Convolutional neural networks for global human settlements mapping from Sentinel-2 satellite imagery

Enhancing surface drainage mapping in eastern Canada with deep learning applied to LiDAR-derived elevation data

Agricultural dykelands in Nova Scotia rely heavily on a surface drainage technique called land forming, which is used to alter the topography of fields to improve drainage. The presence of land-formed fields provides useful information to better understand land utilization on these lands vulnerable to rising sea levels. Current field boundaries delineation and classification methods, such as manual digitalization and traditional segmentation techniques, are labour-intensive and often require manual and time-consuming parameter selection. In recent years, deep learning (DL) techniques, including convolutional neural networks and Mask R-CNN, have shown promising results in object recognition, image classification, and segmentation tasks. However, there is a gap in applying these techniques to detecting surface drainage patterns on agricultural fields. This paper develops and tests a Mask R-CNN model for detecting land-formed fields on agricultural dykelands using LiDAR-derived elevation data. Specifically, our approach focuses on identifying groups of pixels as cohesive objects within the imagery, a method that represents a significant advancement over pixel-by-pixel classification techniques. The DL model developed in this study demonstrated a strong overall performance, with a mean Average Precision (mAP) of 0.89 across Intersection over Union (IoU) thresholds from 0.5 to 0.95, indicating its effectiveness in detecting land-formed fields. Results also revealed that 53% of Nova Scotia's dykelands are being used for agricultural purposes and approximately 75% (6924 hectares) of these fields were land-formed. By applying deep learning techniques to LiDAR-derived elevation data, this study offers novel insights into surface drainage mapping, enhancing the capability for precise and efficient agricultural land management in regions vulnerable to environmental changes.

Landsat-derived impervious surface area expansion in the Arctic from 1985 to 2021

Accurate and timely impervious surface mapping is essential for assessing land cover change, urban heat island, and monitoring human activity intensity and ecological change. While various global impervious surface datasets become available, these datasets exhibit significant omissions in Arctic regions. Hence, in this study, we present a 30-m impervious surface area (ISA) dataset of Arctic from 1985 to 2021 (GISA_Arcitc). To this aim, we proposed to combine visually interpreted ISA samples and automatically generated NonISA samples for Arctic ISA mapping. Then, adaptive random forest (RF) classifiers were used for long time-series ISA mapping and the result was post-processed to improve the spatial-temporal consistency. Finally, the accuracy of GISA_Arcitc was assessed using the 37,800 independent test samples. GISA_Arctic possessed an overall accuracy of 93.59 % and a F-score of 0.934. It is found that the Arctic ISA increased from 857.83 km2 to 2115.49 km2 during the past 37 years. More than 84 % of the Arctic ISA increment is embraced by three countries: Russia, Finland, and Norway. Courtesy of the long time-series GISA_Arctic, the sources of Arctic ISA expansion were further analyzed. It was found that the top three land covers transformed to ISA are tundra, forest and grassland. The GISA_Arctic could contribute to further understanding of human activities and Arctic ecological changes, which can be accessed from http://irsip.whu.edu.cn/resources/resources_v2.php.

A 10-m resolution impervious surface area map for the greater Mekong subregion from remote sensing images

High-resolution and multi-temporal impervious surface area maps are crucial for capturing rapidly developing urbanization patterns. However, the currently available relevant maps for the greater Mekong subregion suffer from coarse resolution and low accuracy. Addressing this issue, our study focuses on the development of accurate impervious surface area maps at 10-m resolution for this region for the period 2016-2022. To accomplish this, we present a new machine-learning framework implemented on the Google Earth Engine platform that merges Sentinel-1 Synthetic Aperture Radar images and Sentinel-2 Multispectral images to extract impervious surfaces. Furthermore, we also introduce a training sample migration strategy that eliminates the collection of additional training samples and automates multi-temporal impervious surface area mapping. Finally, we perform a quantitative assessment with validation samples interpreted from Google Earth. Results show that the overall accuracy and kappa coefficient of the final impervious surface area maps range from 92.75% to 92.93% and 0.854 to 0.857, respectively. This dataset provides comprehensive measurements of impervious surface coverage and configuration that will help to inform urban studies.

Spatially explicit accuracy assessment of deep learning-based, fine-resolution built-up land data in the United States

Geospatial datasets derived from remote sensing data by means of machine learning methods are often based on probabilistic outputs of abstract nature, which are difficult to translate into interpretable measures. For example, the Global Human Settlement Layer GHS-BUILT-S2 product reports the probability of the presence of built-up areas in 2018 in a global 10 m × 10 m grid. However, practitioners typically require interpretable measures such as binary surfaces indicating the presence or absence of built-up areas or estimates of sub-pixel built-up surface fractions. Herein, we assess the relationship between the built-up probability in GHS-BUILT-S2 and reference built-up surface fractions derived from a highly reliable reference database for several regions in the United States. Furthermore, we identify a binarization threshold using an agreement maximization method that creates binary built-up land data from these built-up probabilities. These binary surfaces are input to a spatially explicit, scale-sensitive accuracy assessment which includes the use of a novel, visual-analytical tool which we call focal precision-recall signature plots. Our analysis reveals that a threshold of 0.5 applied to GHS-BUILT-S2 maximizes the agreement with binarized built-up land data derived from the reference built-up area fraction. We find high levels of accuracy (i.e., county-level F-1 scores of almost 0.8 on average) in the derived built-up areas, and consistently high accuracy along the rural-urban gradient in our study area. These results reveal considerable accuracy improvements in human settlement models based on Sentinel-2 data and deep learning, as compared to earlier, Landsat-based versions of the Global Human Settlement Layer.

Urban heat in Johannesburg and Ekurhuleni, South Africa: A meter-scale assessment and vulnerability analysis

Heat stress is an important threat for human health and urban areas are affected at higher rates compared to rural environments. Additionally, climate change will increase the vulnerability towards urban heat stress in the future. Current high-resolution urban heat stress assessments are limited in time and space due to the high computational costs. In this paper, the UrbClim numerical model is used to simulate urban heat accurately at a fast rate and high spatial resolution for the cities of Johannesburg and Ekurhuleni, South Africa. Using detailed terrain information, (future) urban heat stress assessments are provided at 30 m resolution for both city agglomerations, while meter-scale simulations are executed for a selection of neighborhoods. These model simulations are evaluated using an extensive monitoring campaign in which the local community was heavily engaged. Distinct spatial differences in the urban heat island effect are observed, with greatest heat stress in areas with high building densities and low vegetation numbers. These areas are often characterized by lower socio-economic living conditions. The meter-scale analysis further shows the importance of shade provided by vegetation to lower heat stress in both present and future climate. These assessments offer assistance in the design of climate-resilient urban planning strategies.

Sentinel2GlobalLULC: A Sentinel-2 RGB image tile dataset for global land use/cover mapping with deep learning

Land-Use and Land-Cover (LULC) mapping is relevant for many applications, from Earth system and climate modelling to territorial and urban planning. Global LULC products are continuously developing as remote sensing data and methods grow. However, there still exists low consistency among LULC products due to low accuracy in some regions and LULC types. Here, we introduce Sentinel2GlobalLULC, a Sentinel-2 RGB image dataset, built from the spatial-temporal consensus of up to 15 global LULC maps available in Google Earth Engine. Sentinel2GlobalLULC v2.1 contains 194877 single-class RGB image tiles organized into 29 LULC classes. Each image is a 224 × 224 pixels tile at 10 × 10 m resolution built as a cloud-free composite from Sentinel-2 images acquired between June 2015 and October 2020. Metadata includes a unique LULC annotation per image, together with level of consensus, reverse geo-referencing, global human modification index, and number of dates used in the composite. Sentinel2GlobalLULC is designed for training deep learning models aiming to build precise and robust global or regional LULC maps.

Assessing the relationship between morphology and mapping accuracy of built-up areas derived from global human settlement data

It is common knowledge that the level of landscape heterogeneity may affect the performance of remote sensing based land use / land cover classification. While this issue has been studied in depth for land cover data in general, the specific relationship between the mapping accuracy and morphological characteristics of built-up surfaces has not been analyzed in detail, an urgent need given the recent emergence of a variety of global, fine-resolution settlement datasets. Moreover, previous studies typically rely on aggregated, broad-scale landscape metrics to quantify the morphology of built-up areas, neglecting the fine-grained spatial variation and scale dependency of such metrics. Herein, we aim to fill this knowledge gap by assessing the associations between localized (focal) landscape metrics, derived from binary built-up surfaces and localized data accuracy estimates. We tested our approach for built-up surfaces from the Global Human Settlement Layer (GHSL) for Massachusetts (USA). Specifically, we examined the explanatory power of landscape metrics with respect to both commission and omission errors in the multi-temporal GHS-BUILT R2018A data product. We found that the Landscape Shape Index (LSI) calculated in focal windows exhibits, on average, the highest levels of correlation to focal accuracy measures. These relationships are scale-dependent, and become stronger with increasing level of spatial support. We found that thematic omission error, as measured by Recall, has the strongest relationship to measures of built-up surface morphology across different temporal epochs and spatial resolutions. The results of our regression analysis (R2>0.9), estimating accuracy based on landscape metrics, confirmed these findings. Lastly, we tested the generalizability of our findings by regionally stratifying our regression models and applying them to a different version of the GHSL (i.e., the GHS-BUILT-S2) and a different study area. We observed varying levels of model transferability, indicating that the relationship between accuracy and landscape metrics may be sensor-specific, and is heavily localized for most accuracy metrics, but quite generalizable for the Recall measure. This indicates that there is a strong and generalizable association between morphological properties of built-up land and the degree to which it is "undermapped".

Urban growth modelling and social vulnerability assessment for a hazardous Kathmandu Valley

In our rapidly urbanizing world, many hazard-prone regions face significant challenges regarding risk-informed urban development. This study addresses this issue by investigating evolving spatial interactions between natural hazards, ever-increasing urban areas, and social vulnerability in Kathmandu Valley, Nepal. The methodology considers: (1) the characterization of flood hazard and liquefaction susceptibility using pre-existing global models; (2) the simulation of future urban built-up areas using the cellular-automata SLEUTH model; and (3) the assessment of social vulnerability, using a composite index tailored for the case-study area. Results show that built-up areas in Kathmandu Valley will increase to 352 km² by 2050, effectively doubling the equivalent 2018 figure. The most socially vulnerable villages will account for 29% of built-up areas in 2050, 11% more than current levels. Built-up areas in the 100-year and 1000-year return period floodplains will respectively increase from 38 km² and 49 km² today to 83 km² and 108 km² in 2050. Additionally, built-up areas in liquefaction-susceptible zones will expand by 13 km² to 47 km². This study illustrates how, where, and to which extent risks from natural hazards can evolve in socially vulnerable regions. Ultimately, it emphasizes an urgent need to implement effective policy measures for reducing tomorrow's natural-hazard risks.

Urban Change Detection from Aerial Images Using Convolutional Neural Networks and Transfer Learning

Urban change detection is an important part of sustainable urban planning, regional development, and socio-economic analysis, especially in regions with limited access to economic and demographic statistical data. The goal of this research is to create a strategy that enables the extraction of indicators from large-scale orthoimages of different resolution with practically acceptable accuracy after a short training process. Remote sensing data can be used to detect changes in number of buildings, forest areas, and other landscape objects. In this paper, aerial images of a digital raster orthophoto map at scale 1:10,000 of the Republic of Lithuania (ORT10LT) of three periods (2009–2010, 2012–2013, 2015–2017) were analyzed. Because of the developing technologies, the quality of the images differs significantly and should be taken into account while preparing the dataset for training the semantic segmentation model DeepLabv3 with a ResNet50 backbone. In the data preparation step, normalization techniques were used to ensure stability of image quality and contrast. Focal loss for the training metric was selected to deal with the misbalanced dataset. The suggested model training process is based on the transfer learning technique and combines using a model with weights pretrained in ImageNet with learning on coarse and fine-tuning datasets. The coarse dataset consists of images with classes generated automatically from Open Street Map (OSM) data and the fine-tuning dataset was created by manually reviewing the images to ensure that the objects in images match the labels. To highlight the benefits of transfer learning, six different models were trained by combining different steps of the suggested model training process. It is demonstrated that using pretrained weights results in improved performance of the model and the best performance was demonstrated by the model which includes all three steps of the training process (pretrained weights, training on coarse and fine-tuning datasets). Finally, the results obtained with the created machine learning model enable the implementation of different approaches to detect, analyze, and interpret urban changes for policymakers and investors on different levels on a local map, grid, or municipality level.

A crowdsourced global data set for validating built-up surface layers

Several global high-resolution built-up surface products have emerged over the last five years, taking full advantage of open sources of satellite data such as Landsat and Sentinel. However, these data sets require validation that is independent of the producers of these products. To fill this gap, we designed a validation sample set of 50 K locations using a stratified sampling approach independent of any existing global built-up surface products. We launched a crowdsourcing campaign using Geo-Wiki ( https://www.geo-wiki.org/ ) to visually interpret this sample set for built-up surfaces using very high-resolution satellite images as a source of reference data for labelling the samples, with a minimum of five validations per sample location. Data were collected for 10 m sub-pixels in an 80 × 80 m grid to allow for geo-registration errors as well as the application of different validation modes including exact pixel matching to majority or percentage agreement. The data set presented in this paper is suitable for the validation and inter-comparison of multiple products of built-up areas.

Mapping Blue and Red Color-Coated Steel Sheet Roof Buildings over China Using Sentinel-2A/B MSIL2A Images

Accurate and efficiently updated information on color-coated steel sheet (CCSS) roof materials in urban areas is of great significance for understanding the potential impact, challenges, and issues of these materials on urban sustainable development, human health, and the environment. Thanks to the development of Earth observation technologies, remote sensing (RS) provides abundant data to identify and map CCSS materials with different colors in urban areas. However, existing studies are still quite challenging with regards to the data collection and processing costs, particularly in wide geographical areas. Combining free access high-resolution RS data and a cloud computing platform, i.e., Sentinel-2A/B data sets and Google Earth Engine (GEE), this study aims at CCSS material identification and mapping. Specifically, six novel spectral indexes that use Sentinel-2A/B MSIL2A data are proposed for blue and red CCSS material identification, namely the normalized difference blue building index (NDBBI), the normalized difference red building index NDRBI, the enhanced blue building index (EBBI), the enhanced red building index (ERBI), the logical blue building index (LBBI) and the logical red building index (LRBI). These indexes are qualitatively and quantitatively evaluated on a very large number of urban sites all over the P.R. China and compared with the state-of-the-art redness and blueness indexes (RI and BI, respectively). The results demonstrate that the proposed indexes, specifically the LRBI and LBBI, are highly effective in visual evaluation, clearly detecting and discriminating blue and red CCSS covers from other urban materials. Results show that urban areas from the northern parts of P.R. China have larger proportions of blue and red CCSS materials, and areas of blue and red CCSS material buildings are positively correlated with population and urban size at the provincial level across China.

High resolution global spatiotemporal assessment of rooftop solar photovoltaics potential for renewable electricity generation

Rooftop solar photovoltaics currently account for 40% of the global solar photovoltaics installed capacity and one-fourth of the total renewable capacity additions in 2018. Yet, only limited information is available on its global potential and associated costs at a high spatiotemporal resolution. Here, we present a high-resolution global assessment of rooftop solar photovoltaics potential using big data, machine learning and geospatial analysis. We analyse 130 million km2 of global land surface area to demarcate 0.2 million km2 of rooftop area, which together represent 27 PWh yr-1 of electricity generation potential for costs between 40-280 $ MWh-1. Out of this, 10 PWh yr-1 can be realised below 100 $ MWh-1. The global potential is predominantly spread between Asia (47%), North America (20%) and Europe (13%). The cost of attaining the potential is lowest in India (66 $ MWh-1) and China (68 $ MWh-1), with USA (238 $ MWh-1) and UK (251 $ MWh-1) representing some of the costliest countries.

Satellite-Based Human Settlement Datasets Inadequately Detect Refugee Settlements: A Critical Assessment at Thirty Refugee Settlements in Uganda

Satellite-based broad-scale (i.e., global and continental) human settlement data are essential for diverse applications spanning climate hazard mitigation, sustainable development monitoring, spatial epidemiology and demographic modeling. Many human settlement products report exceptional detection accuracies above 85%, but there is a substantial blind spot in that product validation typically focuses on large urban areas and excludes rural, small-scale settlements that are home to 3.4 billion people around the world. In this study, we make use of a data-rich sample of 30 refugee settlements in Uganda to assess the small-scale settlement detection by four human settlement products, namely, Geo-Referenced Infrastructure and Demographic Data for Development settlement extent data (GRID3-SE), Global Human Settlements Built-Up Sentinel-2 (GHS-BUILT-S2), High Resolution Settlement Layer (HRSL) and World Settlement Footprint (WSF). We measured each product’s areal coverage within refugee settlement boundaries, assessed detection of 317,416 building footprints and examined spatial agreement among products. For settlements established before 2016, products had low median probability of detection and F1-score of 0.26 and 0.24, respectively, a high median false alarm rate of 0.59 and tended to only agree in regions with the highest building density. Individually, GRID3-SE offered more than five-fold the coverage of other products, GHS-BUILT-S2 underestimated the building footprint area by a median 50% and HRSL slightly underestimated the footprint area by a median 7%, while WSF entirely overlooked 8 of the 30 study refugee settlements. The variable rates of coverage and detection partly result from GRID3-SE and HRSL being based on much higher resolution imagery, compared to GHS-BUILT-S2 and WSF. Earlier established settlements were generally better detected than recently established settlements, showing that the timing of satellite image acquisition with respect to refugee settlement establishment also influenced detection results. Nonetheless, settlements established in the 1960s and 1980s were inconsistently detected by settlement products. These findings show that human settlement products have far to go in capturing small-scale refugee settlements and would benefit from incorporating refugee settlements in training and validating human settlement detection approaches.

Analysis of Built-Up Areas of Small Polish Cities with the Use of Deep Learning and Geographically Weighted Regression

Small cities are an important part of the settlement system, a link between rural areas and large cities. Although they perform important functions, research focuses on large cities and metropolises while marginalizing small cities, the study of which is of great importance to progress in social sciences, geography, and urban planning. The main goal of this paper was to verify the impact of selected socio-economic factors on the share of built-up areas in 665 small Polish cities in 2019. Data from the Database of Topographic Objects (BDOT), Sentinel-2 satellite imagery from 2015 and 2019, and Local Data Bank by Statistics Poland form 2019 were used in the research. A machine learning segmentation procedure was used to obtain the data on the occurrence of built-up areas. Hot Spot (Getis-Ord Gi*) analysis and geographically weighted regression (GWR) was applied to explain spatially varying impact of factors related to population, spatial and economic development, and living standards on the share of built-up areas in the area of small cities. Significant association was found between the population density and the share of built-up areas in the area of the cities studied. The influence of the other socio-economic factors examined, related to the spatial and economic development of the cities and the quality of life of the inhabitants, showed great regional variation. The results also indicated that the share of built-up areas in the area of the cities under study is a result of the conditions under which they were established and developed throughout their existence, and not only of the socio-economic factors affecting them at present.