Earth Observation Community Research Articles

Evaluation of the SpatioTemporal Asset Catalog for management and discovery of FAIR flood hazard models

Approaches for performing flood hazards modeling and risk assessment at federal, state, and local agencies are undergoing emergent challenge for consistent metadata and cataloging systems to ensure the sharing of flood risk data in a Findable, Accessible, Interoperable, and Reusable (FAIR) manner. This paper explores the suitability of a suite of software and specifications developed by the Earth observation community for environmental modeling, which adhere to the FAIR principles not only for managing published or authoritative data but throughout the model development and flood hazard analysis phases. Specifically, we evaluate the SpatioTemporal Asset Catalog (STAC) in a pilot study undertaken as part of the Future of Flood Risk Data (FFRD) initiative of FEMA. The experimental results indicate the STAC ecosystem offers a flexible cloud native approach for linking data, managing metadata, and cataloging collections of models. Further, the STAC framework shows favorable results in a probabilistic and other use cases.

Ecosystem Resilience Monitoring and Early Warning Using Earth Observation Data: Challenges and Outlook

AbstractAs the Earth system is exposed to large anthropogenic interferences, it becomes ever more important to assess the resilience of natural systems, i.e., their ability to recover from natural and human-induced perturbations. Several, often related, measures of resilience have been proposed and applied to modeled and observed data, often by different scientific communities. Focusing on terrestrial ecosystems as a key component of the Earth system, we review methods that can detect large perturbations (temporary excursions from a reference state as well as abrupt shifts to a new reference state) in spatio-temporal datasets, estimate the recovery rate after such perturbations, or assess resilience changes indirectly from stationary time series via indicators of critical slowing down. We present here a sequence of ideal methodological steps in the field of resilience science, and argue how to obtain a consistent and multi-faceted view on ecosystem or climate resilience from Earth observation (EO) data. While EO data offers unique potential to study ecosystem resilience globally at high spatial and temporal scale, we emphasize some important limitations, which are associated with the theoretical assumptions behind diagnostic methods and with the measurement process and pre-processing steps of EO data. The latter class of limitations include gaps in time series, the disparity of scales, and issues arising from aggregating time series from multiple sensors. Based on this assessment, we formulate specific recommendations to the EO community in order to improve the observational basis for ecosystem resilience research.

Validating Digital Earth Australia NBART for the Landsat 9 Underfly of Landsat 8

In recent years, Geoscience Australia has undertaken a successful continental-scale validation program, targeting Landsat and Sentinel analysis-ready data surface reflectance products. The field validation model used for this program was successfully built upon earlier studies, and the measurement uncertainties associated with these protocols have been quantified and published. As a consequence, the Australian earth observation community was well-prepared to respond to the United States Geological Survey (USGS) call for collaborators with the 2021 Landsat 8 (L8) and Landsat 9 (L9) underfly. Despite a number of challenges, seven validation datasets were captured across five sites. As there was only a single 100% overlap transit across Australia, and the country was amidst a strong La Niña climate cycle, it was decided to deploy teams to the two available overpasses with only 15% side lap. The validation sites encompassed rangelands, chenopod shrublands, and a large inland lake. Apart from instrument problems at one site, good weather enabled the capture of high-quality field data allowing for meaningful comparisons between the radiometric performance of L8 and L9, as well as the USGS and Australian Landsat analysis-ready data processing models. Duplicate (cross-calibration) spectral sampling at different sites provides evidence of the field protocol reliability, while the off-nadir view of L9 over the water site has been used to better compare the performance of different water and atmospheric correction processing models.

Earth Observation to Address Inequities in Post‐Flood Recovery

AbstractFloods impact communities worldwide, resulting in loss of life, damaged infrastructure and natural assets, and threatened livelihoods. Climate change and urban development in flood‐prone areas will continue to worsen flood‐related losses, increasing the urgency for effective tools to monitor recovery. Many Earth Observation (EO) applications exist for flood‐hazard monitoring and provide insights on location, timing, and extent in near real‐time and historically to estimate flood risk. Less attention has been paid to flood recovery, even though differing recovery rates and outcomes can have immediate and enduring distributional effects within communities. EO data are uniquely positioned to monitor post‐flood recovery and inform policy on hazard mitigation and adaptation but remain underutilized. We encourage the EO and flood research community to refocus on developing flood recovery applications to address growing risk. Translation of EO insights on flood recovery among flood‐affected communities and decision‐makers is necessary to address underlying social vulnerabilities that exacerbate inequitable recovery outcomes and advocate for redressing injustices where disparate recovery is observed. We identify an unequivocal need for EO to move beyond mapping flood hazard and exposure toward post‐flood recovery monitoring to inform recovery across geographic contexts. This commentary proposes a framework for remote sensing scientists to engage community‐based partners to integrate EO with non‐EO data to advance flood recovery monitoring, characterize inequitable recovery, redistribute resources to mitigate inequities, and support risk reduction of future floods.

Environmental impacts of earth observation data in the constellation and cloud computing era

Numbers of Earth Observation (EO) satellites have increased exponentially over the past decade reaching the current population of 1193 (January 2023). Consequently, EO data volumes have mushroomed and data storage and processing have migrated to the cloud. Whilst attention has been given to the launch and in-orbit environmental impacts of satellites, EO data environmental footprints have been overlooked. These issues require urgent attention given data centre water and energy consumption, high carbon emissions for computer component manufacture, and difficulty of recycling computer components. Doing so is essential if the environmental good of EO is to withstand scrutiny. We provide the first assessment of the EO data life-cycle and estimate that the current size of the global EO data collection is ~807 PB, increasing by ~100 PB/year. Storage of this data volume generates annual CO2 equivalent emissions of 4101 t. Major state-funded EO providers use 57 of their own data centres globally, and a further 178 private cloud services, with considerable duplication of datasets across repositories. We explore scenarios for the environmental cost of performing EO functions on the cloud compared to desktop machines. A simple band arithmetic function applied to a Landsat 9 scene using Google Earth Engine (GEE) generated CO2 equivalent (e) emissions of 0.042–0.69 g CO2e (locally) and 0.13–0.45 g CO2e (European data centre; values multiply by nine for Australian data centre). Computation-based emissions scale rapidly for more intense processes and when testing code. When using cloud services such as GEE, users have no choice about the data centre used and we push for EO providers to be more transparent about the location-specific impacts of EO work, and to provide tools for measuring the environmental cost of cloud computation. The EO community as a whole needs to critically consider the broad suite of EO data life-cycle impacts.

Analysis-Ready Data and FAIR-AI—Standardization of Research Collaboration and Transparency Across Earth-Observation Communities [Technical Committees

Analysis-Ready Data and FAIR-AI—Standardization of Research Collaboration and Transparency Across Earth-Observation Communities [Technical Committees

Dependent or not? From a daily practice of Earth observation research in the Global South to promoting adequate developmental spaces in science and technology studies

Abstract. Ever since the operation of the first civilian Earth observation (EO) satellites gained momentum in the 1970s, potential benefits and disadvantages of transferring space science and technologies, such as remote sensing techniques, have also been discussed in relation to developing countries. However, this debate has so far largely taken place at a macro-comparative level. This paper presents results from moving to the ethnographic micro level in southwest Nigeria. It sets the experiences of researchers from the Global South, who use remote sensing data, in relation to a critical review of (post-)development theory perspectives and corresponding discourses in postcolonial science and technology studies (STS). The paper discusses how researchers construct collective agency towards capacity building as a shared liberatory language in relation to an amalgam of experienced and contested places in the EO community. At the intersection of STS, geography and the arena of development policies, these experiences create their own spatial references to a developing niche that invites scholars and development practitioners to rethink and reorganise knowledge production and technologies in a postcolonial world.

Satellite-based global maps are rarely used in forest reference levels submitted to the UNFCCC

The Earth observation (EO) community is coordinating a range of activities in support of the Global Stocktake. One objective is to enhance the uptake of satellite-based global-scale maps (hereafter ‘EO products’) in national greenhouse gas (GHG) inventories submitted to the United Nations Framework Convention on Climate Change (UNFCCC). To measure progress towards this objective, we compile information on the use of EO products on land cover, fire, and above-ground biomass to derive carbon flux estimates in forest reference levels from 56 tropical countries submitted to the UNFCCC between 2014–2022. The global forest change (GFC) was the only EO product used to measure land extent and change, and was used by almost half the countries. Only two countries used existing EO products for fire mapping. Four countries used biomass maps, although only indirectly, such as for comparing with biomass estimates from field plot measurements or with IPCC defaults. The uptake is limited but improved the measurement, reporting and verification capacity of 22 countries. The relatively high uptake of the GFC demonstrates the importance of meeting essential conditions in the IPCC guidance when developing EO products, including conditions on spatial and temporal resolution, temporal coverage and consistency, and the flexibility to adapt to biophysical thresholds in national definitions. The limited use of other global land EO products underlines the need for developers of EO products to interact with groups responsible for GHG inventories and experts familiar with IPCC guidance so that their products are suitable for national reporting, and thus contribute to more complete aggregated estimates in the Global Stocktake.

CMID: A Unified Self-Supervised Learning Framework for Remote Sensing Image Understanding

Self-supervised learning (SSL) has gained widespread attention in the remote sensing (RS) and earth observation (EO) communities owing to its ability to learn task-agnostic representations without human-annotated labels. Nevertheless, most existing RS SSL methods are limited to learning either global semantic separable or local spatial perceptible representations. We argue that this learning strategy is suboptimal in the realm of RS, since the required representations for different RS downstream tasks are often varied and complex. In this study, we proposed a unified SSL framework that is better suited for RS images representation learning. The proposed SSL framework, Contrastive Mask Image Distillation (CMID), is capable of learning representations with both global semantic separability and local spatial perceptibility by combining contrastive learning (CL) with masked image modeling (MIM) in a self-distillation way. Furthermore, our CMID learning framework is architecture-agnostic, which is compatible with both convolutional neural networks (CNN) and vision transformers (ViT), allowing CMID to be easily adapted to a variety of deep learning (DL) applications for RS understanding. Comprehensive experiments have been carried out on four downstream tasks (i.e. scene classification, semantic segmentation, object-detection, and change detection) and the results show that models pre-trained using CMID achieve better performance than other state-of-the-art SSL methods on multiple downstream tasks. The code and pre-trained models will be made available at https://github.com/NJU-LHRS/official-CMID to facilitate SSL research and speed up the development of RS images DL applications.

CloudSEN12, a global dataset for semantic understanding of cloud and cloud shadow in Sentinel-2

Accurately characterizing clouds and their shadows is a long-standing problem in the Earth Observation community. Recent works showcase the necessity to improve cloud detection methods for imagery acquired by the Sentinel-2 satellites. However, the lack of consensus and transparency in existing reference datasets hampers the benchmarking of current cloud detection methods. Exploiting the analysis-ready data offered by the Copernicus program, we created CloudSEN12, a new multi-temporal global dataset to foster research in cloud and cloud shadow detection. CloudSEN12 has 49,400 image patches, including (1) Sentinel-2 level-1C and level-2A multi-spectral data, (2) Sentinel-1 synthetic aperture radar data, (3) auxiliary remote sensing products, (4) different hand-crafted annotations to label the presence of thick and thin clouds and cloud shadows, and (5) the results from eight state-of-the-art cloud detection algorithms. At present, CloudSEN12 exceeds all previous efforts in terms of annotation richness, scene variability, geographic distribution, metadata complexity, quality control, and number of samples.

The science-policy interfaces of the European network for observing our changing planet: From Earth Observation data to policy-oriented decisions

This paper reports on major outcomes of the ERA-PLANET (The European network for observing our changing planet) project, which was funded under Horizon 2020 ERA-net co-funding scheme. ERA-PLANET strengthened the European Research Area in the domain of Earth Observation (EO) in coherence with the European participation to Group on Earth Observation and the Copernicus European Union’s Earth Observation programme. ERA-PLANET was implemented through four projects focused on smart cities and resilient societies (SMURBS), resource efficiency and environmental management (GEOEssential), global changes and environmental treaties (iGOSP) and polar areas and natural resources (iCUPE). These projects developed specific science-policy workflows and interfaces to address selected environmental policy issues and design cost-effective strategies aiming to achieve targeted objectives. Key Enabling Technologies were implemented to enhancing ‘data to knowledge’ transition for supporting environmental policy making. Data cube technologies, the Virtual Earth Laboratory, Earth Observation ontologies and Knowledge Platforms were developed and used for such applications.SMURBS brought a substantial contribution to resilient cities and human settlements topics that were adopted by GEO as its 4th engagement priority, bringing the urban resilience topic in the GEO agenda on par with climate change, sustainable development and disaster risk reduction linked to environmental policies.GEOEssential is contributing to the development of Essential Variables (EVs) concept, which is encouraging and should allow the EO community to complete the description of the Earth System with EVs in a close future. This will clearly improve our capacity to address intertwined environmental and development policies as a Nexus.iGOSP supports the implementation of the GEO Flagship on Mercury (GOS4M) and the GEO Initiative on POPs (GOS4POPs) by developing a new integrated approach for global real-time monitoring of environmental quality with respect to air, water and human matrices contamination by toxic substances, like mercury and persistent organic pollutants. iGOSP developed end-user-oriented Knowledge Hubs that provide data repository systems integrated with data management consoles and knowledge information systems.The main outcomes from iCUPE are the novel and comprehensive data sets and a modelling activity that contributed to delivering science-based insights for the Arctic region. Applications enable defining and monitoring of Arctic Essential Variables and sets up processes towards UN2030 SDGs that include health (SDG 3), clean water resources and sanitation (SDGs 6 and 14).

Fast and accurate multi-class geospatial object detection with large-size remote sensing imagery using CNN and Truncated NMS

Multi-class geospatial object detection with remote sensing imagery has broad prospects in urban planning, natural disaster warning, industrial production, military surveillance and other applications. Accuracy and efficiency are two common measures for evaluating object detection models, and it is often difficult to achieve both at the same time. Developing a practical remote sensing object detection algorithm that balances the accuracy and efficiency is thus a big challenge in the Earth observation community. Here, we propose a comprehensive high-speed multi-class remote sensing object detection method. Firstly, we obtain a multi-volume YOLO (You Only Look Once) v4 model for balancing speed and accuracy, based on a pruning strategy of the convolutional neural network (CNN) and the one-stage object detection network YOLO v4. Moreover, we apply the Manhattan-Distance Intersection of Union (MIOU) loss function to the multi-volume YOLO v4 to further improve the accuracy without additional computational burden.Secondly, mainly due to computing limitations, a remote sensing image that is large-size relative to a natural image must first be divided into multiple smaller tiles, which are then detected separately, and finally, the detection results are spliced back to match the original image. In the process of remote sensing image slicing, a large number of truncated objects appear at the edge of tiles, which will produce a large number of false results in the subsequent detection links. To solve this problem, we propose a Truncated Non-Maximum Suppression (NMS) algorithm to filter out repeated and false detection boxes from truncated targets in the spliced detection results. We compare the proposed algorithm with the state-of-the-art methods on the Dataset for Object deTection in Aerial images (DOTA) and DOTA v2. Quantitative evaluations show that mAP and FPS reach 77.3 and 35 on DOTA, and 61.0 and 74 on DOTA v2. Overall, our method reaches the optimal balance between efficiency and accuracy, and realizes the high-speed remote sensing object detection.

Citizen science for Earth Observation (Citzens4EO): understanding current use in the UK

ABSTRACT The role of Earth observation (EO) data in addressing societal problems from environmental through to humanitarian should not be understated. Recent innovation in EO means provision of analysis ready data and data cubes, which allows for rapid use of EO data. This in combination with processing technologies, such as Google Earth Engine and open source algorithms/software for EO data integration and analyses, has afforded an explosion of information to answer research questions and/or inform policy making. However, there is still a need for both training and validation data within EO projects – often this can be challenging to obtain. It has been suggested that citizen science can help here to provide these data, yet there is some perceived hesitancy in using citizen science within EO projects. This paper reports on the Citizen Science 4 EO (Citizens4EO) project that aimed to obtain an in-depth understanding of researchers’ and practitioners’ experiences with citizen science data in EO within the UK. Through a mixed methods approach (online and in-depth surveys and a spotlight case study) it was found that although the benefits of using citizen science data in EO projects were many (and highlighted in the spotlighted “Slavery from Space” case study), there were a number of common concerns around using citizen science. These were around the mechanics of deploying citizen science and the unreliability of a potentially misinformed or undertrained citizen base. As such, comparing the results of this study with those of a similar survey undertaken in 2016, it is apparent that progress towards optimizing citizen science use in EO has been incremental but positive with evidence of the realization of the benefits of citizen science for EO (Citizens4EO). As such, we conclude by offering priority action areas to support further use of citizen science by the EO community within the UK, which ultimately should be adopted further afield.

Earth observation: An integral part of a smart and sustainable city

Over the course of the 21st century, a century in which the urbanization process of the previous one is ever on the rise, the novel smart city concept has rapidly evolved and now encompasses the broader aspect of sustainability. Concurrently, there has been a sea change in the domain of Earth observation (EO) where scientific and technological breakthroughs are accompanied by a paradigm shift in the provision of open and free data. While the urban and EO communities share the end goal of achieving sustainability, cities still lack an understanding of the value EO can bring in this direction, an next a consolidated framework for tapping the full potential of EO and integrating it in their operational modus operandi. The “SMart URBan Solutions for air quality, disasters and city growth” H2020 project (SMURBS/ERA-PLANET) sits at this scientific and policy crossroad, and, by creating bottom-up EO-driven solutions against an array of environmental urban pressures, and by expanding the network of engaged and exemplary smart cities that push the state-of-the-art in EO uptake, brings the international ongoing discussion of EO for sustainable cities closer to home and contributes in this discussion. This paper advocates for EO as an integral part of a smart and sustainable city and aspires to lead by example. To this end, it documents the project’s impacts, ranging from the grander policy fields to an evolving portfolio of smart urban solutions and everyday city operations, as well as the cornerstones for successful EO integration. Drawing a parallel with the utilization of EO in supporting several aspects of the 2030 Agenda for Sustainable Development, it aspires to be a point of reference for upcoming endeavors of city stakeholders and the EO community alike, to tread together, beyond traditional monitoring or urban planning, and to lay the foundations for urban sustainability.

FAIR1M: A benchmark dataset for fine-grained object recognition in high-resolution remote sensing imagery

With the rapid development of deep learning, many deep learning-based approaches have made great achievements in object detection tasks. It is generally known that deep learning is a data-driven approach. Data directly impact the performance of object detectors to some extent. Although existing datasets include common objects in remote sensing images, they still have some scale, category, and image limitations. Therefore, there is a strong requirement for establishing a large-scale object detection benchmark for high-resolution remote sensing images. In this paper, we propose a novel benchmark dataset with more than 1 million instances and more than 40,000 images for Fine-grAined object recognItion in high-Resolution remote sensing imagery which is named as FAIR1M. We collected remote sensing images with a resolution of 0.3 m to 0.8 m from different platforms, which are spread across many countries and regions. All objects in the FAIR1M dataset are annotated with respect to 5 categories and 37 subcategories by oriented bounding boxes. Compared with existing detection datasets that are dedicated to object detection, the FAIR1M dataset has 4 particular characteristics: (1) it is much larger than other existing object detection datasets both in terms of the number of instances and the number of images, (2) it provides richer fine-grained category information for objects in remote sensing images, (3) it contains geographic information such as latitude, longitude and resolution attributes, and (4) it provides better image quality due to the use of a careful data cleaning procedure. Based on the FAIR1M dataset, we propose three fine-grained object detection and recognition tasks. Moreover, we evaluate several state-of-the-art approaches to establish baselines for future research. Experimental results indicate that the FAIR1M dataset effectively represents real remote sensing applications and is quite challenging for existing methods. Considering the fine-grained characteristics, we improve the evaluation metric and introduce the idea of hierarchy detection into the algorithms. We believe that the FAIR1M dataset will contribute to the earth observation community via fine-grained object detection in large-scale real-world scenes. FAIR1M Website: http://gaofen-challenge.com/.

Dual Network Structure With Interweaved Global-Local Feature Hierarchy for Transformer-Based Object Detection in Remote Sensing Image

Frequent and accurate object detection based on remote sensing images is an encouraging approach for monitoring dynamic of the interest object on earth surface. Transformer-based object detection was recently developed to cope with the trade-off dilemma between large computation load and accuracy sacrifice confronted by region-proposal-based and regression-based object detection, and its self-attention mechanism can provide a global understanding that has potential ability for reasoning the location relationship within sparsely heterogeneously distributed geospatial objects. However, Transformer-based object detection is essentially weak at modeling local feature hierarchy to compensate for the large scale variation of geospatial object, and it is extremely difficult to train due to the lack of inductive bias, resulting in a slow convergence. To overcome the problem, this study proposed a Dual network structure with InterweAved Global-local feature hierarchy based on the TRansformer architecture (DIAG-TR), to alleviate the incompatibility of global and local feature form, and hierarchically embed the local features into global representations. Besides, a learnable anchor box is incorporated into the positional query in the decoder part to provide a spatial prior, which can accelerate convergence. The proposed DIAG-TR is validated on the widely used optical remote sensing image DIOR dataset, and the results demonstrate that the global-local feature hierarchy contributes 3.4% mean average precision compared to the original Transformer-based method, and the convergence time is shortened by 2.5-fold. State-of-the-art methods are also participated as benchmark for comparison, and DIAG-TR outperforms Faster-RCNN-FPN by 8.9%, which proves that DIAG-TR has great potential in earth observation community.

LUCAS Copernicus 2018: Earth-observation-relevant in situ data on land cover and use throughout the European Union

Abstract. The Land Use/Cover Area frame Survey (LUCAS) is an evenly spaced in situ land cover and land use ground survey exercise that extends over the whole of the European Union. LUCAS was carried out in 2006, 2009, 2012, 2015, and 2018. A new LUCAS module specifically tailored to Earth observation (EO) was introduced in 2018: the LUCAS Copernicus module. The module surveys the land cover extent up to 51 m in four cardinal directions around a point of observation, offering in situ data compatible with the spatial resolution of high-resolution sensors. However, the use of the Copernicus module being marginal, the goal of the paper is to facilitate its uptake by the EO community. First, the paper summarizes the LUCAS Copernicus protocol to collect homogeneous land cover on a surface area of up to 0.52 ha. Secondly, it proposes a methodology to create a ready-to-use dataset for Earth observation land cover and land use applications with high-resolution satellite imagery. As a result, a total of 63 364 LUCAS points distributed over 26 level-2 land cover classes were surveyed on the ground. Using homogeneous extent information in the four cardinal directions, a polygon was delineated for each of these points. Through geospatial analysis and by semantically linking the LUCAS core and Copernicus module land cover observations, 58 426 polygons are provided with level-3 land cover (66 specific classes including crop type) and land use (38 classes) information as inherited from the LUCAS core observation. The open-access dataset supplied with this paper (https://doi.org/10.6084/m9.figshare.12382667.v4 d'Andrimont, 2020) provides a unique opportunity to train and validate decametric sensor-based products such as those obtained from the Copernicus Sentinel-1 and Sentinel-2 satellites. A follow-up of the LUCAS Copernicus module is already planned for 2022. In 2022, a simplified version of the LUCAS Copernicus module will be carried out on 150 000 LUCAS points for which in situ surveying is planned. This guarantees a continuity in the effort to find synergies between statistical in situ surveying and the need to collect in situ data relevant for Earth observation in the European Union.

Towards Federated Satellite Systems and Internet of Satellites: The Federation Deployment Control Protocol

Presently, the Earth Observation community is demanding applications that provide low latency and high downlink capabilities. An increase in downlink contacts becomes essential to meet these new requirements. The Federated Satellite Systems concept addresses this demand by promoting satellite collaborations to share unused downlink opportunities. These collaborations are established opportunistically and temporarily, posing multiple technology challenges to be implemented in-orbit. This work contributes to the definition of the Federation Deployment Control Protocol which formalizes a mechanism to fairly establish and manage these collaborations by employing a negotiation process between the satellites. Moreover, this manuscript presents the results of a validation campaign of this protocol with three stratospheric balloons. In summary, more than 27 federations with 63.0% of throughput were established during the field campaign. Some of these federations were used to download data to the ground, and others were established to balance data storage between balloons. These federations allowed also the extension of the coverage of a ground station with a federation that relayed data through a balloon, and the achievement of a hybrid scenario with one balloon forwarding data from a ground device. The results demonstrate that the proposed protocol is functional and ready to be embedded in a CubeSat mission.

Automatic aircraft detection in very-high-resolution satellite imagery using a YOLOv3-based process

Aircraft detection in remote-sensing images is a fundamental task in civil and military applications. Deep learning techniques to achieve end-to-end object detection have attracted the attention of the Earth observation community. One of the primary factors behind the success of deep learning techniques is the utilized data. Several previous studies focused on designing the network infrastructure. Instead, this study pays more attention to the data. With the increasing number of available public datasets, whether directly employing a large number of instances with great variation will lead to a good performance has become a research topic. The ways in which these object instances are collected differ greatly. For example, the image sizes, object sizes in the training images, and geospatial resolution are varied. Therefore, herein, the factors influencing the detection performance, such as the object size, ground sampling distance, and Google zoom view, are investigated. A you-only-look-once-v3-based detection process is proposed for automatic aircraft detection. A nonmaximum suppression algorithm strategy is applied to filter unreliable and redundant bounding boxes detected in the overlapping image blocks. The model generalization ability under different training data combinations is evaluated in several challenging cases. The results prove that more variety of training instances from a greater variety of zoom levels will result in more false alarms. Instead, more variety in the object sizes under a constant zoom level is welcome. A large range of aircraft sizes (i.e., 7 to 77 m in length in this study) can be detected, with a promising F1 score of 0.98.

Earth observation remote sensing for oil and gas: A new era

In the last seven years, many earth observation satellites from national agencies and commercial providers have been launched, making huge volumes of data freely available to anyone. Open-access software and cloud computing tools also have been developed by the earth observation community. Those, as well as new sensors and vectors such as drone-borne hyperspectral cameras or gas-sensing systems are opening a number of applications for the oil and gas industry in exploration, production, and environmental monitoring.