Urban Land Cover Classification Research Articles

Machine learning based urban land cover classification using PolInSAR data: a study with ALOS-2 and RADARSAT-2 datasets

A substantial variation in the land cover dynamics has been observed as a consequence of increasing urban expansion. Polarimetric synthetic aperture radar (PolSAR) data is widely being used for land cover studies in urban areas due to its all-weather, day-and-night imaging capabilities. However, in densely built-up areas, challenge arises with buildings having large Azimuth Orientation Angles (AOAs). These buildings are often misclassified as vegetation due to the depolarization of radar signal causing volumetric scattering response from the structures. This study addresses this issue by proposing an approach that integrates polarimetric information with interferometric SAR (InSAR) coherence to improve the differentiation between urban structures and vegetated areas, enhancing the accuracy of urban land-cover classification. Vegetated areas exhibit lower temporal coherence due to changes in the orientation of its leaves and branches caused by wind, seasonal variations, growth phenology, and other factors. In contrast, urban structures, being relatively stable targets, maintain high temporal coherence values. In present research various decomposition and scattering parameters were evaluated, along with PolInSAR coherence derived from L-band (ALOS-2) and C-band (RADARSAT-2), using two machine learning algorithms namely, Random Forest (RF) and Convolutional Neural Network (CNN). The C-band RADARSAT-2 data, particularly with six-component decomposition parameters, performed better, achieving an overall accuracy as 85.85% using RF algorithm. To further improve classification results, optical datasets from Landsat constellation were fused with SAR parameters using Gram-Schmidt fusion technique. This fusion led to significant improvements, achieving an overall accuracy of 94.50% and kappa statistics of 0.92, when CNN algorithm was applied to the fused optical and C-band RADARSAT-2 dataset. These results demonstrate the effectiveness of combining PolInSAR and optical data for more accurate urban land-cover classification, particularly in complex urban environments.

Using cloud computing techniques to map the geographic extent of informal settlements in the greater Cape Town Metropolitan Area

Although remote sensing approaches offer unprecedented opportunities to understand urban land cover dynamics including informal settlements areal extent, challenges such as spectral confusions still persist, particularly when segregating land cover types like informal settlements from planned formal settlements. The improvements in Earth Observation (EO) data analytic tools such as introduction of Google Earth Engine (GEE) cloud computing platform, provide prospects to improve separability of these settlements from other urban land cover classes, via their advanced data processing and filtering algorithm, which allows for the synergic use of multisource and multi-temporal data, thus improving detection and monitoring of these settlements. This study harnessed the advance data analytic powers of GEE cloud computing platform coupled with higher resolution Sentinel-2 data to map the geographical extent of informal settlement in the Cape Town Metropolitan Area. The classification yielded six land cover classes: formal settlements, informal settlements, water, bare or built-up areas, vegetated lands, and croplands. Built-up formal settlement was the most dominant class, accounting for 70% of the total Cape Town surface area, while open water was the least dominant, accounting for 2%. Informal settlements accounted for approximately 7% of all settlements. Although overall accuracy was within acceptable limits (68%), some classes, such as vegetated lands and formal settlements, reported low class accuracies due to spectral similarities with other classes. The findings highlight the importance of the GEE platform, as well as the interaction of contextual and spectral characteristics, as well as various sentinel-2 derived water, built up, and vegetation indices in mapping informal settlements. These findings are critical for the facilitation of improved urban planning, provision of services and assisting in alleviating social as well as environmental issues within the Cape Town Metropolitan area.

An advanced classification method for urban land cover classification

This manuscript presents a detailed comparative analysis of three advanced classification techniques that were used between 2018 and 2020 to classify land cover using Landsat8 imagery, namely Support Vector Machine (SVM), Maximum Likelihood Classification (MLSC), and Random Forests (RF). The study focuses on evaluating the accuracy of these methods by comparing the classified maps with a higher-resolution ground truth map, utilising 500 randomly selected points for assessment. The obtained results show that, compared to MLSC and RT, the Support Vector Machine (SVM) approach performs better. The SVM model demonstrates enhanced precision in land cover classification, showcasing its effectiveness in discerning subtle differences in landscape features. Furthermore, using the precise classification results produced by the SVM method, this study examines the temporal variations in land cover between 2018 and 2020. The results provide insight into dynamic land cover changes and highlight the significance of applying reliable classification techniques for thorough temporal analysis with Landsat8 images.

Comparison of machine learning and parametric methods for the discrimination of urban land cover types

The aim of this study is to compare the performances of different machine learning and parametric techniques for differentiating highly mixed urban land cover classes in Ulaanbaatar, the capital city of Mongolia, using multisource data sets. For data sources, 17 features are chosen, including the original 10 spectral bands of the Sentinel-2 data; VV, VH, average of HH & HV, and simple ratio of Sentinel-1 data; and normalized difference vegetation index (NDVI), Bare soil index (BSI), and modified soil adjusted vegetation index (MSAVI). Six different feature combinations are used to identify available urban land cover classes. To discriminate existing classes, a support vector machine (SVM), artificial neural network (ANN), random forest (RF), and a statistical maximum likelihood classifier (MLC) are employed and compared. In all six feature combinations, the RF method outperforms the others with an overall accuracy ranging from 81.72% to 90.71%. The SVM has an overall accuracy ranging from 77.77%-83.50%, with the second-highest performance in four combinations and the lowest in two. The ANN has an overall accuracy ranging from 74.83%-83.31%, with poorer performance than the SVM. The MLC's performance varies across feature combinations, with an overall accuracy ranging from 70.96%-85.92%, and the second-highest performance in two feature combinations. Overall, the study shows that multisource information along with additional features/indices can significantly improve the classification of mixed urban land cover types, and for the given test site, the RF technique is the best option for producing a dependable land cover map.

An Urban Land Cover Classification Method Based on Segments’ Multidimension Feature Fusion

An Urban Land Cover Classification Method Based on Segments’ Multidimension Feature Fusion

Contrastive Learning for Urban Land Cover Classification With Multimodal Siamese Network

Contrastive Learning for Urban Land Cover Classification With Multimodal Siamese Network

A SHALLOW NEURAL NETWORK MODEL FOR URBAN LAND COVER CLASSIFICATION USING VHR SATELLITE IMAGE FEATURES

Abstract. Recently, image classification techniques using neural networks have received considerable attention in sustainable urban development, since their applications have an extreme effect on building distribution, infrastructural networks, and water resource management. In this research, a back-propagation shallow neural network model is presented for very high resolution satellite image classification in urban environments. Workflow procedures consider selecting and collecting data, preparing required study areas, extracting distinctive features, and applying the classification process. Visual interpretation is performed to identify observed land cover classes and detect distinctive features in the urban environment. Pre-processing techniques are implemented to present the used images in a more suited form for the classification techniques. A shallow neural network model (supported by MathWorks MATLAB environment) is successfully applied and results are evaluated. The proposed model is tested for classifying both WorldView-2 and WorldView-3 multispectral images with different spatial and spectral characteristics to check the model’s applicability to various kinds of satellite imagery and different study areas. Model outcomes are compared to two well-known classification methods; the Nearest Neighbour object-based method and the Maximum Likelihood pixel-based classifier, to validate and check the model stability. The overall accuracy achieved by the proposed model is 86.25% and 83.25%, while the nearest neighbour approach has obtained 84.50% and 82.75%, and the maximum likelihood classifier has accomplished 82.50% and 80.25% for study area 1 and study area 2 respectively. Obtained results indicate that the developed shallow neural network model achieves a promising accuracy for urban land cover classification in comparison with the standard techniques.

Fusion of sentinel-1 SAR and sentinel-2 MSI data for accurate Urban land use-land cover classification in Gondar City, Ethiopia

Effective urban planning and management rely on accurate land cover mapping, which can be achieved through the combination of remote sensing data and machine learning algorithms. This study aimed to explore and demonstrate the potential benefits of integrating Sentinel-1 SAR and Sentinel-2 MSI satellite imagery for urban land cover classification in Gondar city, Ethiopia. Synthetic Aperture Radar (SAR) data from Sentinel-1A and Multispectral Instrument (MSI) data from Sentinel-2B for the year 2023 were utilized for this research work. Support Vector Machine (SVM) and Random Forest (RF) machine learning algorithms were utilized for the classification process. Google Earth Engine (GEE) was used for the processing, classification, and validation of the remote sensing data. The findings of the research provided valuable insights into the performance evaluation of the Support Vector Machine (SVM) and Random Forest (RF) algorithms for image classification using different datasets, namely Sentinel 2B Multispectral Instrument (MSI) and Sentinel 1A Synthetic Aperture Radar (SAR) data. When applied to the Sentinel 2B MSI dataset, both SVM and RF achieved an overall accuracy (OA) of 0.69, with a moderate level of agreement indicated by the Kappa score of 0.357. For the Sentinel 1A SAR data, SVM maintained the same OA of 0.69 but showed an improved Kappa score of 0.67, indicating its suitability for SAR image classification. In contrast, RF achieved a slightly lower OA of 0.66 with Sentinel 1A SAR data. However, when the datasets of Sentinel 2B MSI and Sentinel 1A SAR were combined, SVM achieved an impressive OA of 0.91 with a high Kappa score of 0.80, while RF achieved an OA of 0.81 with a Kappa score of 0.809. These findings highlight the potential of fusing satellite data from multiple sources to enhance the accuracy and effectiveness of image classification algorithms, making them valuable tools for various applications, including land use mapping and environmental monitoring.

Urban land use land cover classification based on GF-6 satellite imagery and multi-feature optimization

Urban land use/land cover (LULC) classification has long been a hotspot for remote sensing applications. With high spatio-temporal resolution and multispectral, the recently launched GF-6 satellite provides ideal open imagery for LULC mapping. In this study, we utilized multitemporal GF-6 images to generate six types of land features, including spectral bands, texture features, built-up, waterbody, vegetation, and red-edge indices. The minimum Redundancy Maximum Relevance (mRMR) algorithm was employed to optimize feature selection. Subsequently, Random Forest (RF) and Extreme Gradient Boosting (XGBT) were assessed using different feature selections. Besides, various feature configurations were designed for LULC classification and comparison. The results indicate that the mRMR-based RF method achieved the highest overall accuracy of 91.37%. The temporal red-edge indices were important features for urban LULC classification and contributed mainly to grassland and cropland. These results supplement existing classification methods and assist in improving LULC mapping in urban areas with complex landscapes.

Multitemporal Feature-Level Fusion on Hyperspectral and LiDAR Data in the Urban Environment

Technological innovations and advanced multidisciplinary research increase the demand for multisensor data fusion in Earth observations. Such fusion has great potential, especially in the remote sensing field. One sensor is often insufficient in analyzing urban environments to obtain comprehensive results. Inspired by the capabilities of hyperspectral and Light Detection and Ranging (LiDAR) data in multisensor data fusion at the feature level, we present a novel approach to the multitemporal analysis of urban land cover in a case study in Høvik, Norway. Our generic workflow is based on bitemporal datasets; however, it is designed to include datasets from other years. Our framework extracts representative endmembers in an unsupervised way, retrieves abundance maps fed into segmentation algorithms, and detects the main urban land cover classes by implementing 2D ResU-Net for segmentation without parameter regularizations and with effective optimization. Such segmentation optimization is based on updating initial features and providing them for a second iteration of segmentation. We compared segmentation optimization models with and without data augmentation, achieving up to 11% better accuracy after segmentation optimization. In addition, a stable spectral library is automatically generated for each land cover class, allowing local database extension. The main product of the multitemporal analysis is a map update, effectively detecting detailed changes in land cover classes.

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.

A Subpixel Mapping Method for Urban Land Use by Reducing Shadow Effects

Urban land use classification is significant for urban development planning. Considering complex environments of urban surface features, traditional semantic segmentation methods are difficult to solve the problems of mixed pixels and limited spatial resolution of images. The subpixel mapping technology is an effective method to solve the above problems in urban land use classification. However, traditional subpixel mapping methods are sensitive to mountain shadow, high-rise building shadow and impermeable surface heterogeneity, resulting in false classification. Therefore, we propose a subpixel mapping method that can reduce the shadow effect. This method uses a multi-index feature fusion strategy to optimize the abundance of the shadow errors in the abundance image, and uses a super-resolution reconstruction neural network model to reconstruct the optimized abundance image for the subpixel mapping of urban land use. Experiments were conducted on sentinel-2 images obtained over Yuelu District of Changsha City, Hunan Province, China. The experimental results show that the method proposed in this article can effectively overcome the influence of building shadows and mountain shadows in urban land cover classification and is superior to traditional subpixel/pixel spatial attraction model, radial basis function, super-resolution subpixel mapping, and other methods in the effect and accuracy of urban land use subpixel mapping.

A Framework for Improving Urban Land Cover Using Object and Pixel-Based Techniques via Remotely Sensed Data

Recently, the advancement of remote sensing technology played a key role in urban land/cover mapping, planning, tourism, and environmental management. Images with a high spatial resolution for urban classification are widely used. Despite the high spectral resolution of the image, spectral confusion happens among different land covers. Furthermore, the shadow problem also causes poor results in the classification based on traditional per-pixel spectral approaches. This study looks at ways of improving the classification of urban land cover using QuickBird images. Maximum likelihood (ML) pixel-based supervised as well as Rule-based object-based approaches were examined on high-resolution QuickBird satellite images in Karbala City, Iraq. This study indicates that the use of textural attributes during the rule-based classification procedure can significantly improve land-use classification performance. Furthermore, the results show that rule-based results are highly effective in improving classification accuracy than pixel-based. The results of this study provide further clarity and insight into the implementation of using the object-based approach with various classifiers for the extended study. In addition, the finding demonstrated the integration of high-resolution QuickBird data and a set of attributes derived from the visible bands and geometric rule set resulted in superior class separability, thus higher classification accuracies in mapping complex urban environments.

A Possibility-Based Method for Urban Land Cover Classification Using Airborne Lidar Data

Airborne light detection and ranging (LiDAR) has been recognized as a reliable and accurate measurement tool in forest volume estimation, urban scene reconstruction and land cover classification, where LiDAR data provide crucial and efficient features such as intensity, elevation and coordinates. Due to the complex urban environment, it is difficult to classify land cover accurately and quickly from remotely sensed data. Methods based on the Dempster–Shafer evidence theory (DS theory) offer a possible solution to this problem. However, the inconsistency in the correspondence between classification features and land cover attributes constrains the improvement of classification accuracy. Under the original DS evidence theory classification framework, we propose a novel method for constructing a basic probability assignment (BPA) function based on possibility distributions and apply it to airborne LiDAR land cover classification. The proposed approach begins with a feature classification subset selected by single-feature classification results. Secondly, the possibility distribution of the four features was established, and the uncertainty relationship between feature values and land cover attributes was obtained. Then, we selected suitable interval cut-off points and constructed a BPA function. Finally, DS evidence theory was used for land cover classification. LiDAR and its co-registration data acquired by Toposys Falcon II were used in the performance tests of the proposed method. The experimental results revealed that it can significantly improve the classification accuracy compared to the basic DS method.

Towards sector-based attribution using intra-city variations in satellite-based emission ratios between CO2 and CO

Abstract. Carbon dioxide (CO2) and air pollutants such as carbon monoxide (CO) are co-emitted by many combustion sources. Previous efforts have combined satellite-based observations of multiple tracers to calculate their emission ratio (ER) for inferring combustion efficiency at the regional to city scale. Very few studies have focused on combustion efficiency at the sub-city scale or related it to emission sectors using space-based observations. Several factors are important for interpreting and deriving spatially resolved ERs from asynchronous satellite measurements, including (1) variations in meteorological conditions given the mismatch in satellite overpass times, (2) differences in vertical sensitivity of the retrievals (i.e., averaging kernel profiles), (3) interferences from the biosphere and biomass burning, and (4) the mismatch in the daytime variations of CO and CO2 emissions. In this study, we extended an established emission estimate approach to arrive at spatially resolved ERs based on retrieved column-averaged CO2 (XCO2) from the Snapshot Area Mapping (SAM) mode of the Orbiting Carbon Observatory-3 (OCO-3) and column-averaged CO from the TROPOspheric Monitoring Instrument (TROPOMI). To evaluate the influences of the confounding factors listed above and further attribute intra-urban variations in ERs to certain sectors, we leveraged a Lagrangian atmospheric transport model with an urban land cover classification dataset and reported ERCO values from the sounding level to the overpass and city level. We found that the differences in overpass times and averaging kernels between OCO and TROPOMI strongly affect the estimated spatially resolved ERCO. Specifically, a time difference of >3 h typically led to dramatic changes in wind directions and urban plume shapes, thereby making the calculation of accurate sounding-specific ERCO difficult. After removing such cases from consideration and applying a simple plume shift method when necessary to account for changes in wind direction and speed, we discovered significant contrasts in combustion efficiencies between (1) two megacities versus two industry-oriented cities and (2) different regions within a city, based on six nearly coincident overpasses per city. Results suggest that the ERCO impacted by heavy industry in Los Angeles is slightly lower than the overall city-wide value (<10 ppb-CO/ppm-CO2). In contrast, the ERCO related to heavy industry in Shanghai is much higher than Shanghai's city mean and more aligned with the city means of two selected industry-oriented cities in China (approaching 20 ppb-CO/ppm-CO2). Although investigations based on a larger number of satellite overpasses are needed, our unique approach (i.e., without using sector-specific information from emission inventories) provides new insights into assessing combustion efficiency within a city from future satellite missions, such as those that will map column CO2 and CO concentrations simultaneously with high spatiotemporal resolutions.

Urban land use and land cover classification with interpretable machine learning – A case study using Sentinel-2 and auxiliary data

The European Commission launch of the twin Sentinel-2 satellites provides new opportunities for land use and land cover (LULC) classification because of the ready availability of their data and their enhanced spatial, temporal and spectral resolutions. The rapid development of machine learning over the past decade led to data-driven models being at the forefront of high accuracy predictions of the physical world. However, the contribution of the driving variables behind these predictions cannot be explained beyond generalized metrics of overall performance. Here, we compared the performance of three shallow learners (support vector machines, random forest, and extreme gradient boosting) as well as two deep learners (a convolutional neural network and a residual network with 50 layers) in and around the city of Malmö in southern Sweden. Our complete analysis suite involved 141 input features, 85 scenarios, and 8 LULC classes. We explored the interpretability of the five learners using Shapley additive explanations to better understand feature importance at the level of individual LULC classes. The purpose of class-level feature importance was to identify the most parsimonious combination of features that could reasonably map a particular class and enhance overall map accuracy. We showed that not only do overall accuracies increase from shallow (mean = 84.64%) to deep learners (mean = 92.63%) but that the number of explanatory variables required to obtain maximum accuracy decreases along the same gradient. Furthermore, we demonstrated that class-level importance metrics can be successfully identified using Shapley additive explanations in both shallow and deep learners, which allows for a more detailed understanding of variable importance. We show that for certain LULC classes there is a convergence of variable importance across all the algorithms, which helps explain model predictions and aid the selection of more parsimonious models. The use of class-level feature importance metrics is still new in LULC classification, and this study provides important insight into the potential of more nuanced importance metrics.

Airborne small-footprint full-waveform LiDAR data for urban land cover classification

Airborne small-footprint full-waveform LiDAR data have a unique ability to characterize the landscape because it contains rich horizontal and vertical information. However, a few studies have fully explored its role in distinguishing different objects in the urban area. In this study, we examined the efficacy of small-footprint full-waveform LiDAR data on urban land cover classification. The study area is located in a suburban area in Beijing, China. Eight land cover classes were included: impervious ground, bare soil, grass, crop, tree, low building, high building, and water. We first decomposed waveform LiDAR data, from which a set of features were extracted. These features were related to amplitude, echo width, mixed ratio, height, symmetry, and vertical distribution. Then, we used a random forest classifier to evaluate the importance of these features and conduct the urban land cover classification. Finally, we assessed the classification accuracy based on a confusion matrix. Results showed that Afirst was the most important feature for urban land cover classification, and the other seven features, namely, ωfirst, HEavg, nHEavg, RAω, SYMS, Srise, and ωRf_fl, also played important roles in classification. The random forest classifier yielded an overall classification accuracy of 94.7%, which was higher than those from previous LiDAR-derived classifications. The results indicated that full-waveform LiDAR data could be used for high-precision urban land cover classification, and the proposed features could help improve the classification accuracy.

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.

Machine Learning Approaches for Developing Land Cover Mapping.

In remote sensing data processing, cover classification on decimeter-level data is a well-studied but tough subject that has been well-documented. The majority of currently existent works make use of orthographic photographs or orthophotos and digital surface models that go with them (DSMs). Urban land cover classification plays a significant role in the field of remote sensing to enhance the quality of different applications including environment protection, sustainable development, and resource management and planning. Novelty of the research done in this area is focused on extracting features from high-resolution satellite images to be used in the classification process. However, it is well known in machine learning literature that some of the extracted features are irrelevant to the classification process with a negative or no effect on its accuracy. In this work, a genetic algorithm-based feature selection approach is used to enhance the performance of urban land cover classification. Neural networks (NNs) and random forest (RF) classifiers were used to evaluate the proposed approach on a recent urban land cover dataset of nine different classes. Experimental results show that the proposed approach achieved better performance with RF classifier using only 27% of the features. The random forest tree has achieved highest accuracy 84.27%; it is concluded that the RF algorithm is an appropriate algorithm for classifying cover land.

Monitoring urbanization and environmental impact in Kigali, Rwanda using Sentinel-2 MSI data and ecosystem service bundles

• U-Net achieved good performance in urban imperviousness density extraction from Sentinel-2 MSI data. • The hybrid approach using Random Forest and U-Net improved urban land cover classification accuracies. • The expansion of built-up areas resulted in high demand for provisioning and regulating ecosystem services. • The decrease in cropland and forest negatively affected ecosystem service supply. • The framework using Sentinel-2 MSI data and ES bundles is effective for urbanization monitoring and environmental impact. Rapid urbanization in developing countries often results in uncontrolled urban growth. In order to support sustainable urban development, reliable and up-to-date information on urban land cover changes and their environmental impact is needed. In this study, we aim at evaluating the potential of Sentinel-2 (S-2) Multispectral Instrument (MSI) data for urban land cover change monitoring and for analyzing resulting impacts on Ecosystem Services (ES) provision in Kigali, Rwanda. Land cover classification into eight distinct urban classes (84% overall accuracies, 0.8 Kappa) was performed on data from 2016 and 2021 using a hybrid approach combining Random Forest with a U-Net-based impervious surface segmentation that improved the delineation of urban areas. The bi-temporal land cover maps were then analyzed regarding landscape structure using Landscape Metrics (LM). Ecosystem service bundles were derived for both years and their changes were summarized. Service providing areas were further evaluated in terms of changes in spatial attributes and structure of patches. ES were aggregated into eight bundles and grouped into provisioning, regulating and supporting services. The bundles were further analyzed using a matrix spatially linking landscape units with service supply and demand budgets. The results illustrated that three urban development scenarios can be distinguished including infill through housing and infrastructures development in core urban areas, urban sprawl in fringe zones and the development of urban patches at distant locations intercepted by cropland. The results revealed that the changes in LM negatively affected ES supply mainly through a decrease in cropland and forest. The expansion of built-up areas resulted in a high demand for provisioning and regulating services, especially food and water provision, surface runoff mitigation and erosion control. This is the first study demonstrating that detailed monitoring of urbanization and resulting environmental impacts can be performed with open access S-2 MSI data in Sub-Saharan Africa. Moreover, the framework developed in this study has the potential to be transferred to other Sub-Saharan cities.