Urban Land Use Classification Research Articles

THE RATIONALITY OF URBAN FINANCIAL NETWORK LAYOUT BASED ON POI DATA

Abstract. Financial outlets are branches or business outlets of financial institutions such as banks, investment and finance companies and insurance companies in cities, which are used to provide a variety of financial services and businesses. The location of financial outlets is usually determined on the basis of factors such as population density, commercial areas, residential areas and transportation convenience. As an important carrier of urban financial services, the rationality of the layout of financial outlets can reflect the financial needs and behavioral patterns of people in the city. Urban functional area is a product after urbanization, different from urban land use type which focuses on the classification of urban land use through natural conditions, urban functional area mainly analyzes the internal spatial structure of the city, focusing on the discernment of the dominant functions within the built-up area of the city. By categorizing urban functions such as residence, commercial service and industry, the spatial distribution of urban functions can be grasped to provide support for urban planning and development. Urban functional areas can reveal human behavior patterns, needs and preferences in different areas. The current research on financial outlets mainly focuses on the accessibility and coverage of financial outlets, and researchers treat financial outlets as spatially undifferentiated points, ignoring the problem of spatial heterogeneity of financial outlets within different functional zones. When studying the rationality of the layout of urban financial outlets, considering different functional zones within the city can better meet the financial needs of residents, enterprises and institutions in the region and improve the accessibility and convenience of financial services. Analyzing the comprehensive service capacity of financial outlets within different functional areas is obviously insufficient in the current study, which is not conducive to the configuration planning of financial outlets and the promotion of the accessibility of urban financial services. The wide application of Point of Interest (POI) data in spatial analysis provides a new perspective for the study of urban financial outlets, for this reason, this paper analyzes and researches the current situation and rationality of the spatial layout of financial outlets based on POI data, under the perspective of the functional areas of the city, and combined with the population travel vitality index. The Term Frequency-Inverse Document Frequency (TF-IDF) algorithm is used to process the POI data to generate the urban functional area, on the basis of which the travel vitality index in the city is calculated by combining the population coverage and road network density, so as to judge the reasonableness of the layout of urban financial outlets. This helps us to better understand the behavioral decision-making process and social interaction of human beings about financial life in the city, and provides guidance and decision-making support for urban financial layout planning, financial public services and financial policies.

Global urban high-resolution land-use mapping: From benchmarks to multi-megacity applications

Timely and reliable land-use maps are of great importance in urban planning and environmental monitoring. With the rich spatial structure information, very high resolution (VHR) imagery is an important data source for identifying complex urban land use. However, the existing scene datasets and land-use mapping products based on VHR images have the following three problems: 1) accurate geographic boundaries of urban land parcels are lacking; 2) the category systems are inconsistent with the definitions in urban land use; and 3) it is difficult to achieve efficient and fully automated mapping in multiple cities. To tackle these problems, the GlobalUrbanNet-based automatic multi-city mapping and analysis (GAMMA) framework is proposed in this article. The GAMMA framework is made up of the GlobalUrbanNet (GUN) dataset, the multi-city fully automatic urban land-use mapping (AutoULUM) method, and the analysis of urban development patterns. Specifically, the large-scale 42-category fine-grained VHR urban land-use dataset—the GUN dataset—was constructed to deal with the above global urban land-use mapping problems, which contains 1,846,151 samples and 42 land-use categories covering six continents. The GUN dataset samples with land-use semantics and parcel boundaries were generated automatically based on the open-source area of interest (AOI) data from OpenStreetMap (OSM). In addition, the AutoULUM method is proposed to automate the process of OSM road network rectification and land parcel generation. On this basis, efficient and complete multi-city land-use maps can be produced using the GUN-pretrained scene classification models. To establish a benchmark for urban land-use classification, the representative urban land-use classification methods were evaluated on the GUN dataset. For further application, eight megacities from six continents were selected for automatic land-use mapping and analysis, i.e., Shanghai, Wuhan, and Chengdu in Asia, Helsinki in Europe, Nairobi in Africa, New York in North America, Rio de Janeiro in South America, and Sydney in Oceania. The results show that the models trained on the proposed GUN dataset have good generalizability in global urban areas, the AutoULUM method achieves efficient and fully automatic land-use mapping, and the GAMMA framework will help boost the coordinated development of multiple cities around the world.

Threshold effect of data amount and grid size on urban land use type identification using multi-source data fusion

Reliable urban land use maps are important for sustainable development and planning. Currently, the effects of different data source combinations and grid sizes on mapping results have rarely been studied. To reduce subjectivity in data selection, 10 collected multi-source spatial data were combined by traversal to create 1013 simulated combination schemes. Considering the size range of these data sources, 10 fusion grid sizes were selected. Then, a multi-source data learning model for urban land use classification (ULUC) was established by combining convolutional neural networks and long short-term memory. By taking Jinshui District (Zhengzhou, China) as an example, 10130 ULUC mappings were obtained. The maximum accuracy (82.9%) was achieved in the combination scheme D1D2D3D5D6D7D8D9D10 at a grid size of 30 m. The optimal solution among simulation 10130 schemes had an accuracy of 82.9%, a 14.7% improvement compared to the average accuracy of 67.6%. It is found that (1) The maximum accuracy showed a tendency to increase and then decrease with the increase in the variety of multi-source data combinations;(2) As the grid size decreases, the maximum accuracy also exhibited a tendency to increase and then decrease; (3) There was a significant threshold effect for both data combination types and grid sizes.

Estimating Urban Green Space Irrigation for 286 Cities in China: Implications for Urban Land Use and Water Management

Urban green space has increased significantly in many cities in China as a result of rapid urbanization. Despite various environmental and societal benefits of urban green space, its irrigation water use has become an important yet under-researched issue in urban water management. Using a newly developed green space evapotranspiration and soil water budgeting model coded in Python, this research estimated the irrigation water requirement of urban green space for 286 cities at the prefecture level and above in China, with daily meteorological data in 1986–2011, the distribution and areas of urban green space from the 30 m resolution “Essential Urban Land-Use Categories in China” database, and green space types from the 10 m resolution “10 m resolution global land cover” database. The estimated annual average irrigation water requirement of urban green space for these 286 cities was 8.7 billion m3, accounting for over 20% of residential water consumption in the cities where data on residential water use were available. We also investigated the spatiotemporal variation of urban green space irrigation water requirement and the factors that influence such variations. To the best of our knowledge, this study develops the first set of results of urban green space irrigation for all major cities in China, thus providing useful insights for green space irrigation and urban water man-agement in the context of rapid urbanization and sustainable water management.

Fungal-bacterial associations in urban allotment garden soils

The soil microbiome in urban agriculture has not received much attention to date despite its important role in soil functionality. In this work, we evaluated the composition and diversity of fungal and bacterial communities through DNA extraction and ITS/16S marker gene sequencing in 40 soil samples collected from 10 urban allotment garden areas in the city of Santiago de Compostela (northwestern Spain). Despite anthropogenic activities are expected to affect negatively the microbial diversity, the richness of both fungal and bacterial communities was comparable to that of soils from other urban land-use categories with lower anthropogenic influence, i.e., urban forests, urban grasslands, and urban agricultural fields. However, the deterministic effect of soil properties and geographical distances was almost negligible in the surveyed allotment gardens. Neutral community models confirmed that the fungal and bacterial communities followed a random distribution (pseudo-R2fungi|bacteria = 0.653 | 0.898) and that they were more random (tNSTfungi|bacteria = 0.54 | 0.74) compared to the other abovementioned land-use categories (tNSTfungi|bacteria = 0.46 | 0.67, p ≪ 0.001). Network analysis showed that in contrast to natural soils with lower anthropogenic influence, microbial associations formed very small modules; and frequently, microbial units remained unconnected. Taken together, our findings provide evidence that the soil microbial communities in the studied urban allotment gardens comprised a random assortment of microbes and their interactions, thereby supporting potential implications of anthropogenic activity for soil health and ultimately ecosystem functionality.

Urban land-use classification using machine learning classifiers: comparative evaluation and post-classification multi-feature fusion approach

ABSTRACT Accurate spatial-temporal mapping of urban land-use and land-cover (LULC) provides critical information for planning and management of urban environments. While several studies have investigated the significance of machine learning classifiers for urban land-use mapping, the determination of the optimal classifiers for the extraction of specific urban LULC classes in time and space is still a challenge especially for multitemporal and multisensor data sets. This study presents the results of urban LULC classification using decision tree-based classifiers comprising of gradient tree boosting (GTB), random forest (RF), in comparison with support vector machine (SVM) and multilayer perceptron neural networks (MLP-ANN). Using Landsat data from 1984 to 2020 at 5-year intervals for the Greater Gaborone Planning Area (GGPA) in Botswana, RF was the best classifier with overall average accuracy of 92.8%, MLP-ANN (91.2%), SVM (90.9%) and GTB (87.8%). To improve on the urban LULC mapping, the study presents a post-classification multiclass fusion of the best classifier results based on the principle of feature in-feature out (FEI-FEO) under mutual exclusivity boundary conditions. Through classifier ensemble, the FEI-FEO approach improved the overall LULC classification accuracy by more than 2% demonstrating the advantage of post-classification fusion in urban land-use mapping.

Exploring the Addition of Airborne Lidar-DEM and Derived TPI for Urban Land Cover and Land Use Classification and Mapping

The classification and mapping accuracy of urban land cover and land use has always been a critical topic and several auxiliary data have been used to improve the classification accuracy. However, to the best of our knowledge, there is limited knowledge of the addition of airborne Light Detection and Ranging (lidar)-Digital Elevation Model (DEM) and Topographic Position Index (TPI) for urban land cover and land use classification and mapping. The aim of this study was to explore the addition of airborne lidar-DEM and derived TPI to reflect data of Landsat Operational Land Imager (OLI) in improving the classification accuracy of urban land cover and land use map- ping. Specifically, this study explored the mapping accuracies of urban land cover and land use classifications derived using: 1) standalone Landsat OLI satellite data; 2) Landsat OLI with acquired airborne lidar-DEM ; 3) Landsat OLI with TPI ; and 4) Landsat OLI with airborne lidar-DEM and derived TPI. The results showed that the addition of airborne lidar-DEM and TPI yielded the best overall urban land cover and land use classification accuracy of about 88%. The findings in this study demonstrated that both lidar-DEM and TPI had a positive impact in improving urban land cover and land use classification.

MANAGING, ASSESSING AND MONITORING URBAN SPRAWL USING REMOTE SENSING; CHANGE DETECTION, URBAN EXPANSION INTENSITY INDEX (UEII), AND SHANNON ENTROPY: A CASE STUDY OF AMMAN CITY, JORDAN (1990–2022)

Mapping and monitoring of land use land cover (LULC) changes in Amman city are important for sustainable regional development, urban planning, and environmental and land use management. To comprehend the temporal and the spatial variability of LULC change, six time-period satellite images of the study area have been examined based on remote sensing (RS) and geographic information system (GIS) techniques. The changes in LULC patterns of Amman city were monitored for the periods 1990-1995, 1995-1999, 1999-2004, 2004-2013, 2013-2017 and 2017-2022. Employing Spatio-temporal Landsat TM and ETM imagery acquired from 1990 to 2022 to comprehend, detect, and quantify the spatial pattern of urban sprawl. Supervised classification was used to extract built-up areas and measure changes in urban land-use class across the metropolis. The result of the urban expansion intensity index shows that urban expansion hits its peak between 2013 and 2017, with a 1.75 index resulting in high-speed development between the time frames. Shannon’s entropy approach was quantified to detect the urban sprawl phenomenon. The results showed that the Relative Shannon entropy values are above 0.5, indicating dispersedly spread-urban sprawl. Land Absorption Coefficient (LAC) and Land Consumption Ratio (LCR) were also used. A strong positive correlation between road networks and urban sprawl (0.7) indicates that urbanization plays a major role in the occurrence of urban sprawl. The findings obtained from this work would provide urban planners and decision-makers in Amman the tools and data they need, to implement planning interventions for sustainable future development and to mitigate the impacts of urban sprawl on different aspects on the environment.

Using Multiple Sources of Data and “Voting Mechanisms” for Urban Land-Use Mapping

High-quality urban land-use maps are essential for grasping the dynamics and scale of urban land use, predicting future environmental trends and changes, and allocating national land resources. This paper proposes a multisample “voting mechanism” based on multisource data and random forests to achieve fine mapping of urban land use. First, Zhengzhou City was selected as the study area. Based on full integration of multisource features, random forests were used to perform the preliminary classification of multiple samples. Finally, the preliminary classification results were filtered according to the “voting mechanism” to achieve high-precision urban land-use classification mapping. The results showed that the overall classification accuracy of Level I features increased by 5.66% and 14.32% and that the overall classification accuracy of Level II features increased by 9.02% and 12.46%, respectively, compared with the classification results of other strategies. Therefore, this method can significantly reduce the influence of mixed distribution of land types and improve the accuracy of urban land-use classification at a fine scale.

Improving Typical Urban Land-Use Classification with Active-Passive Remote Sensing and Multi-Attention Modules Hybrid Network: A Case Study of Qibin District, Henan, China

The study of high-precision land-use classification is essential for the sustainable development of land resources. This study addresses the problem of classification errors in optical remote-sensing images under high surface humidity, cloud cover, and hazy weather. The synthetic aperture radar (SAR) images are sensitive to soil moisture, and the microwave can penetrate clouds, haze, and smoke. By using both the active and passive remote-sensing data, the Sentinel-1A SAR and Sentinel-2B multispectral (MS) images are combined synergistically. The full-band data combining the SAR + MS + spectral indexes is thus constructed. Based on the high dimensionality and heterogeneity of this data set, a new framework (MAM-HybridNet) based on two-dimensional (2D) and three-dimensional (3D) hybrid convolutional neural networks combined with multi-attention modules (MAMs) is proposed for improving the accuracy of land-use classification in cities with high surface humidity. In addition, the same training samples supported by All bands data (SAR + MS + spectral index) are selected and compared with k-Nearest Neighbors (KNN), support vector machine (SVM), 2D convolutional neural networks, 3D convolutional neural networks, and hybridSN classification models to verify the accuracy of the proposed classification model. The results show that (1) fusion classification based on Sentinel-2B MSI and Sentinel-1A SAR data produce an overall accuracy (OA) of 95.10%, a kappa coefficient (KC) of 0.93, and an average accuracy (AA) of 92.86%, which is better than the classification results using Sentinel-2B MSI and Sentinel-1A SAR images separately. (2) The classification accuracy improves upon adding the spectral index, and the OA, KC, and AA improve by 3.77%, 0.05, and 5.5%, respectively. (3) With the support of full-band data, the algorithm proposed herein produces better results than other classification algorithms, with an OA of 98.87%, a KC of 0.98, and an AA of 98.36%. These results indicate that the synergistic effect of active-passive remote-sensing data improves land-use classification. Additionally, the results verify the effectiveness of the proposed deep-learning classification model for land-use classification.

Using different training strategies for urban land-use classification based on convolutional neural networks

Urban land-use scene classification from high-resolution remote-sensing imagery at high quality and accuracy is of paramount interest for urban planning, government policy-making and urban change detection. In recent years, urban land-use classification has become an ongoing task in areas addressable primarily by remote sensing, and numerous deep learning algorithms have achieved high performance on this task. However, both dataset and methodology problems still exist in the current approaches. Previous studies have relied on limited data sources, resulting in saturated classification results, and they have difficulty achieving comprehensive classification results. The previous methods based on convolutional neural networks (CNNs) focused primarily on model architecture rather than on the hyperparameters. Therefore, to achieve more accurate classification results, in this study, we constructed a new large dataset for urban land-use scene classification. More than thirty thousand remote sensing scene images were collected to create a dataset with balanced class samples that includes both higher intra-class variations and smaller inter-class dissimilarities than do the previously available public datasets. Then, we analysed two possible strategies for exploiting the capabilities of three existing popular CNNs on our datasets: full training and fine tuning. For each strategy, three types of learning rate decay were applied: fixed, exponential and polynomial. The experimental results indicate that fine tuning tends to be the best-performing strategy, and using ResNet-V1-50 and polynomial learning rate decay achieves the best results for the urban land-use scene classification task.

Deep convolutional autoencoder for urban land use classification using mobile device data

Mobile phone data can provide insightful location-based information on the interactions between individuals and the urban environment, e.g. urban land-use types. Using mobile phone data, this study aims to recognize citywide land use classes in a highly accurate and timely manner, because land use classes are important for urban planning and urban management. How to extract key and discriminative features from mobile-phone-call time series is an essential problem for urban land-use classification. Considering the high dimensional characteristic of mobile-phone-call time series data, an improved land-use classification framework with deep convolutional autoencoder (DCAE) is proposed to automatically recognize land-use types in urban areas, which are geographical regions clustered based on the similarity of features in waveform and magnitude of mobile-phone-call time series. The method is validated using mobile phone data collected in Wuhan city, China. The experimental result shows that the key and discriminative features in waveform and magnitude can be extracted without human prior knowledge and these features extracted by DCAE can improve the performance to distinguish functionally similar land-use types, such as business and industrial/commercial types.

Fine-Scale Classification of Urban Land Use and Land Cover with PlanetScope Imagery and Machine Learning Strategies in the City of Cape Town, South Africa

Urban land use and land cover (LULC) change can be efficiently monitored with high-resolution satellite products for a variety of purposes, including sustainable planning. These, together with machine learning strategies, have great potential to detect even subtle changes with satisfactory accuracy. In this study, we used PlaneScope Imagery and machine learning strategies (Random Forests, Support Vector Machines, Naïve Bayes and K-Nearest Neighbour) to classify and detect LULC changes over the City of Cape Town between 2016 and 2021. Our results showed that K-Nearest Neighbour outperformed other classifiers by achieving the highest overall classification of accuracy (96.54% with 0.95 kappa), followed by Random Forests (94.8% with 0.92 kappa), Naïve Bayes (93.71% with 0.91 kappa) and Support Vector Machines classifiers with relatively low accuracy values (92.28% with 0.88 kappa). However, the performance of all classifiers was acceptable, exceeding the overall accuracy of more than 90%. Furthermore, the results of change detection from 2016 to 2021 showed that the high-resolution PlanetScope imagery could be used to track changes in LULC over a desired period accurately.

COMPARISON OF MACHINE LEARNING CLASSIFIERS FOR MULTITEMPORAL AND MULTISENSOR MAPPING OF URBAN LULC FEATURES

Abstract. This study compares four machine-learning algorithms comprising of Classification And Regression Trees (CART), Random Forest (RF), Gradient Tree Boosting (GTB) and Support Vector Machine (SVM) for the classification of urban land-use and land-cover (LULC) features. Using multitemporal and multisensor Landsat data from 1984-2020 at 5-year intervals for the Greater Gaborone Planning Area (GGPA) in Botswana, the aim of the study is to determine the performance of the classifiers in the extraction of different urban LULC features as built-up, bare-soil, water, grass, shrubs and forest. The results show that for mapping built-up areas, RF and SVM presented the best results with overall accuracy of 85%. Bare soil is best mapped using RF and CART with accuracy of up to 98%, while SVM and GTB were most suitable for mapping water bodies. The suitable classifiers for mapping the vegetation classes were RF for grass (94.5%), SVM for shrubland (81.5%) and GTB for forest (84.3%). In terms of class specific accuracy, RF achieved the highest performance with average overall accuracy (OA) of 95.9%, SVM (95.8%), GTB (95.6%) and CART (95.1%). The same performance pattern was observed from the F1-score, True Positive Rate (TPR), False Positive Rate (FPR) and Area under ROC curve (AUC) metrices for the class classification accuracies. The overall accuracy for the eight-epoch years were RF (87.8%), SVM (87.5%), GTB (86.4%) and CART (85.3%). To improve on the urban LULC mapping, the study proposes the post-classification feature fusion of the best classifier results.

Fusing Earth observation and socioeconomic data to increase the transferability of large-scale urban land use classification

Monitoring and understanding urban development requires up-to-date information on multiple urban land-use classes. Manual classification and deep learning approaches based on very-high resolution imagery have been applied successfully, but the required resources limits their capacity to map urban land use at larger scales. Here, we use a combination of open-source satellite imagery, constituting of data from Sentinel-1 and Sentinel-2, and socioeconomic data, constituting of points-of-interest and spatial metrics from road networks to classify urban land-use at a national scale, using a deep learning approach. A related challenge for large-scale mapping is the availability of ground truth data. Therefore, we focus our analysis on the transferability of our classification approach, using ground truth labels from a nationwide land-use dataset for the Netherlands. By dividing the country into four regions, we tested whether a combination of satellite data and socioeconomic data increases the transferability of the classification approach, compared to using satellite data only. The results indicate that socioeconomic data increases the overall accuracy of the classification for the Netherlands by 3 percentage points. In a transfer learning approach we find that adding socioeconomic data increases the accuracy between 3 and 5 percentage points when trained on three regions and tested on the independent fourth one. In the case of training and testing on one region and testing on another, the increase in overall accuracy increased up to 9 percentage points. In addition, we find that our deep learning approach consistently outperforms a random forest model, used here as benchmark, in all of the abovementioned experiments. Overall, we find that socioeconomic data increases the accuracy of urban land use classification, but variations between experiments are large.

Identification of Urban Functional Areas Based on the Multimodal Deep Learning Fusion of High-Resolution Remote Sensing Images and Social Perception Data

As the basic spatial unit of urban planning and management, it is necessary to know the distribution status of urban functional areas in time. Due to the complexity of urban land use, it is difficult to identify the urban functional areas using only remote sensing images. Social perception data can provide additional information for the identification of urban functional areas. However, the sources of remote sensing data and social perception data differ, with some differences in data forms. Existing methods cannot comprehensively consider the characteristics of these data for functional area identification. Therefore, in this study, we propose a multimodal deep learning method with an attention mechanism to fully utilize the data features of these two modalities and apply it to the recognition of urban functional areas. First, the pre-processed remote sensing images, points of interest, and building footprint data are divided into block-based target units of features by the road network. Next, the remote sensing image features and social perception data features of the target unit are extracted separately using a two-branch convolutional network. Finally, features are extracted sequentially along two separate dimensions, being channel and spatial, to generate an attention weight map for the identification and classification mapping of urban functional areas. The model framework was finally applied to the Ningbo dataset for testing, and the recognition accuracy was above 93%. The experimental results deduce, as a whole, that the prediction performance of the deep multimodal fusion model framework with an attention mechanism is comparatively superior to other traditional methods. It can provide a reference for the classification of urban land use and provide data support for urban planning and management.

A framework for urban land use classification by integrating the spatial context of points of interest and graph convolutional neural network method

Land-use classification plays an important role in urban planning and resource allocation and had contributed to a wide range of urban studies and investigations. With the development of crowdsourcing technology and map services, points of interest (POIs) have been widely used for recognizing urban land-use types. However, current research methods for land-use classifications have been limited to extracting the spatial relationship of POIs in research units. To close this gap, this study uses a graph-based data structure to describe the POIs in research units, with graph convolutional networks (GCNs) being introduced to extract the spatial context and urban land-use classification. First, urban scenes are built by considering the spatial context of POIs. Second, a graph structure is used to express the scenes, where POIs are treated as graph nodes. The spatial distribution relationship of POIs is considered to be the graph's edges. Third, a GCN model is designed to extract the spatial context of the scene by aggregating the information of adjacent nodes within the graph and urban land-use classification. Thus, the land-use classification can be treated as a classification on a graphic level through deep learning. Moreover, the POI spatial context can be effectively extracted during classification. Experimental results and comparative experiments confirm the effectiveness of the proposed method.

A Combined Convolutional Neural Network for Urban Land-Use Classification with GIS Data

The classification of urban land-use information has become the underlying database for a variety of applications including urban planning and administration. The lack of datasets and changeable semantics of land-use make deep learning methods suffer from low precision, which prevent improvements in the effectiveness of using AI methods for applications. In this paper, we first used GIS data to produce a well-tagged and high-resolution urban land-use image dataset. Then, we proposed a combined convolutional neural network named DUA-Net for complex and diverse urban land-use classification. The DUA-Net combined U-Net and Densely connected Atrous Spatial Pyramid Pooling (DenseASPP) to extract Remote Sensing Imagers (RSIs) features in parallel. Then, channel attention was used to efficiently fuse the multi-source semantic information from the output of the double-layer network to learn the association between different land-use types. Finally, land-use classification of high-resolution urban RSIs was achieved. Experiments were performed on the dataset of this paper, the publicly available Vaihingen dataset and Potsdam dataset with overall accuracy levels reaching 75.90%, 89.71% and 89.91%, respectively. The results indicated that the complex land-use types with heterogeneous features were more difficult to extract than the single-feature land-cover types. The proposed DUA-Net method proved suitable for high-precision urban land-use classification, which will be of great value for urban planning and national land resource surveying.

Effectiveness of Semi-Supervised Learning and Multi-Source Data in Detailed Urban Landuse Mapping with a Few Labeled Samples

Detailed urban landuse information plays a fundamental role in smart city management. A sufficient sample size has been identified as a very crucial pre-request in machine learning algorithms for urban landuse classification. However, it is often difficult to recognize and label landuse categories from remote sensing images alone. Alternatively, field investigation is time-consuming with a high demand in human resources and monetary cost. Therefore, previous studies on urban landuse classification have often relied on a small size of labeled samples with very uneven spatial distribution. This study aims to explore the effectiveness of a semi-supervised classification framework with multi-source data for detailed urban landuse classification with a few labeled samples. A disagreement-based semi-supervised learning approach, the Co-Forest, was employed and compared with traditional supervised methods (e.g., random forest and XGBoost). Multi-source geospatial data were utilized including optical and nighttime light remote sensing and geospatial big data, which present the physical and socio-economic features of landuse categories. Taking urban landuse classification in Shenzhen City as a case, results show that the classification accuracy of the semi-supervised method are generally on par with that of traditional supervised methods, and less labeled samples are needed to achieve a comparable result under different training set ratios. Given a small sample size, the accuracy tends to be stable with training samples no less than 5% in total. Our results also indicate that the classification accuracy by using multi-source data is significantly higher than that with any single data source being applied. Among these data, map POI and high-resolution optical remote sensing data make larger contributions on the classification, followed by mobile data and nighttime light remote sensing data.

Determination of Heavy Metal Contamination and Pollution Indices of Roadside Dust in Dhaka City, Bangladesh

Urban roadside dust samples from Dhaka City in Bangladesh were collected from a planned residential area (PRA), spontaneous residential area (SRA), commercial area (CA), and urban green area (UGA) in winter and summer to study how season and different urban land-use categories influence the concentrations of heavy metals (Cr, Mn, Co, Ni, Cu, Zn, As, and Pb) and different pollution indices. The dust samples were fractionated into <32 μm particles, extracted by acid digestion followed by estimation of heavy metals, using ICP-MS. Pollution indices were calculated from the metal concentrations, using standard protocols. The concentrations of heavy metals in roadside dust varied significantly (all p < 0.05), due to sampling seasons and the land-use category. Higher concentrations of heavy metals (Cr, Mn, Ni, Cu, Zn, and Pb) were found in the dust sampled during the winter season than in the summer season, except for As and Co. The geo-accumulation index (Igeo) indicated that the commercial area was heavily contaminated with Cu and Zn during the winter season. The contamination factor (CF) was higher for Cu and Zn in the CA, PRA, and SRA of Dhaka City in winter than in the summer season. The enrichment factor (EF) suggested that Mn and Co were the least enriched metals, and significant enrichment was seen for Cu and Zn for all land-use categories, both in summer and winter. A moderate potential ecological risk for Cu was estimated in CA and PRA in the winter season.