Improving Large Area Land Cover Classification Using Multi Temporal Remote Sensing Data

Land cover is described in other studies as the (bio)physical cover of the Earth’s surface and includes vegetated areas, artificial areas, bare areas and water bodies. Land cover is prone to changes due to anthropogenic activities and natural processes. These changes influence climate, e.g. by their effect on emissions of CO2 and other greenhouse gases and changes in carbon storage capacity. Therefore, accurate and continuous information on land cover is needed on a global scale. User requirements analysis conducted by the Climate Change Initiative Land Cover consortium (CCI-LC) proved that current land cover products derived from remotely sensed data are lacking accuracy and consistency. These issues often arise due to the inability of the input data to capture temporal dynamics by using a limited time span. Furthermore, land cover changes are often not taken into account in current classification approaches. This research aims to improve current classification approaches by investigating 1) how time series parameters, e.g. phenological metrics, can be extracted from multi-temporal MERIS data and 2) how these can be utilized for classification purposes. Furthermore, a comparison was made between classification results with and without these parameters in order 3) to determine to what extent these influence the classification result. In addition, given the fact that vegetation is highly dynamic, another goal of this study was to investigate 4) how temporally stable locations can be separated from unstable areas in order to ultimately limit classification to the stable period within a time series. The use of phenological metrics was emphasized during this study in order to include vegetation dynamics in the classification approach. During this study an operational method was developed to extract phenological metrics from MTCI and NDVI time series which were successfully used for land cover classification. The use of this method seems to increase the overall accuracy of the classification results and has the potential to be used on a large scale. In addition, an explorative study was conducted on the separation of temporary land cover change from permanent land cover change. This resulted in a fast method that may be effectively added to the classification process and applied on a larger scale.

Land cover assessment and monitoring of its dynamics are essential requirements for the sustainable management of natural resources, environmental protection, food security, humanitarian programmes as well as core data for monitoring and modelling. Land Cover (LC) data are therefore fundamental in fulfilling the mandates of many United Nations (UN), international and national institutions and programmes. Despite the recognition of such importance, current users of LC data still lack access to sufficient reliable or comparable baseline LC data. These data are essential to tackle the increasing concerns in regard to food security, environmental degradation, and climate change. Critically, maintaining and restoring land resources plays a vital task in tackling climate change, securing biodiversity, and maintaining crucial ecosystem services, while ensuring resilient livelihoods and food security.

The development of effective methodologies for the analysis of multi-temporal data is one of the most important and challenging issues that the remote sensing community will face in the coming years. Its importance and timeliness are directly related to the ever-increasing quantity of multi-temporal data provided by the numerous remote sensing satellites that orbit our planet. The synergistic use of multi-temporal remote sensing data and advanced analysis methodologies results in the possibility of solving complex problems related to the monitoring of the Earth's surface and atmosphere at different scales. However, the advances in the methodologies for the analysis of multi-temporal data have been significantly under-illuminated with respect to other remote sensing data analysis topics. In addition, the link between the end-users' needs and the scientific community needs to be strengthened.This volume of proceedings contains 43 contributions from researchers representing academia, industry and governmental organizations. It is organized into three thematic sections: Image Analysis and Algorithms; Analysis of Synthetic Aperture Radar Data; Monitoring and Management of Resources.

Filling the need for a comprehensive book that covers both theory and application, Remote Sensing of Land Use and Land Cover: Principles and Applications provides a synopsis of how remote sensing can be used for land-cover characterization, mapping, and monitoring from the local to the global scale. With contributions by leading scientists from aro

Written by world renowned scientists, this book provides an excellent overview of a wide array of methods and techniques for the processing and analysis of multitemporal remotely sensed images. These methods and techniques include change detection, multitemporal data fusion, coarse-resolution time series processing, and interferometric SAR multitemporal processing, among others. A broad range of multitemporal datasets are used in their methodology demonstrations and application examples, including multispectral, hyperspectral, SAR and passive microwave data. This book features a variety of application examples covering both land and aquatic environments. Land applications include urban, agriculture, habitat disturbance, vegetation dynamics, soil moisture, land surface albedo, land surface temperature, glacier and disaster recovery. Aquatic applications include monitoring water quality, water surface areas and water fluctuation in wetland areas, spatial distribution patterns and temporal fluctuation trends of global land surface water, as well as evaluation of water quality in several coastal and marine environments. This book will help scientists, practitioners, students gain a greater understanding of how multitemporal remote sensing could be effectively used to monitor our changing planet at local, regional, and global scales.

Land degradation and desertification are amongst the most severe threats to human welfare and the environment, as they affect the livelihoods of some 2 billion people in the world's drylands, and they are directly connected to pressing global environmental problems, such as the loss of biological diversity or global climate change. Strategies to combat these processes and mitigate their effects at the land-management and policy level require spatially explicit, up-to-date information, which can be provided based on remote sensing data and using geoinformation processing techniques. Recent Advances in Remote Sensing and Geoinformation Processing for Land Degradation Assessment introduces the current state of the art in this field and provides an overview of both conceptual and technological advances of the recent past. With a specific focus on desertification and land degradation, the volume covers the assessment of related biophysical indicators, as well as complementary qualitative information at different spatial and temporal scales. It is shown how remote sensing data may be utilized in the context of assessing and monitoring affected ecosystems and how this information may be assimilated into integrated interpretation and modelling concepts. In addition, different case studies are provided to demonstrate the implementation of these methods in the frame of different local settings. The volume will be of interest to scientists and students working at the interface of ecosystem services, land degradation/desertification, spatial ecology, remote sensing and spatial modelling, as well as to land managers and policy makers.

"An increasingly important application of remote sensing is to provide decision support during emergency response and disaster management efforts. Land cover maps constitute one such useful application product during disaster events; if generated rapidly after any disaster, such map products can contribute to the efficacy of the response effort. In light of recent nuclear incidents, e.g., after the earthquake/tsunami in Japan (2011), our research focuses on constructing rapid and accurate land cover maps of the impacted area in case of an accidental nuclear release. The methodology involves integration of results from two different approaches, namely coarse spatial resolution multi-temporal and fine spatial resolution imagery, to increase classification accuracy. Although advanced methods have been developed for classification using high spatial or temporal resolution imagery, only a limited amount of work has been done on fusion of these two remote sensing approaches. The presented methodology thus involves integration of classification results from two different remote sensing modalities in order to improve classification accuracy. The data used included RapidEye and MODIS scenes over the Nine Mile Point Nuclear Power Station in Oswego (New York, USA). The first step in the process was the construction of land cover maps from freely available, high temporal resolution, low spatial resolution MODIS imagery using a time-series approach. We used the variability in the temporal signatures among different land cover classes for classification. The time series-specific features were defined by various physical properties of a pixel, such as variation in vegetation cover and water content over time. The pixels were classified into four land cover classes - forest, urban, water, and vegetation - using Euclidean and Mahalanobis distance metrics. On the other hand, a high spatial resolution commercial satellite, such as RapidEye, can be tasked to capture images over the affected area in the case of a nuclear event. This imagery served as a second source of data to augment results from the time series approach. The classifications from the two approaches were integrated using an a posteriori probability-based fusion approach. This was done by establishing a relationship between the classes, obtained after classification of the two data sources. Despite the coarse spatial resolution of MODIS pixels, acceptable accuracies were obtained using time series features. The overall accuracies using the fusion-based approach were in the neighborhood of 80%, when compared with GIS data sets from New York State. This fusion thus contributed to classification accuracy refinement, with a few additional advantages, such as correction for cloud cover and providing for an approach that is robust against point-in-time seasonal anomalies, due to the inclusion of multi-temporal data."--Abstract.

This volume comprises an outstanding variety of chapters on Earth Observation based time series analyses, undertaken to reveal past and current land surface dynamics for large areas. What exactly are time series of Earth Observation data? Which sensors are available to generate real time series? How can they be processed to reveal their valuable hidden information? Which challenges are encountered on the way and which pre-processing is needed? And last but not least: which processes can be observed? How are large regions of our planet changing over time and which dynamics and trends are visible? These and many other questions are answered within this book “Remote Sensing Time Series Analyses – Revealing Land Surface Dynamics”. Internationally renowned experts from Europe, the USA and China present their exciting findings based on the exploitation of satellite data archives from well-known sensors such as AVHRR, MODIS, Landsat, ENVISAT, ERS and METOP amongst others. Selected review and methods chapters provide a good overview over time series processing and the recent advances in the optical and radar domain. A fine selection of application chapters addresses multi-class land cover and land use change at national to continental scale, the derivation of patterns of vegetation phenology, biomass assessments, investigations on snow cover duration and recent dynamics, as well as urban sprawl observed over time.

The development of effective methodologies for the analysis of multi-temporal data is one of the most important and challenging issues that the remote sensing community will face in the next few years. The relevance and timeliness of this issue are directly related to the ever-increasing quantity of multi-temporal data provided by the numerous remote sensing satellites that orbit our planet. The synergistic use of multi-temporal remote sensing data and advanced analysis methodologies results in the possibility of solving complex problems related to the monitoring of the Earth's surface and atmosphere.This book brings together the methodological aspects of multi-temporal remote sensing image analysis, real applications and end-user requirements, presenting the state of the art in this field and contributing to the definition of common research priorities. Researchers and graduate students in the fields of remote sensing, image analysis, and environmental monitoring will appreciate the interdisciplinary approach thanks to the articles written by experts from different scientific communities.