Satellite Images Research Articles

Assessing Tunisia's urban air quality using combined lichens and Sentinel-5 satellite integration.

In Tunisia, urban air pollution is becoming a bigger problem. This study used a combined strategy of biomonitoring with lichens and satellite mapping with Sentinel-5 satellite data processed in Google Earth Engine (GEE) to assess the air quality over metropolitan Tunis. Lichen diversity was surveyed across the green spaces of the Faculty of Science of Tunisia sites, revealing 15 species with a predominance of pollution-tolerant genera. The Index of Atmospheric Purity (IAP) calculated from the lichen data indicated poor air quality. Spatial patterns of pollutants sulfur dioxide (SO2), ozone (O3), nitrogen dioxide (NO2), carbon monoxide (CO), and aerosol index across Greater Tunis were analyzed from Sentinel-5 datasets on the GEE platform. The higher values of these indices in the research area indicate that it may be impacted by industrial activity and highlight the considerable role that vehicle traffic plays in air pollution. The results of the IAP, IBL, and the combined ground-based biomonitoring and satellite mapping techniques confirm poor air quality and an environment affected by atmospheric pollutants which will enable proactive air quality management strategies to be put in place in Tunisia's rapidly expanding cities.

Predicting wheat yield gap and its determinants combining remote sensing, machine learning, and survey approaches in rainfed Mediterranean regions of Morocco

Wheat plays a crucial role in Morocco’s food security, economic stability, and livelihoods of farming communities. Assessing key vegetation indices (as yield predictors), along with understanding potential yield, yield gap, and major determinants for this gap at regional and national scales, is vital for improving food security with resilience in variable climatic conditions. Analysing the yield gap and its causes during drought and optimal weather conditions can reduce crop failure risks and enhance productivity specially in variable rainfed production systems. This study aimed to develop scalable methodology to predict field- and landscape-level yield and yield gaps for wheat and their underlying causes examplifying Morocco’s rainfed production environment combining remote sensing, machine learning, and ground information. By analysing six vegetation indices (EVI2, CGVI, MSR, NDVI, OSAVI, and RVI) derived from Sentinel-2 satellite imagery (10 m resolution) over three successive growing seasons (2018–2019, 2019–2020, and 2020–2021), the study employed advanced vegetation index models for accurate prediction of wheat yields and yield gaps at plot and on a larger regional scale within the Rabat-Sale-Kenitra region. To identify the determinants of yield gap, climate and soil datasets were merged with crop management information and the random forest model was fine-tuned and assessed for each season and cumulatively. The findings highlighted that RVI, GCVI, and NDVI vegetation indices were particularly effective in predicting wheat yields, showing the highest R2 and the lowest prediction errors (RMSE). Such predictive methodologies are crucial for policymakers to proactively plan and mitigate risk minimization and adaption plans at regional and national levels. The models predicted rainfed potential yields of 5.99, 1.53, and 4.66 t ha−1, with corresponding yield gap of 3.38, 0.73, and 1.58 t ha−1 for the seasons of 2018/2019 (favorable); 2019/2020 (drought) and 2020/2021 (favorable), respectively. Across three periods, critical factors determining yield include soil moisture, total rainfall during the crop growing period, evapotranspiration, and soil texture and carbon content. To minimize drought risks and maximize benefits during variable rainfall conditions, it is essential to implement pre-season drought forecasts, customize seeding dates based on soil moisture, adopt technologies that enhance soil moisture retention, and utilize climate-adapted farming practices in the semi-arid and arid rainfed regions.

Automatic detection of methane emissions in multispectral satellite imagery using a vision transformer

Curbing methane emissions is among the most effective actions that can be taken to slow down global warming. However, monitoring emissions remains challenging, as detection methods have a limited quantification completeness due to trade-offs that have to be made between coverage, resolution, and detection accuracy. Here we show that deep learning can overcome the trade-off in terms of spectral resolution that comes with multi-spectral satellite data, resulting in a methane detection tool with global coverage and high temporal and spatial resolution. We compare our detections with airborne methane measurement campaigns, which suggests that our method can detect methane point sources in Sentinel-2 data down to plumes of 0.01 km2, corresponding to 200 to 300 kg CH4 h−1 sources. Our model shows an order of magnitude improvement over the state-of-the-art, providing a significant step towards the automated, high resolution detection of methane emissions at a global scale, every few days.

Advancing spatial mapping for satellite image road segmentation with multi-head attention

Advancing spatial mapping for satellite image road segmentation with multi-head attention

Deep learning-based harmonization and super-resolution of Landsat-8 and Sentinel-2 images

Multi-spectral satellite images of the Earth’s surface are used in various applications, from water quality assessment and urban planning to climate monitoring, disaster response, infrastructure oversight, and agricultural surveillance. Many of these applications would benefit from higher spatial and temporal resolution of observations, which could be achieved by combining observations from several sources. This study introduces a deep learning-based pipeline to harmonize the spectral and spatial discrepancies between the Landsat-8 and Sentinel-2 Earth Observation satellites. Through established image quality metrics, we demonstrate a significant enhancement in the spatial resolution of Landsat-8 images. Field observation experiments show that leveraging unified images from both satellites increases the availability of cloud-free images by 21% annually on average in our study area. Additionally, our pipeline enhances the Normalized Difference Vegetation Index (NDVI) correlation between Landsat-8 and Sentinel-2 observations by about 4.9%, offering significant performance gains in a downstream crop segmentation task. Our 100M parameter model, trained on European data, generalizes to most regions with only minor limitations. Furthermore, we show that the pipeline can provide uncertainty estimates for its outputs, which are valuable for decision-making in downstream applications. Our code base is available online: https://github.com/venkatesh-thiru/Deep-Harmonization.

Estimation of Suspended Sediment Concentration along the Lower Brazos River Using Satellite Imagery and Machine Learning

This article focuses on developing models that estimate suspended sediment concentrations (SSCs) for the Lower Brazos River, Texas, U.S. Historical samples of SSCs from gauge stations and satellite imagery from Landsat Missions and Sentinel Mission 2 were utilized to develop models to estimate SSCs for the Lower Brazos River. The models used in this study to accomplish this goal include support vector machines (SVMs), artificial neural networks (ANNs), extreme learning machines (ELMs), and exponential relationships. In addition, flow measurements were used to develop rating curves to estimate SSCs for the Brazos River as a baseline comparison of the models that used satellite imagery to estimate SSCs. The models were evaluated using a Taylor Diagram analysis on the test data set developed for the Brazos River data. Fifteen of the models developed using satellite imagery as inputs performed with a coefficient of determination R2 above 0.69, with the three best performing models having an R2 of 0.83 to 0.85. One of the best performing models was then utilized to estimate the SSCs before, during, and after Hurricane Harvey to evaluate the impact of this storm on the sediment dynamics along the Lower Brazos River and the model’s ability to estimate SSCs.

Semantic Segmentation for Change Detection in Satellite Imaging

Change detection is a common and actual problem in the field of remote sensing. The classical approaches using raw pixel information are very sensitive to noise. In this study we propose the usage of additional semantic information for change detection. We use the semantic segmentation methods like geospatial Segment Anything Model and encoder based U-Net to evaluate the predictions and tracing the semantic information as well as raw information in change detection. Later the multidimensional time series data is used via the Vector Autoregression model to predict the future changes in the landscape. The observations which fall out of the prediction interval are considered as the changes in the landscape. The proposed method is evaluated on the dataset of the random locations across the Baltic region. The research is accompanied by the data and reproducible code at Github repository.

Global area boom for greenhouse cultivation revealed by satellite mapping.

Greenhouse cultivation has been expanding rapidly in recent years, yet little knowledge exists on its global extent and expansion. Using commercial and freely available satellite data combined with artificial intelligence techniques, we present a global assessment of greenhouse cultivation coverage and map 1.3 million hectares of greenhouse infrastructures in 2019, a much larger extent than previously estimated. Our analysis includes both large (61%) and small-scale (39%) greenhouse infrastructures. Examining the temporal development of the 65 largest clusters (>1,500 ha), we show a recent upsurge in greenhouse cultivation in the Global South since the 2000s, including a dramatic increase in China, accounting for 60% of the global coverage. We emphasize the potential of greenhouse infrastructures to enhance food security but raise awareness of the uncertain environmental and social implications that may arise from this expansion. We further highlight the gap in spatio-temporal datasets for supporting future research agendas on this critical topic.

Detecting Coseismic Landslides in GEE Using Machine Learning Algorithms on Combined Optical and Radar Imagery

Landslides, resulting from disturbances in slope equilibrium, pose a significant threat to landscapes, infrastructure, and human life. Triggered by factors such as intense precipitation, seismic activities, or volcanic eruptions, these events can cause extensive damage and endanger nearby communities. A comprehensive understanding of landslide characteristics, including spatio-temporal patterns, dimensions, and morphology, is vital for effective landslide disaster management. Existing remote sensing approaches mostly use either optical or synthetic aperture radar sensors. Integrating information from both these types of sensors promises greater accuracy for identifying and locating landslides. This study proposes a novel approach, the ML-LaDeCORsat (Machine Learning-based coseismic Landslide Detection using Combined Optical and Radar Satellite Imagery), that integrates freely available Sentinel-1, Palsar-2, and Sentinel-2 imagery data in Google Earth Engine (GEE). The approach also integrates relevant spectral indices and suitable bands used in a machine learning-based classification of coseismic landslides. The approach includes a robust and reproducible training and validation strategy and allows one to choose between five classifiers (CART, Random Forest, GTB, SVM, and Naive Bayes). Using landslides from four different earthquake case studies, we demonstrate the superiority of our approach over existing solutions in coseismic landslide identification and localization, providing a GTB-based detection accuracy of 87–92%. ML-LaDeCORsat can be adapted to other landslide events (GEE script is provided). Transfer learning experiments proved that our model can be applied to other coseismic landslide events without the need for additional training data. Our novel approach therefore facilitates quick and reliable identification of coseismic landslides, highlighting its potential to contribute towards more effective disaster management.

Fire Vulnerability, Resilience, and Recovery Rates of Mediterranean Pine Forests Using a 33-Year Time Series of Satellite Imagery

The modification of fire regimes and their impact on vegetation recovery, soil properties, and fuel structure are current key research areas that attempt to identify the thresholds of vegetation’s susceptibility to wildfires. This study aimed to evaluate the vulnerability of Mediterranean pine forests (Pinus halepensis Mill. and Pinus pinaster Aiton) to wildfires, analyzing two major forest fires that occurred in Yeste (Spain) in 1994 and 2017, affecting over 14,000 and 3200 hectares, respectively. Four recovery regions were identified based on fire severity—calculated using the delta Normalized Burn Ratio (dNBR) index—and recurrence: areas with high severity in 2017 but not in 1994 (UB94-HS17), areas with high severity in 1994 but not in 2017 (HS94-UB17), areas with high severity in both fires (HS94-HS17), and areas unaffected by either fire (UB94-UB17). The analysis focused on examining the recovery patterns of three spectral indices—the Normalized Difference Vegetation Index (NDVI), Normalized Moisture Index (NDMI), and Normalized Burn Ratio (NBR)—using the Google Earth Engine platform from 1990 to 2023. Additionally, the Relative Recovery Indicator (RRI), the Ratio of Eighty Percent (R80P), and the Year-on-Year average (YrYr) metrics were computed to assess the spectral recovery rates by region. These three spectral indices showed similar dynamic responses to fire. However, the Mann–Kendall and unit root statistical tests revealed that the NDVI and NDMI exhibited distinct trends, particularly in areas with recurrence (HS94-HS17). The NDVI outperformed the NBR and NDMI in distinguishing variations among regions. These results suggest accelerated vegetation spectral regrowth in the short term. The Vegetation Recovery Capacity After Fire (VRAF) index showed values from low to moderate, while the Vulnerability to Fire (V2FIRE) index exhibited values from medium to high across all recovery regions. These findings enhance our understanding of how vegetation recovers from fire and how vulnerable it is to fire.

Spatiotemporal distribution of mangrove along the Egyptian Red Sea coast and analysis of hydrological impact on growth patterns

AbstractThis study presents an in-depth spatiotemporal analysis of mangrove ecosystems along Egypt’s Red Sea coast, utilizing satellite imagery and GIS to examine changes from 2003 to 2022. We evaluate the effects of hydrological factors, specifically rainfall and runoff -presented by Stream Power Index-, on mangrove growth patterns. Results indicate a significant increase in mangrove areas, with a notable annual growth rate, despite a reduction in a specific region. This research highlights the integral role of catchment area runoff (R2 = 0.735, R = 0.857, P-value = 0.003 < 0.05, CV = 70.26%), rather than direct rainfall, in mangrove expansion, contributing to the understanding of mangrove resilience and informing sustainable coastal management strategies. The study bridges a significant research gap by mapping decadal mangrove changes, offering insights into the dynamics affecting these crucial ecosystems.

Crop Yield Prediction using Machine Learning

Weather profoundly impacts agricultural outcomes, making accurate crop prediction vital for farmers' decision-making. This abstract presents a comprehensive overview of weather-based crop prediction, emphasizing its significance, key components, and methodologies. The process begins with the collection and analysis of historical weatherdata encompassing variables such as temperature, precipitation, humidity, and sunlight. Utilizing Python programming and data visualization libraries like Pandas and Matplotlib facilitates the exploration and visualization of this data, revealing trends and patterns. Machine learning algorithms, including regression and ensemble methods, are employedto develop predictive models. These models leverage historical weather data to forecast future crop yields accurately. Python's extensive libraries, such as Scikit-learn and TensorFlow, offer robust tools for model development and evaluation. Incorporating advanced technologies like remote sensing and satellite imagery further refines the prediction process. These tools provide real-time insights into crop health and growth, enhancing the precision of forecasts. Ultimately, weather-based crop prediction serves as a valuable decision support tool for farmers, enabling informed choices regarding planting, irrigation, and harvesting practices. By harnessing historical weather data, machine learning algorithms, and innovative technologies, stakeholders can optimize agricultural productivity, mitigate risks, and contribute to global food security

Object Detection using Convolutional Neural Network

During the last years, a noticeable growth is observed in the field of computer vision research. In computer vision, object detection is a task of classifying and localizing the objects in order to detect the same. The widely used object detection applications are human– computer interaction, video surveillance, satellite imagery, transport system, and activity recognition. In the wider family of deep learning architectures, convolutional neural network (CNN) made up with set of neural network layers is used for visual imagery. Deep CNN architectures exhibit impressive results for detection of objects in digital image. This paper represents a comprehensive review of the recent development in object detection using convolutional neural networks

Adaptability and Interconnectedness in Latin American Innerburbs: An Assessment of the Public/Private Interface in Informal Settlements

An “Innerburb” is an urban structure that emerged between the 1950s and 1980s, settled in rural areas, and is considered the first periphery of the Latin American city. This structure results from socio-spatial and territorial evolutionary processes, constituting the pinnacle of informal evolution. However, despite offering a comprehensive perspective on the informal problem, innerburbs have been scarcely reviewed in the literature. This article explores the Latin American innerburb by adopting as a method a comparative study of the public/private interface in the cases of San Cosme in Lima, Villa Rodolfo Ricciardelli in Buenos Aires and Vila Heliopolis in São Paulo, evaluating their adaptation and interconnection with the city through morphological indicators, using morphological mapping through satellite images as an analytical tool, and using Space Syntax as a topological approach in the analysis of connectivity and visibility indicators. The objective of the research is the detection of morphological patterns that alter the functioning of the public/private interface in innerburbs. The results show that the existence of impermeable facades, the lack of public spaces and the illegal appropriation in the development of informal practices are transgressive adaptability patterns at the micro-scale that affect the interface, drastically limiting the interconnection between the innerburb and the formal fabric, restricting its morphological openness and affecting the development of socioeconomic activities. As a discussion and conclusion, corrective measures for progressive improvement in innerburbs are established, focusing on the adaptability of housing and road space as a means of interconnection between the innerburb and the city.

Using satellite image data to identify rice varieties through linear spectral unmixing method (case study: Karangjati Sub District, Ngawi Regency)

<p>Remote sensing technology has increasingly emerged as a potent tool for precision agriculture, particularly in facilitating the mapping and monitoring of crops on a large scale. An application of this technology is the identification of different types of rice by analyzing the pixels acquired in satellite images. Regrettably, the pixels in the image have been mixed from different recorded items. Therefore, they have the potential to influence the outcome of the identification. An effective approach to address this problem is to employ the linear spectral unmixing (LSU) technique. The LSU approach quantifies the ratio of pure objects in every pixel of an image by utilizing the spectral value associated with the endmember of the rice variety. The investigation was carried out in the Karangjati District during the generative stage (70 ± DAP) of the rice planting season. The data indicates that the dominant variety is Inpari 32 HDB. The data validation tests, which involved the use of a confusion matrix and Kappa analysis, resulted in an overall accuracy rate of 85.48% and a Kappa analysis score of 70.6%.<strong></strong></p>

Novel Learning of Bathymetry from Landsat 9 Imagery Using Machine Learning, Feature Extraction and Meta-Heuristic Optimization in a Shallow Turbid Lagoon

Bathymetry data is indispensable for a variety of aquatic field studies and benthic resource inventories. Determining water depth can be accomplished through an echo sounding system or remote estimation utilizing space-borne and air-borne data across diverse environments, such as lakes, rivers, seas, or lagoons. Despite being a common option for bathymetry mapping, the use of satellite imagery faces challenges due to the complex inherent optical properties of water bodies (e.g., turbid water), satellite spatial resolution limitations, and constraints in the performance of retrieval models. This study focuses on advancing the remote sensing based method by harnessing the non-linear learning capabilities of the machine learning (ML) model, employing advanced feature selection through a meta-heuristic algorithm, and using image extraction techniques (i.e., band ratio, gray scale morphological operation, and morphological multi-scale decomposition). Herein, we validate the predictive capabilities of six ML models: Random Forest (RF), Support Vector Machine (SVM), CatBoost (CB), Extreme Gradient Boost (XGB), Light Gradient Boosting Machine (LGBM), and KTBoost (KTB) models, both with and without the application of meta-heuristic optimization (i.e., Dragon Fly, Particle Swarm Optimization, and Grey Wolf Optimization), to accurately ascertain water depth. This is achieved using a diverse input dataset derived from multi-spectral Landsat 9 imagery captured on a cloud-free day (19 September 2023) in a shallow, turbid lagoon. Our findings indicate the superior performance of LGBM coupled with Particle Swamp Optimization (R2 = 0.908, RMSE = 0.31 m), affirming the consistency and reliability of the feature extraction and selection-based framework, while offering novel insights into the expansion of bathymetric mapping in complex aquatic environments.

Geostatistical Analysis of Lineament Domains: The Study Case of the Apennine Seismic Province of Italy

Regional-scale swarms of subparallel linear topographic features, known as lineament domains, are a common feature of planetary surfaces. Lineament domains are superficial manifestations of the crustal stress field trajectory. Notably, one of the effects of active tectonics is seismicity. Italy is one of the most seismically active regions in the Mediterranean, with many destructive earthquakes that have occurred in past centuries. Here, we assess the seismic meaning of the main lineament domain in the tectonically active region of Central Italy. We describe the use of an automated analysis of satellite imagery coupled with spatial grid analysis to identify three lineament domains of the Central Apennines. Spatial and azimuthal comparisons of the main lineament domain (i.e., the Apennine Domain), with the known locations of earthquakes (moment magnitude of Mw > 5.5) that occurred during the past century, revealed the most seismically active tectonic areas and their spatial distributions. Further, we present a conceptual seismo-geodynamic model for the Central Apennines, which is characterized by regional arching and explains the presence of an extensional tectonic regime in the upper crustal layer of the active Apennines fold-and-thrust belt.

Development of Multi-Strip Image Mosaicking for KOMPSAT-3A Images

Abstract. High-resolution satellite imagery has a limitation in terms of coverage area. This limitation presents challenges for extensive-scale analysis at regional or national levels. To maximize the utility of high-resolution satellite imagery, the implementation of image mosaicking techniques is essential. In this paper, we have developed seamline extraction techniques and relative geometric correction optimized for high-resolution satellite imagery. Ultimately, we proposed a multi-strip image mosaicking method for KOMPSAT-3A (Korea Multi-Purpose Satellite-3A) images. We applied the Dijkstra's shortest path algorithm to efficiently extract seamlines. we also performed image registration based on feature matching and homography transformation to correct the relative geometric errors between input images. We conducted experiments with our methods using 29 scenes from KOMPSAT-3A L1G data. The results indicated high relative geometric accuracy, with an average error of 1.63 pixels. Furthermore, we were able to obtain high-quality seamless mosaic images. Our proposed method is expected to enhance the utility of KOMPSAT-3A imagery for large-scale environmental and urban analysis and to provide more accurate and comprehensive data.

Geometric accuracy evaluation and analysis of ZY-1 02E IRS thermal infrared image data using GCP extraction based on phase correlation matching method

Abstract. The Ziyuan-1 (ZY-1) 02E launched on December 26, 2021 is equipped with a thermal infrared sensor (IRS), which has a ground resolution of better than 16m and a width priority of 115km, balancing the advantages of high resolution and large wide observation. The geometric performance of image data is the premise of remote sensing application, and the difficulty in evaluating the geometric performance of thermal infrared image data lies in the extraction of well-distributed, reliable and accurate GCPs. To extract GCP from high-precision reference images, it is necessary to overcome the feature differences between images caused by different spectral responses. This paper adopts a phase correlation matching method based on frequency domain to realize the fine registration of the data obtained by the emission thermal spectral band with the data from the reflectance spectral band, which can not only solve the GCP extraction of conventional thermal infrared images collected during the day, but also obtain satisfactory GCP data from thermal infrared data acquired at night. In order to test the GCP method proposed in this paper, three typical areas are selected as the experimental areas, including Yiyang City in Hunan, Nagqu City in Xizang and Hami City in Xinjiang, and the internal geometric accuracy and absolute geolocation accuracy of the thermal infrared data spanning one year are evaluated and analyzed by using the reference data composed of the DOM with an accuracy of 2m and the DEM with an accuracy of 10m. The research results indicate that the internal geometric accuracy of ZY-1 02E IRS satellite image data is better than 1.0 pixels, and the performance is satisfactory. However, its absolute geolocation accuracy needs to be continuously improved, especially there are systematic errors in the ascending data at night that require further research. Overall, it meets the design accuracy indicators of satellites and can meet the application requirements of thermal infrared remote sensing.

Triplet Epipolar Images for Spaceborne Three Linear Cameras

Abstract. Epipolar images can reduce conjugate points searching from a two-dimensional to a one-dimensional space, which significantly improves the robustness and efficiency of dense matching. However, two images matched the lack of redundancy and suffered from occlusions. The third image could introduce two more stereos and improve the accuracy and completeness of the three-dimension reconstruction, which were adopted by spaceborne three linear cameras (TLCs), the ALOS PRSIM and ZY-3. In this study, we propose a new triplet epipolar images generation method for spaceborne TLCs. Triplet epipolar images were defined as any two of the three epipolar image pairs. The triplet epipolar geometry requires the imaging rays of any conjugate point to be coplanar. For high-resolution satellite images, parallel projection can be used to approximate the imaging geometry of satellite images. Therefore, a coplanar principal optical axis is a fundamental requirement for spaceborne TLCs. We propose a general workflow to generate triplet epipolar images (TEIs), which includes free-net bundle adjustment of TLCs, building a triplet epipolar geometry, correcting the y-parallaxes, and generating a rational function model (RFM) for TEIs. The ZY3-02 satellite images were used to validate the proposed method. The root means square error (RMSEs) of the free-net bundle adjustment in the image space was 0.185 pixels, which proved the fine intrinsic accuracy. After compensation, the RMSEs of the y parallaxes of the three epipolar image pairs were 0.295, 0.310, and 0.370 pixels. Owing to the simple geometry of the TEIs, the RMSEs of the RFM replacements were within 0.001 pixels.