The PDF file contains the front matter associated with SPIE Proceedings Volume 10790, including the Title Page, Copyright information, Table of Contents, Introduction, Author and Conference Committee lists

Machine learning algorithms have shown a surprisingly successful development within the last years. Several data intensive technical and scientific fields – like search engines, speech recognition, and robotics – have an enormous benefit of these developments. Remote sensing tasks belong to data intensive applications as well. Today, remote sensing provides data over a wide range of the electromagnetic spectrum (UV, VIS, NIR, IR, and Radar). The capabilities of the sensors include single band images as well as multi- and even hyperspectral data. Due to the fact that remote sensing applications are often monitoring tasks, long time series data are in the focus of image exploitation. Several machine learning algorithms have been used in the remote sensing community since decades, ranging from basic algorithms such as PCA and K-Means to more sophisticated classification and regression frameworks like SVMs, decision trees, Random Forests, and artificial neural networks. Through a combination of data availability, algorithmic progress, and specialized hardware, deep learning methods and convolutional networks (ConvNets) came in the focus of the image exploitation community during the last years and are now on the verge between revolutionary success and illusionary hype. This overview aims to explore in which situations these new approaches are useful in remote sensing applications, which problems are actually solved, and which are still open.

Semantic land cover classification of satellite images or airborne images is becoming increasingly important for applications like urban planning, road net analysis or environmental monitoring. Sensor orientations or varying illumination make classification challenging. Depending on image source and classification task, it is not always easy to name the most discriminative features for a successful performance. To avoid feature selection, we transfer aspects of a feature-based classification approach to Convolutional Neural Networks (CNNs) which internally generate specific features. As land covering classes, we focus on buildings, roads, low (grass) and high vegetation (trees). Two different approaches will be analyzed: The first approach, using InceptionResNetV2, stems from networks used for image recognition. The second approach constitutes a fully convolutional neural network (DeepLabV3+) and is typically used for semantic image segmentation. Before processing, the image needs to be subdivided into tiles. This is necessary to make the data processible for the CNN, as well as for computational reasons. The tiles working with InceptionResNetV2 are derived from a superpixel segmentation. The tiles for working with DeepLabV3+ are overlapping tiles of a certain size. The advantages of this CNN is that its architecture enables to up-sample the classification result automatically and to produce a pixelwise labeling of the image content. As evaluation data for both approaches, we used the ISPRS benchmark of the city Vaihingen, Germany, containing true orthophotos and ground truth labeled for classification.

Remote sensing is useful technique to not only detect harmful algal bloom but also quantify the concentration of the harmful algae by using surface reflectance data. The hyperspectral image allows to investigate detail spatial information of harmful algae. Atmospheric correction is critical process in order to achieve an accurate optical signal in water surface. However, the influence of atmospheric correction performance on the estimation of PC concentration has rarely been studied. Thus, this study investigated the influence of three atmospheric correction method on the spatial distribution and concentration of phycocyanin (PC) concentration from estimation of the bio-optical algorithms. Atmospheric correction is an important image processing technique of hyperspectral image because the result of the correction is expected to influence bio-optical algorithms for quantifying phycocyanin (PC) concentration. From August to October, four field and airborne monitoring campaigns in the Baekje Weir in South Korea were implemented to measure water surface reflectance. PC concentrations in the surface water were analyzed using freezing and thawing method. The two band ratio algorithm, the three band ratio algorithm, the Li algorithm, and the Simis algorithm were utilized to estimate PC concentration. And, atmospheric correction of hyperspectral image was conducted by MODTRAN 6, ATCOR 4, and an artificial neural network (ANN). The ANN model utilized atmospheric parameters which generated from MODTRAN 6 to simulate water surface reflectance. Bio-optical algorithms were applied to the atmospherically corrected image for generating PC distribution map. Even though the atmospheric correction result from the ANN showed Nash-Sutcliffe efficiency (NSE) values of 0.80 and 0.76 for the training and validation steps, respectively, the PC concentration from the bio-optical algorithms showed NSE values ranged from 0.17 to 0.57. The ANN model was turned out to be required having large quantities of input data in order to have precise simulation performances. Atmospheric correction from MODTRAN 6 showed an NSE value over 0.8, whereas the correction from ATCOR 4 had a negative NSE value. However, the accuracy in certain regions of the reflectance spectra (λ<500 nm and λ>700 nm) was relatively low compared to the other spectra region because of insufficient atmospheric observation data during monitoring period. The relationship between atmospheric correction and bio-optical algorithm performance for MODTRAN 6, ATCOR 4, and ANN simulation showed different estimation of PC concentration in the atmospherically corrected images. The precise atmospheric correction by MODTRAN 6 enabled an accurate bio-optical algorithm performance and thus was critical to generating a satisfactory spatial distribution map of PC. In addition, the Li and Simis algorithms were revealed to be much more sensitive to the performance of the atmospheric correction than the other algorithms. Therefore, this study provides a useful tool for understanding importance of atmospheric correction influence of hyperspectral images on quantification of harmful algal blooms.

Fluvial terraces are significant geomorphological features justifying the presence of rivers at high altitudes and constitutes the remains of the old river along the valleys, qualifying their incision capacity. Fill terraces can preserve lithic artefacts of Paleolithic communities that used the terrace clasts as raw material to manufacture their stone tools. The identification of fluvial terraces from cartography (analogical or digital) is based on different techniques and methods. Geographical Information Systems (GIS) has been used for the detection and identification of these fluvial geomorphologic features. Digital Elevation Model (DEM) derived from airborne Light Detection And Ranging (LiDAR) and slope-derived are the most used in the literature. The use of Remote Sensing (RS) data provides detailed information about the surface features, so it can be very useful to help in the identification of fluvial terraces. Recently, along the fluvial terraces of Minho river (border between Spain and Portugal), a significant amount of Paleolithic artefacts was found. The objective of this work was to study different approaches combining GIS and RS data to identify fluvial terraces and define the staircase levels along the Minho river valley. An approach was tested on the Minho River based on DEM and several auxiliary parameters. The analysis was based on several maps, such as DEM, slope, Normalized Difference Vegetation Index (NDVI), Land Use Land Cover (LULC) and hydrological data. The data were in WGS84 UTM zone 29 (EPSG:32629) and the spatial resolution adopted was 10 meters. Different scenarios were tested and validated in order to find the best methodology.

Since the adoption of the United Nations Educational, Scientific and Cultural Organization (UNESCO) World Heritage Convention in 1972, the UNESCO compiled a list of 1,073 World Heritage Sites, based on cultural, natural and mixed properties, respectively. Designated sites of UNESCO are distinguished by their outstanding universal value for mankind. However, an increasing number of these sites are labeled as “In Danger”, with the possibility that those values will get lost. Currently, 54 UNESCO World Heritages Sites, especially in Africa and the Arab States, are considered endangered, often caused by armed conflicts, natural disasters, and urban sprawl. For the management and monitoring of these sites, Earth observation data offer considerable potentials. The Research Group for Earth Observation (rgeo) as part of the Department of Geography at Heidelberg University of Education developed “Space2Place” as contribution to the Copernicus Master Challenge in 2016. As winner of the category “University Challenge”, the e-learning course aims at empowering stakeholders of UNESCO sites to incorporate Earth observation data into their daily working routines. Backbone of Space2Place is an e-learning environment, established in the framework of the project “Space4Geography” and funded by the DLR from 2013 – 2017. The whole module is developed as a blended-learning approach. It provides key features such as optimized presentation of the learning units on different end user devices, introduction to various Earth observation applications, e.g. deforestation, forest fire mapping, drought mapping, monitoring of air pollution, integration of various optical satellite images, e.g. Sentinel 2, Landsat 8, MODIS, and RapidEye, interactive approach and integration of different media, and interim and final quizzes to check learning success. Similar to other developed learning modules by rgeo, Space2Place is adaptable and can be dynamically combined to personalized learning paths with real-time adaptation of content and complexity, depending on the student’s performance in test units. The introduction to Earth observation methods and tools is linked with an easy-to-use web-based remote sensing application, called “BLIF”. The connected application comprises a large set of useful satellite data processing functions, including image enhancement, analysis and digital image processing tools, which are explained by and linked to practical exercises. The improved understanding of inherent benefits of Earth observation technologies will enable participants of “Space2Place” to incorporate Earth observation data for customized monitoring of ongoing environmental, economic and social processes according to the Sustainable Development Goals (SDGs) and related indicators. Furthermore, participants will stimulate the communication about Earth observation methods, datasets, and tools as well as actively promote Earth observation in a wider context as valuable contribution to preserve our cultural and natural heritage.

The study examines the documentation of the Asinou Monument within the auspices of the H2020-SC6-R&IINCEPTION project. The project realises innovation in 3D modelling of cultural heritage through an inclusive approach for time-dynamic 3D reconstruction of artefacts, built and social environments. It enriches the European identity through understanding of how European cultural heritage continuously evolves over long periods of time. In this study, digital techniques for data acquisition and methodologies for data processing were examined using the Asinou Church as a case study. Asinou Church is a 11th century shrine to the Virgin Mary, located in the Troodos Mountains of Cyprus and is a UNESCO World Heritage Site and contains some of the finest Byzantine wall paintings in Cyprus which date between the 12th to the 17th century. Different techniques, such as photogrammetry, laser scanning, drones, video and photographs were used for the data acquisition of all features of the church, which were then processed to create a 3D model and document the church using Building Information Modeling (BIM). The church was digitally reconstructed in a 3D BIM model, where it was then processed to produce a Heritage building Information Model (H-BIM) in order to create an information database for further study.

Currently, assessing geo-hazards in cultural heritage sites takes place after the geo-hazard has occurred. The long-term vulnerability of cultural heritage is commonly focused on the site itself, in response to environmental risks, without fully considering or understanding the entire geological and geotechnical context. However, the high costs of maintenance of cultural heritage sites directly enforce the prioritisation of the monitoring and conservation policies to ensure sustainable conservation. Monitoring the deformation of structures as well as their surroundings facilitates the early recognition of potential risks and enables effective conservation planning. This paper will present the results of the case study of the UNESCO World Heritage Site of Choirokoitia, Cyprus, where long-term low-impact monitoring systems such as UAVs and geodetic techniques were used to monitor and assess the risk from natural hazards on the archaeological site to evaluate potential geo-hazards.

An effective and accurate study of the slow dynamic behavior of a site or a structure requires the implementation of a distributed monitoring system characterized by a large number of broadband inertial sensors in synergy with FEA (Finite Elements Analysis) technique for its modeling. The sensors are most critical part of this approach are the sensors, that must be sensitive enough not only to measure the effect of critical events (e.g. earthquakes, site subsidence) on the site or structure, but mainly to continuously monitor the effects of continuous natural events (microseism, night and day insulation, etc.) and/or anthropic actions (gas and oil extraction, city car traffic, etc.), which may mine in the long term their stability. Among the possible different sensors architectures, the UNISA Folded Pendulum technological platform is very effective for the implementation of large band inertial sensors (horizontal, vertical and angular), well described by the Extended Folded Pendulum Model (EFPM), developed for a quantitative description of the dynamical behavior of a folded pendulum generically oriented in space. In this paper, after the presentation of the EPFM model, we present monolithic implementations of inertial mechanical seismometers/accelerometers based on the UNISA Folded Pendulum mechanical configuration, optimized for low frequency characterization of large structures, and characterized by large measurement band 10−⁷ Hz ÷ 1 kHz, sensitivity down to 10−¹⁵m/√Hz, low weight (down to 40 g), size (<5 cm), directivity > 10⁴.

This paper will examine the use of traditional photogrammetry and LIDAR for documenting cultural heritage site. The case study area was Foinikas village, in the Limassol district of Cyprus, which dates back to the 11th century and has been abandoned from 1960, following the construction of the nearby Asprokremmos dam. Traditionally, photogrammetry has been used for documentation, by processing aerial images acquired from UAVs. However, with the recent development of new lightweight LiDAR scanners, it is now possible to mount professional grade LiDAR sensors on UAVs, which can be used to document areas with high accuracy. In this study, the abandoned village of Fionikas was documented using both photogrammetry using an RGB camera and a LiDAR scanner attached to a UAV. The results of the study found that both methods used provided high accuracy in the documentation of the site.

The present paper is under the auspices of the ATHENA project, which receives funding from the European Union's Horizon 2020 research and innovation programme under grant agreement No 691936. Work programme H2020 under “Spreading Excellence and Widening Participation”, call: H2020-TWINN-2015: Twinning (Coordination and Support Action).

Cabyle is one of the most important ancient cities in Southern Bulgaria with almost uninterrupted habitation for over a millennium between 4th c. BC and 6th c. AD. Philipp II found it as a Hellenistic town and it was later overbuilt by the biggest Early Roman castrum in Upper Thrace and a later castellum with Episcopal seat, while occasionally devastated by invading Celts, Goths and Avars. Disentangling the details of such dramatic history among the multilayered remains of one of the biggest archaeological sites in Bulgaria is a complex challenge calling for complex solutions. The integration of various remote sensing and geophysical methods at different scale showed the high potential of their complementary use, as well as considerable heuristic value under challenging stratigraphic conditions. The following basic survey methods were applied: satellite imagery; imagery from airborne platforms - airplane and unmanned aerial vehicle (UAV), magnetic gradiometry, ground penetrating radar (GPR), electrical resistivity tomography (ERT) and kappametry. The combination of the chosen techniques provided huge amount of data. Their analysis and interpretation served to solve different tasks related to spatial analysis of ancient city of Cabyle and its surroundings, 3D modelling of the terrain and the ruins of several buildings, directing future archaeological excavation works, searching for new targets and locating completely unknown structures buried under ground, etc. The use of car-towed 16-probe magnetic gradiometer array was instrumental for quick and detailed mapping of extensive areas and turning the focus on prospective locations within the city limits, while also unexpectedly revealing a new bath complex beyond the city walls. ERT helped to map the thickness of the anthropogenic layer, the depth to the fundament rocks and to distinguish geological faults from massive walls. The GPR was most successful in outlining the plan of an important part of the Roman town hidden underground, whereas aerial photography indicated part of massive building with East-West orientation indicated by significant differences in the color of the crops. Archaeological excavations informed by the survey results demonstrated a very close match between the linear features visible both on the vertical gradient map and on the GPR model on one hand and the excavated architectural remains on the other. This case study may serve as a perfect illustration of the usefulness of integrated application of remote sensing methods for informing strategic decision-making and long-term management plans of archaeological parks and monuments.

In 2016, the European Unions’ Research and Innovation program Horizon 2020 launched the multi-national project called HEritage Resiliance Against CLimatic Events on Site (HERACLES). The goal of this project is to design, validate and promote effective and sustainable solutions against potential threats climatic changes can bring about on the cultural heritage. For this purpose, knowledge and experiences of multiple research facilities and versatile groups of end-users in different European countries will be bundled and synergetically benefited of. In this paper, we will provide an overview about HERACLES project while a particular focus will be put on the activities concerning close-range Remote Sensing data processing. By exploiting image and laser data acquired from UAVs and from the ground, we strive for precise and reliable 3D models that are useful for representation of the scene on the desired level of detail and thus for assessing potential damages and estimating risks. From oblique UAV-borne imagery, we will obtain textured airborne 3D models using photogrammetric methods of image alignment, dense point cloud reconstruction, meshing and texturing. To provide coverage for the objects’ interiors, very dense point clouds from terrestrial laser scans can be additionally captured. A triangle mesh is obtained from these points and textured by means of terrestrial high-resolution photos, whereby the registration took place using a point-based method. We identified two main challenges: first, the reconstruction results from high-resolution UAV images were not always satisfying due to a low coverage; to cope with this, extensive interactive corrections must be undertaken. Moreover, seamless merging of triangle meshes provided by aerial photogrammetric reconstruction and by terrestrial laser scans is cumbersome because of varying density and accuracy of 3D points in both meshes.

The present study is a continuation of the previous monitoring studies on floating reed islands based on remote sensing methods, but this time the study is much more precise in order to create a sustainable operating model for subsequent monitoring studies on this specific type of habitats. The aim of this study is to create a precise model for the movement and dynamic of the floating reed islands in Srebarna Lake. This was done by creating a hybrid model (based on optical and SAR data), assessing the actually condition of floating reed islands, and applying it to quantify of the movement of floating reed islands to perform an actual and seasonal habitats monitoring. To create the hybrid model, the advantages of SAR data – Sentinel-1 for the hydrological dynamics monitoring of Srebarna Lake were used. The SAR data used were obtained for different time periods, within the observed seasons. Multispectral satellite data from Sentinel-2 was also used in order to apply an orthogonal transformation model called Tasselled Cap Transformation (TCT). The Tasselled Cap model is a very effective method for classifying and analyzing processes related to the dynamics of changes affecting the main components of the Earth's surface: soil, water, and vegetation. This model proved to be very effective in recognizing specific types of vegetation and habitats, such as floating reed islands and their transformation over a period of time. The results for the reconciliation of TCT images and SAR data define very well the precise boundaries of both the central water body in Srebarna Lake, and the floating reed islands. The results obtained by means of comparative analysis confirm both methods as being equally effective to determine the floating reed islands dynamics in the hybrid model proposed in this study.

Both Terrestrial Laser Scanner and Structure from Motion photogrammetry are considered as reliable methods for collecting relief elevation data in the form of dense point clouds. The point density is their common big advantage compared to classical survey measurements. Although both methods produce millions of XYZ points, the manner of collecting data is different resulting in point clouds with serious differences. In the current study these two methodologies for 3D point cloud generation are compared. As a test site is considered a part of a well mapped active landslide in western Greece. In this landslide a terrestrial laser scanner was deployed for the collection of 3D information of the steep slopes. At the same time UAVs flights were performed and dense point clouds were also generated. The derived products from the TLS and UAV are compared and the results are presented.

Landslides represent one of the greatest natural disasters and that makes their research notably crucial. The development of Sentinel missions, which implemented by European Space Agency (ESA) for the operational needs of the Copernicus programme that is directed by the European Commission, gives a new impetus to Earth Observation, providing timely, operational, easily accessible, global coverage and robust datasets. It is worth mentioning that the specific datasets are freely available to users from a web-based system, called ‘’Copernicus Open Access Hub”. This work deals with the exploitation of Sentinel-1 radar data as well as Sentinel-2 optical data in order to map the sliding areas of an active landslide. The landslide occurred in January 2017 as result of snow melting, while the area is located in Western Peloponnese, Greece. The total length of the landslide zone was about 300 m with a head width of 330 m while the landslide material is constituted from three different geologic formations mirroring the complex geotectonic of the wider area. In that context, radar data as well as optical data were collected and processed appropriately. Concerning radar data, Sentinel-1 images before and after the occurrence of the slide were acquired and interferometric processed yielding to a displacement map through the estimation of vertical displacements. On the other hand, corresponding Sentinel-2 optical data before and after the landslide event were collected and they were submitted to digital processing techniques such as change detection analysis. The results of the two methodologies were compared and also, they were evaluated for their accuracy using the actual displacement values obtained from GPS and UAV surveys.

Extraction of landslides from a pair of Lidar data taken before and after the 2016 Kumamoto, Japan, earthquake was carried out. The spatial correlation coefficient of the two Lidar data was calculated, and the horizontal shift of the April-23 DSM with the maximum correlation coefficient was considered as the crustal movement by the April-16 main-shock. By taking the difference of the co-registered DSMs, the change of the surface elevation was calculated. This elevation change includes many effects due to the earthquake, such as landslides and building collapses, and the other temporal changes, such as parking cars and construction/rescue activities. Thus in this study, only large-scale elevation changes more than plus and minus 2.0 m and the areas of larger than 200 square meters were extracted as possible landslides. The extracted areas were compared with aerial photos taken after the Kumamoto earthquake and other soil movement maps made for this event. The result shows that large-scale landslides were easily extracted by the difference of the DSMs and even ground deformations along surface ruptures, where trees were torn down, could be identified.

The U.S. Geological Survey (USGS) pointed out that landslides caused a loss of billions and thousands of deaths each year all over the world. In Taiwan, because of vulnerable geological conditions, the invasion of typhoon in summer, and over 70% of mountain area in the whole island, landslides not only cause people’s property damage but also threaten personal safety. Current landslide mapping techniques include field inventorying, photogrammetry, and the use of digital elevation models (DEMs) to highlight regions of past instability. Most researchers focus on automatically producing landslide inventory maps by using different kinds of remote sensing data. However, not many researches pay attention to deepseated landslides. Nowadays, deep-seated landslides are detected by experts from DEMs and there are few methods for the automatic detection of deep-seated landslides. One of the landslide mapping methods, contour connection method (CCM), utilizes bare earth light detection and ranging (LiDAR) to consistently detect landslide deposits on a landscape scale in an automated manner. However, CCM with computational complexity not only requires user inputs but focuses on general landslide geometry. This paper focuses on the automatic detection of landslide features by modifying CCM. The detection results were evaluated by the ground truth and marked some features of landslides and high susceptibility areas. It shows that by using the proposed method, the accuracy will be the same as that obtained by using CCM, but with less computation time.

Land use and land cover (LULC) information is an important component for wildfire modeling as an understanding of the underlying vegetation conditions is needed to propagate fires accurately across the region. While government entities around the globe often create their own maps of available fuels, political boundaries within a country or across international borders can create inaccuracies and mismatched labeling that changes the speed, direction, and characteristics of the modeled fire. Government-generated fuels layers can also utilize different classification schemes with the United States relying on the 13 category Anderson Fire Behavior Fuel Model or the 46 category Scott Burgan Fire Behavior Fuel Model and Canada using a 19 category Fire Behavior Prediction System. To address these differences and provide a higher resolution understanding of conditions on the ground, a methodology based on the ISODATA Unsupervised Classification method was developed and applied to 30-meter Landsat 8 imagery. The methodology was utilized to create a fuels layer for Australia in 2016 and the western provinces of Canada in 2017 and will be used in future updates to wildfire models at AIR Worldwide.

2.5D terrain model generation from a data stream provides high quality data, which can be used for assisting situational awareness, conducting operations and training in simulated environments. The objective of our research is to design and implement a real-time texturing and visualization of a 2.5D terrain model from live LiDAR and a RGB data streaming in a high performance remote sensing workflow.

To achieve real-time processing, the incoming data streams are evaluated in small patches. In addition, the calculation time per patch must be lower than the recording/sampling time to ensure a real-time processing. Data meshing and projection of the images onto the mesh cannot be implemented in real-time using an off-the-shelf CPU. However, most of these steps are highly vectorizable (e.g., the projection of each LiDAR point into the camera images). In fact, modern graphics cards are highly specialized in computing such data types. Therefore, all computationally intensive steps were performed in the graphics card. Most of the steps for the terrain model generation have been implemented in CUDA and OpenCL. We compare both technologies regarding calculation times and memory management. The fastest technology was selected for each calculation step. Since the model generation is faster than the data acquisition time, the implemented software is real-time.

Our approach has been embedded and tested in a real-time system consisting of a modern reconnaissance system connected to a ground control station via a radio link. During a flight, a human operator in the ground control station is able to observe a texturized terrain model, which was recently generated. The user is able to zoom in an interesting area.

Operations in a military naval context require a detailed planning and information gathering. For this purpose, remote sensing is a useful technique without in-situ survey. A collaboration of Fraunhofer IOSB (Institute of Optronics, System Technologies and Image Exploitation), DLR (German Aerospace Center), and Bundeswehr Geoinformation Centre ZGeoBw, engineered a desktop Geoinformation (GIS) plug-in to generate bathymetric charts and land-use-classes (vegetation, soil types and minerals) using hyperspectral data of coastal areas. These data are basis for further analyses like trafficability, barrier detection and change detection. To evaluate the potential of satellites launched in the near future with hyperspectral sensors onboard (EnMAP), aerial hyperspectral data (HySpex, AISA) were collected over a test site near the Wismar Bay in Germany and are used to simulate the satellite hyperspectral data with corresponding recording terms. Additionally, a field campaign was conducted at the Wismar Bay to acquire a ground truth dataset for model validation, including soil spectra and water depths as well. For generating the bathymetric charts, the WASI (Water Color Simulator) approach was adapted, which offers additional information besides water depth (e.g. dissolved matter, brightness of sand, relative amount of sea grass and other properties). Resulting bathymetric charts with a depth up to eight meters and unsupervised classifications of land cover are free of artefacts and accurate. A validation process is in progress. The engineered desktop GIS plug-in for HEXAGON ERDAS IMAGINE software was developed using the native SDK in addition with interoperable scripts like Python. The existing plug-in framework is variable and adaptable to different kind of GIS.

This paper presents a case study about the detection of illegal dumps from optical satellite images in a large territory falling in the provinces of Naples and Caserta, Southern Italy. This location is also known with the term "Terra dei Fouchi" because in this area is particularly widespread the phenomenon of waste burning and, over the past decades, there have been many landfills of hazardous waste of industrial origin. In order to contrast this phenomenon, the government of the Campania Region organized some prevention, monitoring and repression activities. In particular, the monitoring activities are employed by periodic inspection of sites, which are often object of illegal deposits (former quarries, illegal dumps, as well as city and country roads). The periodic inspection is usually performed by patrols of the company SMA Campania (the in-house regional company, specialized in environmental protection), and law enforcement agencies. Since early warning of illegal and the quick characterization of the detected sites is a key factor for effective resource allocation in the fight against illegal waste dumping, as part of a project, the periodic monitoring inspections of the local environmental agency have been supported with multi-temporal satellite images. The detection has been performed via expert photo-interpretation in order to achieve a high level of accuracy and reliable maps. However, in order to reduce this time consuming task, a multi-features detection algorithm has been implemented. Detected sites have been then classified according to 4 major characteristics (type, state, location and activity) of the dumps. Moreover, a multi-temporal analysis has been performed and it allowed following the evolution of the phenomenon. This approach was effective in both finding new illegal spills (with associated macro classification) and to follow the evolution (in terms of extension and persistence) of landfills already found in the past.

Detecting onshore hydrocarbon is a major topic for both environmental monitoring and exploration. In this work, a hyperspectral image acquired nearby an old oil extraction site in tropical area is analyzed. The area of interest includes a pit filled with bio-degraded heavy oil, surrounded by herbaceous vegetation and many lagoons.

First, we focused on methodologies that can detect oil pollution in an unsupervised manner. Based on the assumption that such oil pits are rare events in the image, statistical approach for anomalies detection, derived from the Reed-Xiaoli detector, is used. In order to decrease the number false alarms, some a priori knowledge about the spectral signature of the pits and about the background is introduced. This approach succeeds in detecting the pit with very few false alarms.

Hydrocarbon pollution can have an impact on vegetation and leads to change in vegetation (bio)physical parameters (pigments, water content, …), according to species, pollutant type and exposition time . In order to map the polluted area without any a priori knowledge, several un-supervised classification, including an original method of automatic classification combining unmixing approach and SVM (support Vector Machine) are applied and compared. The results are compared with a partial “ground truth map” that has been derived from visual observations on the field, and with areas of stressed vegetation that have been mapped using combination of specific spectral indices. The classification results are consistent with the ground truth map and the retrieved stressed vegetation areas.

The 2011 tsunami caused by the Tohoku Region Pacific Offshore Earthquake caused severe damage. In particular, the city of Rikuzentakata, Iwate Prefecture, a coastal city in the Tohoku district, sustained enormous damage, and long-term revitalization process is in progress. For this disaster, several satellite observations and remote disaster investigations by aerial photography were conducted, and various damage situations were reported. Although there are many interpretations of images, few continuously evaluate the recovery process after the disaster. Here, land cover was investigated using image information observed time-sequentially by high-resolution satellite remote sensing. In addition, the re-urbanization process of the disaster area was evaluated from evaluation index values obtained from several types of filtering analysis. The distribution of debris caused by the tsunami, the change in the characteristics of the bare ground due to debris removal, and the characteristics of the land cover shape were quantitatively evaluated from the regularity and identity of the land cover distribution. In addition, the early stages of re-urbanization, which is still in progress, were effectively evaluated.

The Chilean-Peruvian coastal desert stretches between 8 and 28 degrees southern latitude and is one of the ecologically most extreme geo-ecosystems in the world with annual precipitation of less than 1 mm for some parts. Plant growth is limited to so-called fog oases, in which fog is deposited on highly specialized plant’s surfaces such as Tillandsia spp. so that they can mainly cover their entire water requirements by periodically occurring coastal fog. Against this background, our study shows the potential of a multi-scale remote sensing approach for the monitoring and analysis of fog geo-ecosystem distribution patterns (locally also known as Loma-formation). The results of our study can be used as bio-indicators for climate change. In a first step, a vegetation inventory for the entire study area is established. For this purpose, a convolutional neural network was designed which detects vegetation cover based on WorldView-3 imagery. The results are used to stratify the study area to sites, where detailed analyses are conducted. On this scale, remotely piloted aircraft systems (RPAS) are used on the selected test sites to generate orthophoto-mosaics and digital surface models with a spatial resolution between 3 and 10 cm. From these products, structural characteristics (e.g. degree of coverage, horizontal and periodic patterns) and a detailed recording of Tillandsia vegetation is derived. For this purpose, object-based image analysis modelling techniques are developed. In cooperation with the Heidelberg University Institute for Biodiversity and Plant Systematics, the endemic Tillandsia landbeckii is being examined for vitality and genetic diversity. The relationship between the distribution of Loma-formation and fog properties (like occurrence, frequency and intensity) is analyzed by correlation of the measured field data, which are recorded by a structural network of weather stations. To achieve a supra-regional extrapolation of the vegetation degree and structural features, the approach is extrapolated to very high-resolution, multi-spectral WorldView-3 images. The products derived from the RPAS missions serve as a reference and are used to calibrate and validate the classification of WorldView-3 data. The described approach thus comprises several scale levels. First, the vegetation cover is recorded via satellite images. From these observations, two test sites are selected for a detailed investigation of the vegetation structure parameters. Finally, these results are upscaled to the entire study area. The study shows the integrative approach of several remote sensing products with a multi-level hierarchy. Furthermore, in-situ observations are integrated into the procedure to gain better understanding of the interaction between atmosphere and biosphere in the Chilean-Peruvian coastal desert. The results provide a framework for the protection of fragile fog ecosystems and their endemic species.

The impact of wildfires on society is increasing. Annually burned area, fire season length and fire severity have increased during the last decade in several regions of the world, and a number of tragic events have occurred recently in fire-prone areas. Whereas forest fires are natural phenomena, emergency management must be improved in order to minimise human and economic losses as well as undesired effects to ecosystems not resilient enough to fire. However, wildland fire behaviour is not completely understood. Fire rate of spread depends on terrain, weather and vegetation, but the exact relationship between all involved variables is unknown. Remote sensing can help to solve this issue because it has a great potential to measure real-scale fire behaviour with a high spatio-temporal resolution. In this paper, we propose the use of a Geographic Information System (GIS) to combine wildfire remote sensing data with spatial information about vegetation, weather and terrain. This integrated framework facilitates the systematic analysis of multisource data and the study of observed relationships between variables. We describe which operations may be applied to remote sensing and geospatial data in order to display the observed fire evolution and extract relevant statistical relationships. Among others, fire spread is displayed over a topographic map; burned area and fire perimeter evolution are measured; spatially-explicit rates of spread (ROS) are computed; surface ROS is derived from horizontal ROS using a Digital Elevation Model (DEM); and relationships between ROS, intensity, slope and vegetation type are studied.

Only recently remotely sensed data and image processing techniques have been used in the detection of lithium (Li)- bearing pegmatites because of the current growing importance and demand of Li for the construction of Li-ion batteries for electric cars. The study area of this work was the Fregeneda (Spain)-Almendra (Portugal) region, where different known types of Li-pegmatites have been mapped. The objective of this study is to develop a methodology considering remotely sensed data, capable of identifying Li-pegmatites based on the recognition of the associated alteration halos. For that, Level 1-C cloud free Sentinel-2 images with low vegetation coverage were used. The normalized difference vegetation index (NDVI) was computed and a threshold was considered (NDVI<0.2). This study also aims exploiting the potential of Sentinel-2 data for this purpose. Minerals have important absorption features in the visible and infrared regions. Based on these features and with the aim of determining the most accurate methodology, Sentinel-2 reflectance data was used to compute RGB combinations, band ratios, principal component analysis (PCA) and image classification (supervised). These methodologies allow to predict the occurrence of iron oxides and clay minerals, and to discriminate between non-altered and hydrothermally altered zones. Band ratio 3/8 also predicts the occurrence of Li-bearing minerals. The results of the application of remote sensing data, in particular, Sentinel-2 images and image processing techniques to Li-mineralizations is of great interest to mining companies since they could lead to an increased efficiency and sustainability of mineral exploration.

The main objective of this study is to investigate the environmental impacts of Daitari iron-ore mines zone using Landsat TM (2000), Landsat ETM7+ (2010), Landsat8 (2015) and Landsat8 (2017) satellite data. This paper also delineates to propose a land use- land cover (LULC) classification scheme, evaluate their accuracy in different years and to detect the land cover changes from 2000 to 2017 in this mining area using digital image-processing. Supervised classification methodology using maximum likelihood technique was employed in ENVI 4.5 software. It was observed that the built up area and mining area were increased from 8.32% to 29.13% and 14.85% to 30.7% respectively between 2000 and 2017. Also the area with vegetation land was decreased by 47.16% and waste dump increased from 18.2% to 28.7% in Daitari test site area. Finally the mine waste hazard map has been prepared from the spectral characteristics of digital image analysis, laboratory assessment of different elements present at Daitari iron ore mines and getting the idea of generating various maps like contour map, aspect map, flow direction map etc. of this iron ore mines. From the Markov chain analysis, the projected LULC statistics for the year of 2020 has been calculated and the result suggested that iron ore mines area as well as the settlement area will be increased rapidly whereas, vegetation, forest land, barren land and water bodies will be decreased drastically by the year 2020 leading to severe environmental impact.

The vegetation in the riparian zone of a watercourse influences the water state with multiple factors, first via direct substance discharge and secondly via shadow casting on the water surface. Shadowing directly regulates the solar radiant energy arriving at the water surface. Solar radiation input to aquatic environments is the most important abiotic factor for aquatic flora and fauna habitat development. Thus, to adequately asses the ecological state of water courses it is necessary to quantify the solar surface irradiance E (W/m²) arriving on the water surface. When estimating the solar surface irradiance the complex coherence between incoming solar radiation, atmospheric influences, and spatial-temporal geometries need to be investigated. This work established a work flow to compute the solar surface irradiance for water bodies using different remote sensing data. The work flow was tested on regional level for a section of the river Freiberger Mulde, Saxony, for the year 2016. Product of the calculations is a map visualising the annual sum of the solar surface irradiance (kWh/m²) arriving on the Freiberger Mulde water surface and the surrounding terrain. Based on these information bio-hydrological issues can be further examinated.

The analyses of trends in vegetation dynamics require a profound knowledge of its seasonality. For the determination of the seasonality conventional methods of time series analyses often use a simple averaging of measured values of the identical time in different cycle of the whole time series (e.g. bfast). Then it is assumed that the resulting seasonal portion of a time series is constant and stable for the entire time series. However, analyses of vegetation time series show that trends in vegetation dynamics do not always run steadily, but show structural breaks, especially in regions with high potential for possible landscape changes. For such conversion areas, the assumption of a constant seasonality is not always ensured. The dynamic or variability of the seasonality can have temporal effects by a shift of the start of the season (SOS) or the end of the season (EOS) and therefore also on the length of the vegetation period. To show whether seasonal dynamics can be detected in vegetation time series, two requirements must be fulfilled. (1) High-temporal resolution vegetation information provided for example as MODIS-NDVI. (2) Indicators are needed which allows the description of the variability of seasonality. As a result these metrics allow a better modeling of long-term vegetation dynamics in the trend, taking into account the variability of the seasonality. But at the same time the metrics itself serve as indicators for long term vegetation dynamics. The aim of the present study is to analyse phenological and greenness metrics for the modelling of vegetation dynamics in the nature reserve Konigsbrucker Heide. Detailed analyses of key metrics like SOS and EOS using different metric approaches and interpolation methods are applied and compared. The results show that it is difficult to determine consistent information for example for the trend of single phenology metrics.

Puerto Rico suffered severe damage from the category 5 hurricane (Maria) in September 2017. Total monetary damages are estimated to be ~92 billion USD, the third most costly tropical cyclone in US history. The response to this damage has been tempered and slow moving, with recent estimates placing 45% of the population without power three months after the storm. Consequently, we developed a unique data-fusion mapping approach called the Urban Development Index (UDI) and new open source tool, Comet Time Series (CometTS), to analyze the recovery of electricity and infrastructure in Puerto Rico. Our approach incorporates a combination of time series visualizations and change detection mapping to create depictions of power or infrastructure loss. It also provides a unique independent assessment of areas that are still struggling to recover. For this workflow, our time series approach combines nighttime imagery from the Suomi National Polar-orbiting Partnership Visible Infrared Imaging Radiometer Suite (NPP VIIRS), multispectral imagery from two Landsat satellites, US Census data, and crowd-sourced building footprint labels. Based upon our approach we can identify and evaluate: 1) the recovery of electrical power compared to pre-storm levels, 2) the location of potentially damaged infrastructure that has yet to recover from the storm, and 3) the number of persons without power over time. As of May 31, 2018, declined levels of observed brightness across the island indicate that 13.9% +/- ~5.6% of persons still lack power and/or that 13.2% +/- ~5.3% of infrastructure has been lost. In comparison, the Puerto Rico Electric Power Authority states that less than 1% of their customers still are without power.

In the last few years, image acquisition modes have been developed for spaceborne SAR systems aiming on an enhancement of geometric resolution. This offers new opportunities for the calculation of so-called sub-aperture images and their analysis. Due to the increase of illumination time, one single image can be subdivided into multiple subapertures, which are temporarily closely arranged to each other with a lower geometric resolution than the original image. Focusing on this increase of illumination time, the Staring Spotlight (ST) mode of the German TerraSAR-X (TSX) and TanDEM-X (TDX) satellite constellation is mentionable. Here, the azimuth bandwidth can be divided into subset images, in which e.g. vehicles or ships are still well observable. Consequently, moving targets can be detected without utilizing SAR raw data. For this, only a Single Look Complex (SLC) image has to be considered.

In this study, different stacks consisting of sub-aperture images are calculated, which are used for incoherent change detection. The detection is performed as an adaption of a method for change analysis in SAR time series data developed earlier. This method comprises the detailed description of changes, concerning their categorization and classification. Therefore, suitable features have to be calculated leading to a clear distinction of different change categories.

As test data, TSX SLC images are used which were acquired both in ST and in High Resolution Spotlight (HS) mode. An assessment concerning their suitability for the incoherent change detection method applied to the sub-apertures is given. With future studies, it will be tested, whether the temporal aspect can be meaningfully regarded for the change categorization step. For this, the concept of so-called high activity objects, which were the basis of the earlier developed change analysis scheme, might be of relevance for the analysis of changes in sub-aperture sequences.

This paper discusses the detection of borders vulnerability through remote sensing. Remote sensors enable government to collect rapidly synoptically data from areas where the access is restricted. Accordingly, we propose a preliminary survey of Brazil’s borders and its structures by using remote sensing techniques. The extent of the area is a difficulty, as Brazilian borders measures more than 15,000 kilometers in total and marks the limit to 10 countries. The methodology used can be summarized in the following steps: (i) choosing twelve sample-areas along Brazil’s frontier vis-à-vis each neighboring country, focusing on the existent infrastructure; (ii) defining classes of land-use and land-cover based on the elements of visual image interpretation; (iii) creating an image interpretation key; (iv) measuring spectral signature from different targets. Results allow creating a hierarchical model of bodies related to the visual image interpretation element and spectral curves. This model has been divided between natural and anthropogenic features, it is compared the infrastructure, land-cover and land-use from distinct parts along the Brazilian land border. It is concluded there are serious problems of frontier governance due to Brazil’s border is very heterogeneous.

Earthquake science has entered a new era with the development of space-based technologies to measure surface geophysical parameters and deformation at the boundaries of tectonic plates and large faults. Satellite time-series data, coupled with ground based observations where available, can enable scientists to survey pre-earthquake signals in the areas of strong tectonic activity. Cumulative stress energy in seismic active regions under operating tectonic force manifests various earthquakes’ precursors. Space-time anomalies of Earth’s emitted radiation (thermal infrared in spectral range measured from satellite months to weeks before the occurrence of earthquakes, radon in underground water and soil, etc.), and electromagnetic anomalies are considered as pre-seismic signals. This paper presents observations made using time series MODIS Terra/Aqua and NOAA-AVHRR satellite data for derived multi-parameters land surface temperature (LST), outgoing long-wave radiation (OLR), net surface latent heat flux (LHF), mean air temperature (AT), Relative Humidity (RH), and reanalysis data sets for some moderate or strong seismic events recorded in Vrancea region in Romania, which is one of the most active intracontinental seismic areas in Europe. Starting with almost one week prior to a moderate or strong earthquake a transient thermal infrared rise in LST of several Celsius degrees (oC) and the increased OLR values higher than the normal have been recorded around epicentral areas, function of the magnitude and focal depth, which disappeared after the main shock. These observations support seismic LSAIC (Lithosphere – Surfacesphere – Atmosphere -Ionosphere Coupling) Model.

In 2017, wildfires and bushfires caused an estimated 14 billion USD in losses worldwide with the most severe events occurring in the state of California during October and December. The October outbreak in northern California alone cost USD 10.9 billion to the insurance industry, constituting the fourth largest loss event due to a natural catastrophe in 2017. While catastrophe modeling has begun to help bridge the knowledge gap about the risk of wildfire to life and property, understanding the extent of a wildfire in near real-time allows insurance companies around the world to quantify the loss of property their clients face during an event and proactively respond. To develop this real-time analysis of affected areas and the losses associated with the event, a methodology utilizing optical multi-spectral imagery from DigitalGlobe’s WorldView satellite suite is proposed for delineation of burned and unburned areas. This methodology involves utilizing either the Normalized Burn Ratio or the Normalized Difference Vegetation Index, depending upon which satellite imagery is available. Full-automation of the methodology allows for rapid identification of affected areas to occur around the world based on industry request. The methodology has been tested and proven effective for the October 2017 Tubbs, Atlas, Oakmont, and Nuns fires in California, USA.

Permanent grasslands (meadows and pastures) are the most common agricultural land use type covering 34% (0.65 million hectares) of agricultural land in Latvia. The Common Agriculture Policy (CAP) stipulates that the EU Member States have to designate permanent grasslands, ensure that farmers do not convert or plough them and that the ratio of permanent grasslands to the total agricultural area does not decrease by more than 5% in order to receive support payments. Mapping of grasslands and assessment of their biomass (productivity) is of interest for evaluation of bioeconomical potential. Field sampling is the most precise approach assessment of biomass but it is expensive and timeconsuming when applied to a larger territory. In contrast, remote sensing can provide large coverage and mapping of grass biomass distribution for further use in the assessment of the available fodder for livestock and/or the optimal location for biomass-based renewable energy production sites.

The study was carried out in Cesis Municipality in Latvia using airborne flying laboratory ARSENAL – the constellation of hyperspectral imagers in the visible to mid-wave infrared (400-5000 nm) spectral range, topographic LiDAR and high-resolution RGB camera for simultaneous multi-sensor data acquisition. Hyperspectral data were used for both mapping of grasslands and assessment of grass biomass. Different spectral ranges and machine learning algorithms were tested in order to find the best one. The performance of Sentinel-2 like spectral bands also was tested for further possible further use of multispectral satellite data.

Object Raku’s Tree Species Identifier (TSI) is a multi-faceted and automated process that analyzes LiDAR point cloud data to determine the location and species of individual trees. TSI is a system designed to readily incorporate any tree species and the code was written to enable an almost limitless species catalog. TSI provides stem-based information and has been tested & validated on over 2,000,000 hectares of timber area and has segmented over 650 million trees. TSI outputs adhere to standard GIS specifications and so can be easily integrated into existing GIS networks and applications. TSI requires LiDAR data with point density exceeding 10 pts/m2 and does not require correlated imagery or other remote sensing data. In this paper we describe the process and highlight the challenges and results from an operational case study.

Estimating probability of strata convergence in relation to the distribution of stresses around the excavated wall as well as quantifying the time-dependent displacement are of paramount importance. Convergence is defined as the gradual decrease in the interval between to specified rock units or geological horizons as a result of thinning of intervening strata. The heterogeneous distribution of induced stresses with time resulting into simultaneous strata displacement which is extremely difficult to predict under the robust mining environment. As a result of unpredictable stresses, during the time of subsequent excavation around a specified geologic condition, significant amount of gravitational potential energy decreases due to increasing kinematic energy of overburden. Furthermore, the equilibrium of underground mining conditions is disturbed subsequently with time after the initial uncovering of side wall and roof top. This paper proposes an approach to construct a functional equation of convergence purely based on idealized geological conditions. There are several implications which are associated with the prediction of convergence such as reducing health risk of mining workers, contributions of fractures over the dynamic convergence mechanism and so on.

Landslide is among the most common geologic hazard around the world. They cause many formidable damages and numerous deaths annually. There are multiple causes for every landslide. Landslide causes are classified into three categories, namely, physical, natural, and human. Maps are a valuable and appropriate tool for presenting information on landslide. Landslide susceptibility maps is one of the most useful source of information for landuse planners. A landslide susceptibility map represents zones that have the potential for landslide. These zones are identified by correlation between the past distribution of landslide occurrence and conditioning factors that contribute to landslide. This study employs logistic regression (LR) and artificial neural networks (ANN) models to assess landslide susceptibility in Dodangeh Watershed, Mazandaran Province, Iran. The spatial database included landslide inventory, altitude, slope angle and aspect, plan and profile curvatures, distance from faults and from stream, stream power index (SPI), topographic wetness index (TWI), sediment transport index (STI), terrain roughness index (TRI), landuse and lithology. Validation of the models using receiver operating characteristics and overall accuracy indicates that both models display satisfactory performance, and LR model exhibits the most stable and best performance. Given the outcomes of the study, the LR model, which has an AUC value of 0.872 and an overall accuracy of 82.59%, and the ANN model, which has a AUC value of 0.77 and an overall accuracy of 71% are promising techniques for landslide susceptibility mapping.

The regular acquisition of Earth Observations by remote sensing satellites provides long-term Satellite Image Time Series (SITS). Land surface spectral variability provides the capacity for the assessment of Land Use/Cover Change (LUCC) information through SITS. As the reduction of deforestation rates is a matter of global concern, we selected a test area in the Brazilian Amazon Rainforest to assess LUCC information through long-term SITS. Top of Atmosphere reflectance images acquired from Landsat satellites between 1984 and 2017 were downloaded. A filtering process was carried out through the analysis of cloud and shadow masks. A total of 279 images were used to build a long-term Normalized Difference Vegetation Index (NDVI) SITS for every pixel. Images from across 7 years were used to identify SITS for the classes No Change, Anthropic Change, and Natural Change in order to define reference SITS. The Fast- Dynamic Time Warping (FastDTW) algorithm was used to compute the similarity between the reference SITS and the SITS to be labeled. The K-Nearest Neighbor algorithm was applied to classify the SITS based on the similarity measurements. Two different values of the FastDTW radius parameter were used to build two LUCC maps. The overall accuracies of the LUCC maps were 58.06% and 55.02%, for the radius parameter of one and 20, respectively. It was observed that atmospheric effects, clouds, cloud shadows, smoke, among other noisy agents, can modify the real SITS shape. As a result, the use of raw SITS can lead to a reduction in the accuracy of LUCC maps. Furthermore, high cloud coverage in the Brazilian Amazon Rainforest results in high-frequency irregularity in the SITS, which further reduces the accuracy of the LUCC maps. However, the study showed that the long-term NDVI SITS can describe land cover types and the classes defined despite the constraints mentioned.

To implement the Alte Land Pflanzenschutzverordnung 2015 (Plant Protection Ordinance in the Area of “Altes Land” in the region near Hamburg Germany) regarding the determination of the exposition classes off the ditch structures in thee “Alte Land”, a method is developed based on current optical satellite data that is intended to support the current manual working method. The focus of the investigation is on the three optical satellites Landsat-8, Seentinel-2 and WWorldView-3. It considers both the potential and possible limitations of this data base in the context of the task. The result of this investigation indicates the suitability of optical satellite data for the tasks under the “Alte Land Pflanzenschutzverordnung” 2015.

Land cover changes (LCC) in Mediterranean lands are generally connected with anthropogenization and climate changes. The territory of Haskovo region is located in south-eastern part of Bulgaria, where, in the last years, intensification of degradation processes in landscapes, induced by drought is occurred. There is a real threat from intensification of desertification processes in these landscapes in the next decades. The precise investigation of degradation processes requires in-depth knowledge of landscape components, of the ongoing processes that take place in them and functions that ensures existence and maintenance of ecosystems. In the case study, we have determined the LCC that have occurred in the last 18 years. Remote sensing methods and GIS have been applied. The LCC have been determined by using supervised classification, as for the basis of the classification, vector layers of Land cover and vegetation have been used. Two landscape maps were created. The first is valid to 2000 and the second is a map of the modern landscapes. In order type of the change and its magnitude to be determined, Tasseled Cap Transform (TCT) and modified Change Vector Analysis (mCVA) techniques were applied. The application of this approach allows differentiation of the most affected by the LCC landscapes, determination of the type and direction of change, and assessment of the potential of the landscape to deliver ecosystem services in changing environmental conditions.

The Naipa and Muiane mines are located on the Nampula complex, a stratigraphic tectonic subdivision of the Mozambique Belt, in the Alto Ligonha region. The pegmatites are of the Li-Cs-Ta type, intrude a chlorite phyllite and gneisses with amphibole and biotite. The mines are still active. The main objective of this work was to analyze the pegmatite’s spectral behavior considering ASTER and Landsat 8 OLI data. An ASTER image from 27/05/2005, and an image Landsat OLI image from 02/02/2018 were considered. The data were radiometric calibrated and after atmospheric corrected considered the Dark Object Subtraction algorithm available in the Semi-Automatic Classification Plugin accessible in QGIS software. In the field, samples were collected from lepidolite waste pile in Naipa and Muaine mines. A spectroadiometer was used in order to analyze the spectral behavior of several pegmatite’s samples collected in the field in Alto Ligonha (Naipa and Muiane mines). In addition, QGIS software was also used for the spectral mapping of the hypothetical hydrothermal alterations associated with occurrences of basic metals, beryl gemstones, tourmalines, columbite-tantalites, and lithium minerals. A supervised classification algorithm was employed - Spectral Angle Mapper for the data processing, and the overall accuracy achieved was 80%. The integration of ASTER and Landsat 8 OLI data have proved very useful for pegmatite’s mapping. From the results obtained, we can conclude that: (i) the combination of ASTER and Landsat 8 OLI data allows us to obtain more information about mineral composition than just one sensor, i.e., these two sensors are complementary; (ii) the alteration spots identified in the mines area are composed of clay minerals. In the future, more data and others image processing algorithms can be applied in order to identify the different Lithium minerals, as spodumene, petalite, amblygonite and lepidolite.

Virtual technologies find more and more applications in many areas of our lives, ranging from the entertainment industry to advanced real-time information systems. They can also be successfully used in spatial or geographical information systems, which increases the information potential of such solutions.

What virtual technologies can bring to GIS is related to, among others with the three-dimensionality of imaging and the possibility of its full perception with the use of immersive virtual reality (VR) or augmented reality (AR) solutions. VR/AR solutions offer the possibility of virtualizing various types of data as well as advanced methods of interacting with them, for example with controllers, hands, sight or verbal communication.

An example of this can be image data from systems: active (Range Gated Imaging) or panoramic. New possibilities in the field of visualization of this type of data result in the increase of analytical capabilities, among others in the field of terrafotogrammetry.

The possibility of using virtual technologies in GIS 3D solutions applies to both stationary applications (management centers) and mobile applications (field patrols). Each of these solutions brings a number of synergistic effects (including the effect of cognitive enhancement). The presented content is more adapted to the human perceptive capabilities.

The study contains selected issues and examples of solutions currently being carried out at the Institute of Optoelectronics of the Military University of Technology (IOE-MUT) by the authors.

This work evaluated the performance of an Artificial Neural Network (ANN) in the temporal generalization of the Land Use and Land Cover (LULC) classes in the surroundings of the Salto Grande reservoir, located in a highly urbanized region of the São Paulo State, Brazil. Landsat-8 OLI (Operational Land Imager) multispectral images acquired in 2015, 2016 and 2017 were submitted to an ANN supervised classification. The ANN was trained with the image acquired in May 2015 to recognize five types of land cover (continental waters, forest, bare soil, agricultural area and urbanized area), using a learning rate of 0.1 and momentum of 0.5. As a classifier, the Multilayer Perceptron (MLP) ANN was used and the training algorithm was the backpropagation. To estimate classifications accuracies, checkpoints were randomly selected, and the error matrix was constructed for each date. The measures used in accuracy assessment were kappa, overall accuracy and the commission and omission errors per class. The results show that the image classification of 2015, the same year as the training data, resulted in a kappa index of 0.96, while the 2016 and 2017 classifications had kappa values of 0.72 and 0.74, respectively. Therefore, the experiments carried out to LULC classification from multitemporal scenes using a single-date trained ANN indicate the ANN's generalization capability and its potential in multitemporal analyzes. In addition, the 2016 classification, however, indicates the need to add non-spectral input data, which allows separate types of coverage of the body of water and landfill to be present when similar spectral responses.

Mangrove species inventory and mapping is very important as an effort to preserve the ecosystem and biodiversity of mangrove forests. One way of efficient mangrove species inventory and mapping is to use remote sensing imagery, especially through the analysis of its spectral reflectance pattern. This study aims to map the fourteen mangrove species on Karimunjawa Island, Central Java, Indonesia by: (1) measuring the mangrove species spectral reflectance pattern in the field, (2) characteristic analysis of the mangrove species reflectance pattern, and (3) mapping the dominant mangrove species distribution. The spectral reflectance measurement of mangrove species objects in the field was done by using JAZ EL-350 VIS-NIR (ranges from 300 to 1100 nm). The JAZ field spectrometer was pointed at a distance of 2 cm from the target objects with 10 reading repetitions for each species. Field measurements results were then taken to the laboratory for analysis of spectral reflectance and absorbance patterns, which served as key object recognition in this study. To combine the field and image spectral reflectance patterns, the field reflectance patterns were resampled to the spectral resolution of WorldView-2 image (8 bands, 2 m pixel size). The spectral angle mapper (SAM) method was the used to locate and map the distribution of each targeted mangrove species. As expected, the results showed that the largest difference of spectral curves between species was at the NIR wavelength spectrum (700-900nm). Hence, it is potential to be used as the basis for identification of species mangrove from remote sensing imagery. However, the result of this mapping approach only showed a low accuracy of 62%. The low value of map accuracy was attributed to the inaccuracy in defining threshold in SAM for each class. This study provides a basic understanding of the use of spectral reflectance for mangrove species mapping from remote sensing imagery.

The predominant geologic context of Nappes zone of northern Tunisia is favorable to host Pb-Zn and Fe mineralization. Particularly, the most conspicuous reddish color soil, revealed the tremendous amount of iron oxide minerals around Fe Tamera mining site as well as Sidi Driss site, which are both located in the Nefza district, in northern Tunisia. Reflectance spectroscopy is an alternative approach to traditional discrimination soil’s technique that use absorption features between visible (VIS) and Short-wave (SWIR) wavelengths (350-2500nm) to discriminate the soil mineralogical composition. The analysis of the spectral behavior which is quantified according to the ratio between reflected energy and incident energy on a defined target is applied at the surface of soils. All collected soil samples were located using a Global Positionning System GPS and analyzed with X-Ray Diffraction XRD method. In this study a procedure is proposed for determining the quantitative estimation of oxides/hydroxides within soils based on XRD results. Moreover the discrimination of iron oxide and iron sulfate minerals (hematite, goethite and jarosite respectively) even clay minerals was verified by ground-based spectroradiometer measurements.

The aim of this study is to monitor the post-fire recovery of forest ecosystems on the basis of remote aerospace methods and data. To achieve this goal, a hybrid model for styding the dynamics of recovery processes of forest ecosystems after fire was developed. Based on the Greenness Tasseled cap component, Normalized Differential Greenness Index (NDGI) was obtained and used as input data in combination with vegetation indices (NDVI, MCARI2). NDGI is an index for vegetation dynamic assessment based on orthogonal transformation of satellite images from Sentinel-2. NDGI shows the vegetation dynamic change depending on temporal periods. The values of this index range from +1 to -1. Using NDGI assessment can be made of negative and positive changes of the vegetation. This study uses a new approach for forest ecosystems assessment, based on this index using the Greenness component obtained from orthogonalization of satellite images in combination with generated vegetation indices (NDVI and MCARI2). Optimization of model performance and automatization of Sentinel-2 MSI data processing were conducted. Sentinel-2 MSI model for orthogonalization of multispectral data was used for Tasseled cap transformation. Results obtained by implementation of the proposed approach show that the integrated composite images of NDGI, NDVI and MCARI2 represent the condition of forest ecosystems.

Remote sensing-derived maps contain errors. The magnitude of such errors is evaluated through accuracy assessment based on sub-sampling the total ‘population’ (i.e. map). Several strategies have been proposed to define the optimal sampling design leading to a statistically robust accuracy assessment. In this work, a stratified random sampling approach as proposed by Padilla et al.¹ was applied to validate two burned area (BA) products as part of the ESA’s Firecci project: a SAR-based product generated from S1 data and the MODIS MCD64. The sampling design considers sample allocation as a function of burned area proportion inside each biome. In our study the sampling size was computed as suggested by Olofsson et al.². The objective of this study was to assess to which extent a reduction in the sampling size influences the accuracy metrics. The validation was carried out for BA detected from Sentinel-1 as well as a MODIS based-product, the MCD64, generated for the year 2017 in the Amazon region (8M km2). The reference BA dataset was generated using optical time-series acquired by the Landsat-7 ETM+, Landsat-8 OLI, and Sentinel-2 MSI sensors. The BA products were validated three times: i) over n = 44 sample units (as computed from Olofsson et al.²; ii) considering a sample size of n/2, and iii) considering a sample size of n/4, to test the sensitivity of the accuracy assessment to changes in the sample size over the tropics.

The results showed that halving the sample size while maintaining the stratified allocation method, yielded similar results when compared to the original sample size (differences in OE and CE did not exceed 5% in any of the products, while differences in DC did not exceed 2%). For a sample size of n/4, the validation results were more unstable (differences in DC reached up to 9% and confidence intervals were higher). The results provide evidence for the optimal sampling size for the accuracy assessment of different BA products over the Amazon basin.

In this paper, the negative impact of extreme climate events and urbanization and the subsequent effects on increasing trend of land surface temperature and degradation of urban green system of Bucharest metropolitan region in Romania was assessed using the geospatial and field survey data. Due to anthropogenic and natural factors, urban land covers changes are reflected in the land surfaces albedo changes. The main aim of this paper is to investigate the urban green albedo patterns dynamics due to the impact of atmospheric pollution, deforestation and climate variations on urban green land cover of Bucharest metropolitan area based on satellite remote sensing Landsat TM/ETM/OLI and MODIS Terra/Aqua time series data over 2000–2017 time period. This study is based on MODIS derived biogeophysical parameters land surface BRDF/albedo, LST (land surface temperature) and NDVI products and in-situ monitoring ground data (as air temperature, aerosols distribution, relative humidity, etc.). Due to deforestation in the periurban areas albedo change appears to be the most significant biogeophysical effect in temperate forests. Satellite data and climate station observations show that surface albedo changes of Baneasa forested zone placed close to a large urban area highly respond to atmospheric pollution influence and climate variations. As the physical climate system is very sensitive to surface albedo, urban/periurban vegetation systems could significantly feedback to the projected climate change modeling scenarios through albedo changes.

Forest fires continue to burn large territories, both within and outside Europe. It is suitably to assess fire-induced changes in the vegetation, which in turn affects infiltration, runoff, and erosion potential. Therefore it is important to identify potential areas of concern and prioritize field reconnaissance. The development of a burn severity map will facilitate quantifying of the post-fire assessment phase. In this study the potential of Normalized Burn Index (NBR), Normalized Difference Vegetation Index (NDVI) and normalized difference greenness indices (NDGI) derived from remote sensing methods (satellite data from different sensors Sentinel and Landsat) and Geographical Information System (GIS) have been analyzed for forest fire severity assessment. For more accurate assessment of the fire severity, a hybrid model was developed, using satellite data from different sensors - Sentinel and Landsat. For this purpose, the area, affected by fires occurred in august 2017 on the northwest slopes of the Ajtovska Mountain (East part of the Stara planina mountain) in the Eastern part of Bulgaria was studied. The forest fire events were spread on the area of (508.5 ha) and the affected vegetation was composed by deciduous forests (309.4ha), coniferous (62.4ha), mixed forests (61.4 ha) and grass and shrubs (75.3ha). Through the model developed, results applicable to the actual forest ecosystem conditions for different time intervals have been obtained. These results provide quantitative information about fire severity for distinct forest types, thus allowing for designing relevant fire severity maps.

The digital image processing is extremely important for numerous areas as a significant one is Earth observation. The identification of the land cover using satellite images is very important for most of the economics spheres. The image segmentation is known as a basic option for the process of classification. It works as an improving element for the performance and for the accuracy. The main role behind the image processing is to provide the recognition of the shapes and objects in an image. In this process a segment has a significant role. A segment is actually a homogenous part of any image. The survey of image processing applications shows examining, refining and combining of already outlined segments are featured. The delineation of the segments is momentous, too, when it comes to the quality of the results. The main goal behind this article is to make a comparison between effectiveness of several graph-based image segmentation algorithms in segmenting of the roads. These are the best merge algorithm of Beaulieu, Goldberg and Tilton, the tree merge segmentation of Felzenszwalb, the minimum mean cut segmentation of Wang and Siskind and the normalized cut algorithm of Shi and Malik. The represented methods in this article are used in segmentation of orthophoto image of an urbanized zone including roads. The image is determined as a matrix of pixels, while they are also vectors of the intensity numbers, which are usually registered by the remote sensing sensors. The results from the experiments are shown, discussed and lead to following conclusions. The tree-merge segmentation by Felzenszwalb and Huttenlocher is not suitable in the road segmentation. Thematic preciseness of the normalized cut segmentation by Shi and Malik is not shown the needed accuracy in this experiment. The best merge method by Tilton shows the most satisfying indexes.

In order to recognize the urban development status and select urban spatial development path, this paper, with the spatial analysis method of GIS and the spatial data of Remote Sense, simulates the Harbin urban construction land expansion with two scenarios, the development suitability and ecological importance. According to the landscape ecology method, the spatial patterns of two scenarios are compared and analyzed using patch number (NP), patch area ratio (PLAND), landscape shape index (LSI) and aggregation index (AI). The results show that the degree of fragmentation and dispersion of urban construction land patch in Harbin increases with the expansion intensity. And the degree of spatial fragmentation in ecological importance scenarios is lower than that of development suitability scenarios.