing based input parameters and on the general relation between climate parameters and spatial indicators by comparison with other cities.
Increasing urbanisation is one of the biggest pressures to vegetation in the City of Cape Town. The growth of the city dramatically reduced the area under indigenous Fynbos vegetation, which remains in isolated fragments. These are subject to a number of threats including atmospheric deposition, atypical fire cycles and invasion by exotic plant and animal species. Especially the Port Jackson willow (Acacia saligna) extensively suppresses the indigenous Fynbos vegetation with its rapid growth.
The main objective of this study was to investigate indicators for a quick and early prediction of the health of the remaining Fynbos fragments in the City of Cape Town with help of remote sensing.
First, the productivity of the vegetation in response to rainfall was determined. For this purpose, the Enhanced Vegetation Index (EVI), derived from Terra MODIS data with a spatial resolution of 250m, and precipitation data of 19 rainfall stations for the period from 2000 till 2008 were used. Within the scope of a flexible regression between the EVI data and the precipitation data, different lags of the vegetation response to rainfall were analysed. Furthermore, residual trends (RESTREND) were calculated, which result from the difference between observed EVI and the one predicted by precipitation. Negative trends may suggest a degradation of the habitats. In addition, the so-called Rain-use Efficiency (RUE) was tested in this context. It is defined as the ratio between net primary production (NPP) – represented by the annual sum of EVI – and the annual rainfall sum. These indicators were analysed for their suitability to determine the health of the indigenous Fynbos vegetation.
Furthermore, the degree of dispersal of invasive species especially the Acacia saligna was investigated. With the specific characteristics of the tested indicators and the spectral signature of Acacia saligna, i.e. its unique reflectance over the course of the year, the dispersal was estimated. Since the growth of invasive species dramatically reduces the biodiversity of the fragments, their presence is an important factor for the condition of ecosystem health.
This work focused on 11 test sites with an average size of 200ha, distributed over the whole area of the City of Cape Town. Five of these fragments are under conservation and the others shall be protected in the near future, too, which makes them of special interest. In January 2010, fieldwork was undertaken in order to investigate the state and composition of the local vegetation.
The results show promising indicators for the assessment of ecosystem health. The coefficients of determination of the EVI-rainfall regression for Fynbos are minor, because the reaction of this vegetation type to rainfall is considerably lower than the one of the invasive species. Thus, a good distinction between indigenous and alien vegetation is possible on the basis of this regression. On the other hand, the RESTREND method, for which the regression forms the basis, is only of limited use, since the significance of these trends is not given for Fynbos vegetation. Furthermore, the RUE has considerable potential for the assessment of ecosystem health in the study area. The Port Jackson willow has an explicitly higher EVI than the Fynbos vegetation and thus its RUE is more efficient for a similar amount of rainfall. However, it has to be used with caution, because local and temporal variability cannot be extinguished in the study area over the rather short MODIS time series.
These results display that the interpretation of the indicators has to be conducted differently from the literature, because the element of invasive species was not considered in most of the previous papers. An increase in productivity is not necessarily equivalent with an improvement in health of the fragment, but can indicate a dispersal of Acacia saligna. This shows the general problem of the term ‘degradation’ which in most publications so far is only measured by productivity and other factors like invasive species are disregarded.
On the basis of the EVI-rainfall regression and statistical measures of the EVI, the distribution of invasive species could be delineated. Generally, a strong invasion of the Port Jackson willow was discovered on the test sites. The results display that a reasoned and sustainable management of the fragments is essential in order to prevent the suppression of the indigenous Fynbos vegetation by Acacia saligna. For this purpose, remote sensing can give an indication which areas changed so that specific field surveys can be undertaken and subsequent management measures can be determined.
Agriculture is mankind’s primary source of food production and plays the key role for cereal supply to humanity. One of the future challenges will be to feed a constantly growing population, which is expected to reach more than nine billion by 2050. The potential to expand cropland is limited, and enhancing agricultural production efficiency is one important means to meet the future food demand. Hence, there is an increasing demand for dependable, accurate and comprehensive agricultural intelligence on crop production. The value of satellite earth observation (EO) data for agricultural monitoring is well recognized. One fundamental requirement for agricultural monitoring is routinely updated information on crop acreage and the spatial distribution of crops. With the technical advancement of satellite sensor systems, imagery with higher temporal and finer spatial resolution became available. The classification of such multi-temporal data sets is an effective and accurate means to produce crop maps, but methods must be developed that can handle such large and complex data sets. Furthermore, to properly use satellite EO for agricultural production monitoring a high temporal revisit frequency over vast geographic areas is often necessary. However, this often limits the spatial resolution that can be used. The challenge of discriminating pixels that correspond to a particular crop type, a prerequisite for crop specific agricultural monitoring, remains daunting when the signal encoded in pixels stems from several land uses (mixed pixels), e.g. over heterogeneous landscapes where individual fields are often smaller than individual pixels.
The main purposes of the presented study were (i) to assess the influence of input dimensionality and feature selection on classification accuracy and uncertainty in object-based crop classification, (ii) to evaluate if combining classifier algorithms can improve the quality of crop maps (e.g. classification accuracy), (iii) to assess the spatial resolution requirements for crop identification via image classification.
Reporting on the map quality is traditionally done with measures that stem from the confusion matrix based on the hard classification result. Yet, these measures do not consider the spatial variation of errors in maps. Measures of classification uncertainty can be used for this purpose, but they have attained only little attention in remote sensing studies. Classifier algorithms like the support vector machine (SVM) can estimate class memberships (the so called soft output) for each classified pixel or object. Based on these estimations, measures of classification uncertainty can be calculated, but it has not been analysed in detail, yet, if these are reliable in predicting the spatial distribution of errors in maps. In this study, SVM was applied for the classification of agricultural crops in irrigated landscapes in Middle Asia at the object-level. Five different categories of features were calculated from RapidEye time series data as classification input. The reliability of classification uncertainty measures like entropy, derived from the soft output of SVM, with regard to predicting the spatial distribution of error was evaluated. Further, the impact of the type and dimensionality of the input data on classification uncertainty was analysed. The results revealed that SMVs applied to the five feature categories separately performed different in classifying different types of crops. Incorporating all five categories of features by concatenating them into one stacked vector did not lead to an increase in accuracy, and partly reduced the model performance most obviously because of the Hughes phenomena. Yet, applying the random forest (RF) algorithm to select a subset of features led to an increase of classification accuracy of the SVM. The feature group with red edge-based indices was the most important for general crop classification, and the red edge NDVI had an outstanding importance for classifying crops. Two measures of uncertainty were calculated based on the soft output from SVM: maximum a-posteriori probability and alpha quadratic entropy. Irrespective of the measure used, the results indicate a decline in classification uncertainty when a dimensionality reduction was performed. The two uncertainty measures were found to be reliable indicators to predict errors in maps. Correctly classified test cases were associated with low uncertainty, whilst incorrectly test cases tended to be associated with higher uncertainty.
The issue of combining the results of different classifier algorithms in order to increase classification accuracy was addressed. First, the SVM was compared with two other non-parametric classifier algorithms: multilayer perceptron neural network (MLP) and RF. Despite their comparatively high classification performance, each of the tested classifier algorithms tended to make errors in different parts of the input space, e.g. performed different in classifying crops. Hence, a combination of the complementary outputs was envisaged. To this end, a classifier combination scheme was proposed, which is based on existing algebraic operators. It combines the outputs of different classifier algorithms at the per-case (e.g. pixel or object) basis. The per-case class membership estimations of each classifier algorithm were compared, and the reliability of each classifier algorithm with respect to classifying a specific crop class was assessed based on the confusion matrix. In doing so, less reliable classifier algorithms were excluded at the per-class basis before the final combination. Emphasis was put on evaluating the selected classification algorithms under limiting conditions by applying them to small input datasets and to reduced training sample sets, respectively. Further, the applicability to datasets from another year was demonstrated to assess temporal transferability. Although the single classifier algorithms performed well in all test sites, the classifier combination scheme provided consistently higher classification accuracies over all test sites and in different years, respectively. This makes this approach distinct from the single classifier algorithms, which performed different and showed a higher variability in class-wise accuracies. Further, the proposed classifier combination scheme performed better when using small training set sizes or when applied to small input datasets, respectively.
A framework was proposed to quantitatively define pixel size requirements for crop identification via image classification. That framework is based on simulating how agricultural landscapes, and more specifically the fields covered by one crop of interest, are seen by instruments with increasingly coarser resolving power. The concept of crop specific pixel purity, defined as the degree of homogeneity of the signal encoded in a pixel with respect to the target crop type, is used to analyse how mixed the pixels can be (as they become coarser) without undermining their capacity to describe the desired surface properties (e.g. to distinguish crop classes via supervised or unsupervised image classification). This tool can be modulated using different parameterizations to explore trade-offs between pixel size and pixel purity when addressing the question of crop identification. Inputs to the experiments were eight multi-temporal images from the RapidEye sensor. Simulated pixel sizes ranged from 13 m to 747.5 m, in increments of 6.5 m. Constraining parameters for crop identification were defined by setting thresholds for classification accuracy and uncertainty. Results over irrigated agricultural landscapes in Middle Asia demonstrate that the task of finding the optimum pixel size did not have a “one-size-fits-all” solution. The resulting values for pixel size and purity that were suitable for crop identification proved to be specific to a given landscape, and for each crop they differed across different landscapes. Over the same time series, different crops were not identifiable simultaneously in the season and these requirements further changed over the years, reflecting the different agro-ecological conditions the investigated crops were growing in. Results further indicate that map quality (e.g. classification accuracy) was not homogeneously distributed in a landscape, but that it depended on the spatial structures and the pixel size, respectively. The proposed framework is generic and can be applied to any agricultural landscape, thereby potentially serving to guide recommendations for designing dedicated EO missions that can satisfy the requirements in terms of pixel size to identify and discriminate crop types.
Regarding the operationalization of EO-based techniques for agricultural monitoring and its application to a broader range of agricultural landscapes, it can be noted that, despite the high performance of existing methods (e.g. classifier algorithms), transferability and stability of such methods remain one important research issue. This means that methods developed and tested in one place might not necessarily be portable to another place or over several years, respectively. Specifically in Middle Asia, which was selected as study region in this thesis, classifier combination makes sense due to its easy implementation and because it enhanced classification accuracy for classes with insufficient training samples. This observation makes it interesting for operational contexts and when field reference data availability is limited. Similar to the transferability of methods, the application of only one certain kind of EO data (e.g. with one specific pixel size) over different landscapes needs to be revisited and the synergistic use of multi-scale data, e.g. combining remote sensing imagery of both fine and coarse spatial resolution, should be fostered. The necessity to predict and control the effects of spatial and temporal scale on crop classification is recognized here as a major goal to achieve in EO-based agricultural monitoring.