This PDF file contains the front matter associated with SPIE Proceedings Volume 11864, including the Title Page, Copyright information and Table of Contents

The Land Surface Temperature (LST) is among the Essential Climatic Variables defined by the World Meteorological Organization as it enables the monitoring of climate change and in extension can provide an insight about its causes and effects in urban areas. By taking advantage of all the Landsat satellites with thermal sensor, starting from Landsat 4 up to the most recent Landsat 8, we can produce extensive LST timeseries spanning from 1982 to present on global scale. The RSLab has developed in 2017 an application that calculates Landsat LST on-the-fly with configurable emissivity sources. It is built upon Google cloud with the use of Google Earth Engine. It requires no data input, process power from the users or installation of third-party applications, making it very lightweight and fast. It uses a single channel algorithm LST calculation for the whole Landsat archive. USGS Landsat Collection 2 surface temperature product was made available in December 2020 after the initial provisional product which covered only the United States. With the new product available we are able to further validate the RSLab LST within cities and compare it against the USGS one. Radiometers mounted on towers in the center of Heraklion and Basel are used to gather in-situ data for the LSTs assessment. The web application is accessible through: rslab.gr/downloads_LandsatLST.html

The majority of the population in Europe that is exposed to air pollution levels exceeding the WHO limit values lives in metropolitan areas. There are already several studies that assess the linkage between air pollution and adverse effects on health. With the technology at our disposition, today, we can identify air pollution hotspots. The assessment of the pollution situation alone represents, however, only one component of the whole picture. In order to be able to build a scale that identifies the most critical regions in higher need of intervention, also the probability of exposure and the number of people exposed to defined pollution concentrations must be considered. For this purpose, we can benefit from satellite-derived data products of settlement extent, population density and land use. To improve the health risk assessment, novel data sets have been synergistically exploited for the first time. In this work a method is proposed to perform an assessment of the increased health risk within urban areas in Europe due to the exposure to PM2.5 and to calculate the health burden index HBI: a useful parameter for the assessment of health risk that provides a measure of the impact of air pollution and enables to perform comparisons between different cities. This is a first approach showing the potential of this easily scalable tool that can be of support in the decision-making process and in the research on air pollution/health relationship. Further work is required for the verification and tuning of the initial hypotheses by means of validation with real-life data.

In the context of global warming, the urban heat island (UHI) has become more and more serious, aggravating the severity of heat waves impacts, which not only cause pollution and increase energy consumption, but also pose a threat to human life and health. Many studies have shown that urban surfaces are warmer by absorbing more solar radiation than non-urban areas during the day and stay warm at night, which makes a larger UHI effect at night. Considering that urban surface characteristics are one of the main factors leading to UHI effect, it is necessary to study urban land surface temperature. Due to the lack of high-resolution nighttime remote sensing data, most of the studies on nighttime land surface temperature uses weather station data or Modis remote sensing images with a resolution of 1km, which cannot present more detailed urban climate behaviors, and there are certain limitations to urban climate research. Therefore, this paper aims to find a new approach to solve the problems existing in the traditional method and develop a new model to obtain a more accurate and higher resolution nighttime LST. In this paper, Landsat 8 daytime images (with spatial resolution of 30 square meters per pixel) will be used to build a night cooling model by means of multiple regression analysis in order to obtain the nighttime LST, and compare the day and night models to study the impact of UHI effects. The case study is Madrid Municipality (604.45 km², 3.3 million inhabitants).

This study assessed 30 years (1989–2019) of urban heat phenomena variations for the case study of Gaborone City (Botswana) through the analysis of the land surface temperatures (LST). The LST variability was determined using the Landsat Thermal Infrared Sensor (TIRS) band and the land surface emissivity (ε_λ) factor. The study investigated the influence of land-use and land-cover change (%LULC), normalized difference vegetation index (NDVI) and the normalized difference building index (NDBI) on the variability LST. For the city boundary which occupies 190.96 km², the vegetation cover decreased by nearly 40%, built-up area increased by 38.9%, water bodies decreased by 3% and bare-land increased by approximately 4.1%, while the 30-year mean near-air temperature was observed to have increased by +2.6 ºC. The urban LST variations were observed to increase exponentially with LSTmin of -2.5 ºC – 14.4 ºC and LSTmax of 24.4 ºC – 30.2 ºC respectively from 1989-2019. Using multiple linear regression, the mean LST was observed to be inversely proportional to NDVI (-0.934) and directly proportional to NDBI (+0.949). In correlation with %LULC, the land surface temperature increased with increase in density of the built-up area and bare-land but decreased with increase in vegetation cover and water bodies. Regression of NDVI, NDBI and %LUCC indices for the prediction of LST showed their suitability in the estimation of LST in the arid urban environment with R² of 0.996

Ambient air quality (AQ) is a recurrent issue in cities, exacerbated in low- and middle-income countries. Global AQ impacts on health can be assessed using an aggregated health risk indicator (ƩRIs), derived from in-situ stations, chemical transport models (CTMs) or satellite data. AQ monitoring is well covered in the city of Munich (Germany): in-situ stations, POLYPHEMUS CTM with three domains, CAMS-Reanalysis of the regional model ensemble and satellite data from MODIS. From these data sets, the ƩRIs was calculated considering four major air pollutants (NO2, PM10, PM2.5, O3), using their respective relative risk for the mortality-all causes health end-point. Then, the ƩRIs from the models and the satellite data were compared to the ƩRIs in-situ by means of basic statistics, time series and violin plots and the mean relative difference (MRD). Using the ƩRIs allows to further observe the contribution of individual pollutants to the index. For the mortality all causes health end point between 2017 and 2018, ground observations and CTMs show an increase of ca. 12-13% when exposed to ambient air pollution. The difference between traffic and background stations can be observed: ƩRIs in situ mean is higher at the traffic station than at the background stations. This order is however reversed when considering ƩRIs mean from the models. The four CTMs simulate the ƩRIs well and its seasonality is also represented. Most of the data are spread around the mean and the median for all data sets and stations with an overall distribution skewed towards high values. With 0.5<r2<0.6, POLYPHEMUS/DLR yields medium correlation, regardless the domain, while CAMS-Reanalysisreturns high correlation (r2 ≈ 0.8) for all the studied stations. The MRD indicates an underestimation of the ƩRIs by CAMS-Reanalysis, while POLYPHEMUS tends to overestimate it for the larger domains (positive MRD). The difference in the r2 between the two CTMs is due to their singularities: POLYPHEMUS/DLR uses free runs while CAMS-Reg uses a data assimilation process with station measurements (among them the two studied background stations). The overestimation of Johanneskirchen by POLYPHEMUS/DLR comes from its location nearby a power plant and the wind direction. Finally, the very high values in early 2017 can be explained by fireworks, which are not reproduced by models. It is show in this study that estimating a global health risk from air pollution is possible using in-situ measurements, models and satellite. Finally, satellite data can be helpful to assess the ƩRIs worldwide

The paper shows the possibility of using small UAVs with a rotary wing to monitor the state of atmospheric turbulence at different altitudes. The measurements were carried out at the Basic Experimental Observatory (BEC) of the V.E. Zuev Institute of Atmospheric Optics SB RAS. The turbulence spectra at 4, 10, and 27 m, as well as turbulence scale profiles obtained with three DJI Mavic Mini and one DJI Mavic Air quadcopters are reported. The turbulence spectra measured at different altitudes and turbulence scale profiles are compared with the data obtained from three AMK-03 automated meteorological systems installed at the 4-m and 30-m meteorological towers. It has been found that the turbulence spectra obtained with the AMK-03 and quadcopters are generally in a good agreement with some differences observed in the high-frequency spectral region nearby Hz. During the experiment, Kolmogorov turbulence was observed in the atmosphere in a wide frequency range at all altitudes. This type of turbulence was confirmed by both the AMK-03 and quadcopter data. When determining the longitudinal and lateral turbulence scales at altitudes of 4, 10, and 27 m, the least square fit method was used with the von Karman model as the regression curve. The turbulence scales calculated from AMK-03 and quadcopter data are shown to agree well. The condition describing the relation between the longitudinal and lateral scales in an isotropic atmosphere is true to sufficient accuracy.

A building’s window-to-wall ratio (WWR) has critical influence on heat loss, solar gain, and daylighting levels, with implications for visual and thermal comfort as well as energy performance. However, in contrast to characteristics such as floor area, existing building WWRs are rarely available. In this work we present a machine learning based approach to parse windows from drone images and estimate the WWR. Our approach is based on firstly extracting the building 3D geometry from drone images, secondly performing semantic segmentation to detect windows and finally computing the WWR. Experiments show that our approach is effective in estimating WWR from drone images.

It is important for the electricity transmission and distribution (T&D) companies to patrol their own assets frequently in a wide area. however, the cost of patrolling throughout the area is budget threatening. The work on detecting the maintenance places where the vegetation encroachment problems occurred, is labor intensive, costly, and time-consuming, sometimes inapplicable due to the poor accessibility, and is thus, only practical on relatively small areas. Satellite imagery-based monitoring is reasonable and repeatable; hence it has a potential to replace the helicopter surveillance. Sentinel-2 imagery is one of the most famous satellite imageries with completely free of charge, however, its spatial resolution is relatively lower than high-cost satellite imagery such as PlanetScope or WorldView-3. In this research, we explored the effectiveness of super resolution. The refinement of spatial resolution from 10m/pix to 3.3m/pix (x3 SR) seemed to be extremely useful to assess trigonometric risk assessment, which leveraged the number of the pixels between transmission line and vegetation, and tree height information at the vegetation pixels. We employed the deep learning based super resolution model RDN (Residual Dense Network) to upsample the Sentinel-2 images. The training data is generated from the PlanetScope imagery whose resolution is 3.7m/pix. Deep learning based super resolution is generally effective to get 2-4 times finer resolution, therefore, the PlanetScope imagery is suitable to obtain the RDN model for x3 super resolution. We evaluated the performance of vegetation segmentation performance with and without super resolution in the areas along the transmission line. The experimental results showed that the imagery with super resolution yielded better result than the result without super resolution by 9.3% in weighted F1-score.

Among emerging 3D scanning and imaging techniques that are commercially available, simultaneous localization and mapping (SLAM) is being substantially studied to generate 2D/3D maps of an unknown environment while reliably keeping track of the user’s pinpoint locations. Its ubiquitous mobility has demonstrated great mapping capabilities for infrastructures where vertical information may frequently be occluded using unmanned aircraft system (UAS) structure from-motion (SfM) photogrammetry. In addition, indoor mapping with terrestrial laser scanning can be a cumbersome task due to possible multiple scan locations. Intending to provide a cohesive 3D model by fusing point clouds collected via aerial SfM photogrammetry, terrestrial laser scanning (TLS), and SLAM, the purpose of this work is to assess the performance of SLAM point cloud generated by a proprietary mobile backpack laser scanner (BLS). Considering maximum scanning range and information integration strategy as variables, the point clouds generated by the BLS were evaluated against SfM and TLS datasets in terms of the internal consistency as well as external accuracy. TLS, SfM and SAM data collection efforts were made in a typical university campus environment. For the internal consistency, the SLAM-based point cloud with a maximum scanning range of 70 m presented a root mean square error (RMSE) of 2 mm. The SLAM+GNSS-based point cloud presented the lowest internal precision of RMSE = 0.861 m. The SLAM+GNSS 70 point cloud after a fine adjustment of misalignment presented the highest vertical accuracy with an RMSE = 0.069 m, while the point cloud generated from SfM photogrammetry presented RMSE = 0.297 m. The BLS was able to generate point cloud with an accuracy similar to GNSS-RTK surveying and it can be considered as a viable solution for indoor and outdoor mapping applications.

New satellite images of the Earth at night can be achieved with earth observation by continuous remote sensing throughout all day. The images give the most complete view of contemporary global human settlement, especially cities. Beijing is the capital, one of the most important and typical big cities of China. This study estimated economic activities of Beijing using remote sensing nighttime earth surface light data from S-NPP VIIRS DNB night data with corrections, with a focus on the relationship between economic index and city lights. Our study aims to eliminate the influence of cloud, moonlight and atmosphere on artificial light sources at night, in order to achieve more accurate inversion of ground artificial light source information. The results proved that there is a strong linear regression relationship between corrected DNB nighttime data and GDP with 0.79405 fitting coefficient, which was higher than the linear fitting coefficient (0.2817) of average radiance composite images data and GDP. The linear fitting coefficient of the tertiary industry and corrected DNB nighttime data is 0.76102 is higher than 0.1836 of the tertiary industry and average radiance composite images data. Therefore, the approach was provided for the dynamic evaluation of social and economic data, and the developed urban light fusion product will lay a foundation for the derivative application of backend and the inversion and application of night light data in other locations.

Digital Surface Models (DSMs) and Digital Terrain Models (DTMs) are crucial for automatic detection of buildings and other ground features in urban areas. They are the basis for numerous geospatial applications, such as 3D city modelling, 3D mapping infrastructures, drainage pattern studies, etc. LiDAR based DSMs usually contain less holes than image-based DSMs. This is mainly because of the mismatching error acquired due to occlusion in photogrammetry. However, if high-rise buildings are present in the scene, large holes are most likely to occur in LiDAR derived DSM. In this research, a modified algorithm for LiDAR-based DSM to DTM filtering approach is presented. The algorithm deals with the original DSM without aiming for precise outcomes. Firstly, pixels representing holes are automatically detected and labelled with unique value. Since available DTM extraction algorithms are mainly based on finding pixels having minimal elevation, this unique value must belong to the off-terrain to avoid using them as minima. Afterwards, the Network of Ground Points algorithm is applied for extracting DTM from DSM to be further used for building detection purposes. The proposed method is tested on the benchmark dataset for Toronto city provided by the ISPRS working group III / 4. The evaluation analysis was applied based on calculating correctness and completeness of the detected building segments. Results showed the practical effectiveness of the proposed technique in achieving average correctness of 91 %.

Floods are the most common disaster in Canada. As results of rapid urbanization and climate changes, both frequency and risks of floods have been increased in Canadian urbanized areas, where the disasters have usually costlier impacts than in rural areas. Imagery data and technologies of optical remote sensing are helpful and can be applied for urban flood response and pre-disaster preparation. Especially high and very high optical remote sensing can be used for precise mapping of the floodwater distribution in dense urban areas and providing key information for disaster response management. In addition, the geospatial information about urban land surface and urban growth derived from optical remote sensing imagery can be the key inputs for urban flood risk analyses. In recent years, several case studies for different urban flood types, including fluvial (Calgary 2013, Ottawa-Gatineau 2017) and pluvial (the Greater Toronto Area) floods in Canada, have been carried out at Canada Centre for Mapping and Earth Observation, Natural Resources Canada. Methodologies/framework for urban floodwater mapping have been developed based on high resolution optical data, as well as the impacts of urban growth on the urban flash flood risks have been investigated using model simulations with remote sensing derived maps as inputs. This presentation demonstrates results from three Canadian urban flood case studies and introduces remote-sensing-based methodologies for different types of urban floods.

Images acquired by RGB cameras on Unmanned Aerial Vehicles (UAVs) can be particularly useful to detect crowd in urban areas when restrictive conditions are imposed for the sake of public safety or health , such as during the Covid 19 pandemic. Together with acquired images, opportune pattern recognition techniques have to be considered to extract useful information. In this framework, features capturing the semantic rich information encompassed in VHR images have to be computed. In particular, Deep Neural Networks (DNN) have been recently proved to be able to extract useful features from data [1]. Moreover, in a transfer learning approach, a DNN pre-trained on a data set can be used to extract opportune features, named deep-features, from another data set, belonging to a different applicative domain [1], [2]. Here, a transfer learning technique is presented to produce change maps, detecting how people gathering increases, from VHR images. It is based on deep-features computed by using some pre-trained convolutional layers of AlexNet. The proposed methodology has been tested on a data set composed of several synthetic VHR images, that simulate crowd collecting in a park, as they can be acquired by RGB camera on a UAV flying at 10 meters height from the ground. The experimental results show that the proposed technique is able to efficiently detect change due to new people incoming in the scene or people that get away, with a low computational cost and in a near-real time operative mode.

Green urban areas play a major role in the quality of life of the citizens in terms of public health, environment and recreation. They help to improve air quality, control temperature and provide ecosystem services that contribute, like forests, to climate change mitigation. The health of urban tree stands is, therefore, essential to maintain the benefits that the urban and peri-urban parks provide to people and cities. Early detection of a reduction in vegetation health is essential to apply measures in order to prevent tree mortality or damage. The use of remote sensing technology for forest health monitoring has been widely demonstrated but the number of published papers using these techniques applied to urban parks is significantly lower. Nevertheless, the use of remote sensing offers excellent opportunities for monitoring the state of health of urban and peri-urban parks. It is a tool that can be very useful to complement on-site visits and make them more efficient by warning of suspected areas of damage. The main objective of this work was to evaluate the potential of PlanetScope images to estimate the degree of tree vegetation decay in the Cerro Almodovar urban park. It is located in the neighborhood of Aluche, in the Latina district, Madrid (Spain), and it had been observed that there were individuals that were beginning to show defoliation problems. Machine learning techniques were used to generate a mapping of the damage levels in the park as well as information on the uncertainty of the estimates. The good performance of the model obtained encourages further development of remote sensing health monitoring in urban green areas.

Mapping cities is difficult due to the dense and varied development of infrastructures in urban contexts. To adapt cities for modern faculties, regular monitoring of developing infrastructures is essential. The built-up land use map derived from remote sensing images shows how cities expand horizontally across the landscape as they grow over decades. Recent studies have shown that when planning for a more sustainable future in metropolitan settings, adding vertical components in the computation of built-up is beneficial for better understanding. As illustrated in this paper, the built-up area of an urban region is extracted from high-resolution LISS IV satellite sensor image data using deep learning techniques. The height of an urban built-up is estimated from Cartosat-1 stereo images using photogrammetric methods and automatic terrain extraction. This study combines the use of a deep learning model with the use of high-resolution remote sensing data to extract the urban built up volume. The results demonstrate that the western and northwestern areas of the study region see greater changes in urban volume compared to the central business district and other parts of the city. This work demonstrates a step-by-step strategy for studying urban growth patterns. It helps in understanding urban growth, population mobility patterns, societal typology, environmental indicators. The outcome of such study could be useful in the evaluation of various indicators of growth such as economic growth, traffic density conditions, utility planning, policy framework such as master plan preparation, and so on for governmental authorities.