Urban 3

The number and accuracy of global urban maps is constantly growing due to the increased availability of high-level EO data, urban area extraction algorithms and computing resources. This trend is without doubt going to be continued and therefore developing updating strategies for existing urban maps based on up-to-date satellite imagery is a reasonable alternative to the generation of completely new maps.

This work is devoted to the combination and integration of two already existing urban area mapping procedures, namely the Urban EXTractor (UEXT) and the Urban Area Detection Procedure (UADP), for the purpose of updating the urban class of the global land cover map extracted within the ESA Climate Change Initiative – Land Cover (CCI-LC) project. Both algorithms have been originally developed for ENVISAT ASAR Wide Swath Mode data at a spatial resolution of 75 meters in Phase 1 of the CCI-LC project and were successfully adapted and applied to Sentinel-1A data. Due to numerous occurrences of double bounce effects, urban areas are characterized by high radar backscatter values in SAR imagery. Moreover, urban areas are relatively static objects with little temporal changes in terms of radar backscatter and therefore can be easily recognized in multitemporal image stacks due to the fact that – in contrast to built-up areas - the surrounding areas usually change their appearance throughout annual phenological cycles. Starting from the multitemporal image stack, the UEXT algorithm searches for very bright pixels, consecutively used as starting points for a region growing procedure, which is iterated until a specific threshold is reached. The UADP algorithm uses the multitemporal backscatter statistics and texture measures as input for an unsupervised classification procedure. Both methods are capable of extracting accurate information about the spatial distribution of human settlements, in the form of urban probability maps or binary urban/non-urban classifications. Based on these (intermediate) results the update process is performed via a simple majority voting. Finally, object-based post-processing steps have been applied in order to refine and improve the results.

The proposed approach has so far been applied on multiple test sites in arid and semi-arid areas (located in Tunisia, Turkey, Portugal and Israel), as in these areas the lowest classification accuracies were achieved in Phase 1 of the CCI-LC project. The obtained updated urban maps have been evaluated qualitatively through visual analysis and quantitatively through comparison against reference data sets based on manual labeling of random discrete global grid hexagons. Results show that the proposed method is able to significantly improve the accuracy of urban class in the latest global CCI-LC map due to the incorporation of information derived from Sentinel-1A data by UEXT and UADP.

A new generation of satellite instruments, represented by the Suomi National Polar-Orbiting Partnership (Suomi-NPP) Visible Infrared Imaging Radiometer Suite (VIIRS), offer global measurements of nocturnal visible and near-infrared light suitable for urban science research. While many promising urban-focused applications have been developed using nighttime satellite imagery in the past 25 years, most studies to-date have been limited by the quality of the captured imagery and the retrieval methods used in heritage (DMSP/OLS) products. Instead, science-quality products that are temporally consistent, global in extent, and local in resolution were needed to monitor human settlements worldwide —particularly for studies within dense urban areas. Since the first-light images from the VIIRS were received in January 2012, the NASA/Yale Day/Night Band team has worked on maximizing the capabilities of these low-light measurements to generate a wealth of new information useful for understanding urbanization processes, urban functions, and the vulnerability of urban areas to climate hazards. In a recent case study, our team demonstrated that tracking daily dynamic VIIRS nighttime measurements can provide valuable information about the character of the human activities and behaviors that shape energy consumption and vulnerability (Roman and Stokes, 2015). Moving beyond mapping the physical qualities of urban areas (e.g. land cover and impervious area), VIIRS measurements provide insight into the social, economic, and cultural activities that shape energy and infrastructure use. Furthermore, as this time series expands and is merged with other sources of remote sensing data (e.g., GHSL, Landsat-8 and Sentinel 2), VIIRS has the potential to increase our understanding of changes in urban form, structure, and infrastructure—factors that may also influence urban resilience—and how the increasing frequency and severity of climate-related hazards can ultimately affect development pathways and urban policies in the long term.

The recent launch of Sentinel-2A satellite in June 2015 allows the availability of first ever data with a minimum spatial resolution of 10m, 13 spectral bands, wide acquisition coverage and short time revisits. Sentinel-2 images are promised to countless uses at a European and global scale, its contribution over existing data and images processing techniques must be evaluated.

In the framework of the preliminary provision of Sentinel-2 images, this work proposes to evaluate the benefit from the Sentinel-2 images for the production of European Corpernicus Land services.  More precisely, it focuses on urban environment and the update of the High Resolution Layer imperviousness. The High Resolution Layer Imperviousness consists in a 20x20m grid covering all Corpernicus member countries and aims to represent a degree of imperviousness from 0 to 100%. This layer was established in a monitoring prospect of urban areas and was already produced in 2006, 2009, 2012. Nevertheless, update remains a difficult task since urban change is a small and irregular phenomenon.

The characteristics of Sentinel-2 should improve the monitoring of urban environment. First, its fine spectral and spatial resolutions provide detailed information that enhances detection of man-made object and their shapes. Secondly, frequent revisits provide time series that can reduce false positives like bare soil in rural areas. Finally, wide acquisition coverage is more suitable to large scale processing.

In this study, Sentinel-2 images are evaluated on the basis of an operational framework. Classification and change detection tools initially setup for the production of the 2012 layer are used. The approach relies on (1) a separate image classification and (2) a combination of each classification probabilities using a data fusion technique. The classification step is based on a neural network algorithm and the fusion step is performed with the Dempster-Shafer theory of evidence.

A study area was defined in the Czech Republic at the extent of the available Sentinel-2 images. Images extents cover about 70% of the Czech. Detected changes highlight the apparition of new housing and industrial areas. In addition, results show Sentinel-2 images can improve the previous High Resolution Layers by detecting missing objects such as bridges or motorways. A comparative study with Landsat-8 image acquisition was also performed. Results show gains associated with the use of the Sentinel-2A images and more especially the complementarity of the two sensors. Finally, this work shows that Sentinel-2 is suitable to provide high quality information at a large-scale and meets the needs of the Copernicus Land program and its downstream services.

Starting from the beginning of the years 2000, more than half of the global human population is living in urban environments and the dynamic trend of urbanization is growing at an unprecedented speed. Rapid urban growth brings several challenges, including meeting accelerated demand for basic services, infrastructure, and affordable housing (particularly for the nearly 1 billion people living in informal settlements). Moreover, as cities develop, their exposure to climate and disaster risk increases (e.g., almost half a billion urban residents live in coastal areas, thus increasing their vulnerability to storm surges and sea level rise). In this framework, an effective monitoring of urban sprawl represents a key issue to analyze and understand the complexity of urban environments and ensure a sustainable development of urban and peri-urban areas.

To this purpose, the ESA DUE Innovator III SAR4Urban project aims at implementing - in support of its users the World Bank and GEO Global Urban Observation and Information Task for Societal Benefits (GEO SB-04) - a novel service that allows to automatically and reliably derive maps of past and current extent of urban areas by means of archived ERS/ASAR and novel Sentinel-1 data, respectively.

The basic assumption of the intended approach is that given a series of multi-temporal images for a given study area, the temporal dynamics of urban settlements are sensibly different than those of all other non-urban classes. As an example, the backscattering temporal mean of urban areas (due to double bounce reflection) is higher than that of forest areas (which might result in high backscattering in one/few acquisitions due to specific conditions, but in general exhibit lower values).

After applying orbit correction, calibration, and terrain correction to the multi-temporal images available over a region of interest in the selected time interval, for each pixel we extract key temporal statistics (i.e., backscattering temporal mean, standard deviation, minimum, maximum, etc.). It is worth noting that for different pixels in the study area, different number of scenes might be available. However - in the hypothesis of a sufficient minimum number of acquisitions for computing consistent statistics - this does not represent an issue. Indeed, we always expect a more stable behavior of the urban class compared to the others (for which the temporal variability is higher). Heterogeneity features are also extracted to ease the detection of lower-density settlements and, finally, specific unsupervised classification schemes are applied to ERS/ASAR and Sentinel-1 data, respectively.

Output of SAR4Urban will include the 2002-2003 urban extent map of entire Africa derived from ASAR WSM data, as well as the urban extent maps of Athens, Beijing, Los Angeles, Mexico City, Atlanta and the Pearl River Delta derived from ERS-1/2 PRI and ASAR IMP scenes. Moreover, the current built-up extent of both these and several African cities will be delineated by means of Sentinel-1A imagery. Experimental results are extremely promising and confirm the great potential of ESA SAR data for mapping urbanization over time and the capability of the proposed method to be effectively applied even at global scale.

With more than half of the world population now living in cities, and 2.5 billion more people expected to move into cities by 2050, urban areas pose significant challenges on the environment. Although only a small percentage of global land cover, urban areas significantly alter climate, biogeochemistry, and hydrology at local, regional, and global scales. Thus, accurate and timely information on urban land cover and their changing patterns is of critical importance to support sustainable urban development.

EO4Urban is a new project within the ESA DUE INNOVATOR III program. The overall objective of this research is to evaluate multi-temporal Sentinel-1A SAR and Sentinel-2A MSI data for global urban services using innovative methods and algorithms, including KTH-Pavia Urban Extractor, a robust algorithm for urban extent extraction and KTH-SEG, a novel object-based classification method for detailed urban land cover mapping. Ten cities around the world in different geographical and environmental conditions are selected as study areas. In the first phase of the project, multitemporal Sentinel-1A SAR and Sentinel-2A MSI data as well as SPOT Take 5 data over Beijing, Stockholm and Lagos were acquired during 2015 vegetation season.

The preliminary urban extraction results showed that urban areas and small towns could be well extracted using multitemporal Sentinel-1A SAR data with the KTH-Pavia urban extractor. The results are being further improved using multitemporal Sentinel-1A SAR and a single-date Sentinel-2A MSI or SPOT-5 data.  The preliminary results also show that the fusion of multitemporal Sentinel-1A SAR and Sentinel-2A MSI data could produce detailed land cover maps with improved accuracy than that of SAR or optical data alone.  Compared to the urban extraction results from ENVISAT ASAR or ERS SAR data in 2005, the urbanization patterns and trends in Beijing, Stockholm and Lagos will also be presented at the Living Planet Symposium. 

This research and development is expected to produce a pilot global urban services demonstrator using multitemporal Sentinel-1A SAR and Sentinel-2A MSI data. The project will contribute to i). better understanding of the urban products and services that the end users require; ii). development of novel and robust methods and algorithms for improved urban services to planners to support smart and sustainable urban development; ; iii). better understanding of the capacity of Sentinel-1A SAR and Sentinel-2A optical data for detailed urban land cover mapping and urbanisation monitoring; iv). the goals and activities of GEO SB-04 Global Urban Observation and Information Task and GEO SB-02 Global Land Cover Task.