We present a proof-of-concept onboard cloud detection system for the PRISMA Second Generation (PSG) mission, which combines a secondary forward-looking RGB camera with deep learning (DL) models on a system-on-a-chip (SoC) field-programmable gate array (FPGA). The proposed system enables real-time cloud coverage assessment, optimizing primary hyperspectral payload data collection and supporting adaptive acquisition scheduling, offering an effective solution for enhancing onboard data processing in Earth Observation (EO) missions. To support efficient model development, we derived the design specifications of the secondary camera and constructed a dedicated dataset using astronaut-captured images from the International Space Station – the CloudISS-RGB dataset – employing a self-training approach to generate high-quality pseudo-labels. Through an extensive architectural design exploration and systematic optimization, we tested two fully convolutional networks: U-Net, offering higher segmentation accuracy, and a lightweight convolutional autoencoder (CAE) designed for lower latency inference. The models were deployed on an AMD/Xilinx Zynq UltraScale+ MPSoC using the deep learning processing unit (DPU) IP core for hardware acceleration. The FPGA-deployed U-Net achieved 98.16% with a false positive rate of 0.9%, providing robust segmentation even in challenging conditions, making it the preferred model for reliable inference onboard PSG. The CAE model maintained 97.02% accuracy while achieving over 2× faster inference (31.81 ms vs. 74.29 ms per image). Both models enable accurate cloud segmentation with real-time inference, meeting the operational constraints derived from the secondary camera design, while operating at an average power consumption of 2.6 W (U-Net) and 2.4 W (CAE), well within the mission power constraint for the HW accelerator. Our results validate the feasibility of integrating DL-based cloud coverage assessment onboard PSG, contributing to the broader effort of advancing artificial intelligence-powered computing for EO missions to enable more autonomous data processing and decision-making.

A significant amount of remotely sensed data is generated daily by many Earth observation (EO) spaceborne and airborne sensors over different countries of our planet. Different applications use those data, such as natural hazard monitoring, global climate change, urban planning, and more. Many challenges are brought by the use of these big data in the context of remote sensing (RS) applications. In recent years, the employment of machine learning (ML) and deep learning (DL)-based algorithms has allowed a more efficient use of these data, but the issues in managing, processing, and efficiently exploiting them have even increased as classical computers have reached their limits. This article highlights a significant shift toward leveraging quantum computing (QC) techniques in processing large volumes of RS data. The proposed Quanv4EO framework introduces a quanvolution method for (pre)processing multidimensional EO data. Its effectiveness was first demonstrated on standard image classification datasets (MNIST and FashionMNIST), achieving accuracies of 99.84% and 96.81%, respectively, with a significantly reduced model size of 42 k parameters and 16 frozen qubits. Its capabilities were then checked on EO datasets, such as EuroSAT, with a mean accuracy of 96% using balanced iterative reducing and clustering using hierarchies (BIRCHs) clustering and 93% using automated DL (AutoDL), surpassing or matching state-of-the-art (SOTA) classical nonquantum models. Applying the framework to synthetic aperture radar (SAR) data, the QSPeckleFilter demonstrates notable improvements in speckle noise reduction, achieving a peak signal-to-noise ratio (PSNR) of 21.72 and a structural similarity index measure (SSIM) of 0.81, surpassing all tested classical counterparts. The proposed results underscore the potential of quantum-enhanced approaches in RS data analysis, paving the way for more efficient and effective solutions for wide geographical area EO data exploitation.

Accurate fuel mapping is vital for wildfire risk assessment and management. This study combines Remote Sensing (RS) data and Machine Learning (ML) to differentiate fire behavior fuel models. Three ML approaches - Random Forest (RF), Support Vector Machine (SVM), and Convolutional Neural Network (CNN) - are compared in terms of accuracy, recall, and F1 score. Employing Sentinel-2 imagery, the ML-based classification accurately categorizes fuel types into eight main classes: broadleaf, conifer, shrub, grass, bare soil, burned area, urban area, and water. Past results in a Sardinia test case were promising, with CNN achieving impressive metrics - accuracy, recall, and F1 score - each at 99%. Notably, the network exhibits high validation score in identifying classes in unseen pixels: broadleaf at 99%, conifer at 79%, shrub at 76%, and grass at 84%. Subclasses, aligned with the Standard Scott and Burgan (2005) system, were derived from the eight classes using Above Ground Biomass (AGB) and Bio-Climatic (BC) maps to refine fuel mapping. A significant enhancement of this work involves testing the proposed method in Portugal and Greece to validate its robustness across diverse geographical regions. Furthermore, the final fuel type maps are rigorously validated by comparing them with FirEUrisk’s pre-existing validated fuel maps, providing a benchmark to assess the accuracy and reliability of the new maps.

This paper presents the initial findings of wheat mapping activities conducted in 2023 as part of the Afghanistan Emergency Food Security Project, led by the Food and Agriculture Organization (FAO) of the United Nations and supported by the World Bank. The study focuses on mapping irrigated wheat cultivation, analyzing data from 2017 to 2023 to assess the impact of recent drought years. Using Sentinel-2 imagery and a random forest classification model in the Google Earth Engine cloud computing platform, a pixel-based classification was performed. Results indicate a significant decrease in irrigated wheat acreage at national levels from 2017 to 2023.

Volcanic eruptions are a natural hazard that can devastate people and property. In recent years, the number of volcanic eruptions has been on the rise, and the effects of climate change are making them more frequent and more powerful. This research proposes a new methodology for monitoring volcanoes in real-time or very close to real-time using an intelligent Distributed Satellite System (iDSS). The iDSS is made up of a constellation of satellites that are all connected to one another by means of Inter-Satellite Links (ISL). This allows the data to be processed and distributed in real-time, which is essential for early warning of volcanic eruptions. In previous studies, the on-board volcanic eruption detection was proven to be possible and feasible by utilising appropriate Artificial Intelligence (AI) techniques. Multispectral optical data were used to assess if an active volcanic eruption was captured in the image. The proposed iDSS architecture is practical and can be used to monitor volcanoes in real-time or near real-time. The system has been tested by taking Mount Etna as a case study and the results have been reported. The findings and conclusions of this research can be applied to and expanded upon in the context of similar natural disasters occurring around the globe.

Climate action (SDG-13) is an integral part of the Sustainable Development Goals (SDGs) set by the United Nations (UN), and wildfire is one of the catastrophic events related to climate change. Large-scale forest fires have drastically increased in frequency and size in recent years in Australia and other nations. These wildfires endanger the forests and urban areas of the world, demolish vast amounts of property, and frequently result in fatalities. There is a requirement for real-time/near real-time catastrophic event monitoring of fires due to their growing frequency. In order to effectively monitor disaster events, it will be feasible to manage them in real time or near real time due to the advent of the Distributed Satellite System (DSS). This research examines the possible applicability of DSS for wildfire surveillance. For spacecraft to continually monitor the dynamically changing environment, satellite missions must have broad coverage and revisit intervals that DSS can fulfill. A feasibility analysis, as well as a model and scenario prototype for a satellite artificial intelligence (AI) system, is included in this letter to enable prompt action and swiftly provide alerts. In our previous research, it is shown that on- board implementation, i.e., data processing utilizing hardware accelerators, is feasible. To enable Trusted Autonomous Satellite Operation (TASO), the same will be included in the proposed DSS architecture, and the outcomes will be provided. To demonstrate the applicability, the suggested DSS architecture will be tested in several geographic locations to demonstrate the system-wide coverage.

Crop type mapping represents one of the most challenging problems in remote sensing. Spatial, spectral, and temporal information are required in order to obtain a unambiguous distinction among the types of crop. This paper presents a multi-sensor approach, where labelled high-resolution images from drones, limited to small areas, are used to enhance the classification ability of machine learning models based on Sentinel 2 time series. The project described in this paper is organized into three major activities. The first part focused on the exploitation of RGB drone images by using transfer learning and convolutional networks, and it has already been described in a previous work by the team. The second part deals with preliminary analysis of multi-spectral Sentinel 2 time-series using the labelled data from the drones campaign and trees-based machine learning algorithms. Finally, the third ongoing part deals with the combination of drones and satellite data in order to show how drones data can help the Sentinel 2 classification by reducing the effort needed to collect reference crop type information.

Over the last few years, crop type mapping has gained importance in remote sensing as it represents one of the most challenging problems in this field. Precise and continuous spectral signatures can significantly help to obtain a unambiguous distinction among the types of crop. This paper presents a discussion about the application of different types of hyperspectral imagery to crop-type mapping. This project is part of a collaboration among the Italian Space Agency (ASI), the Φ-lab in the ESRIN centre of the European Space Agency (ESA), and the Italian National Research Council (CNR). This works is mainly focused on the analysis of the PRISMA hyperspectral images, comparing them to airborn imagery from the Compact Airborne Spectrographic Imager (CASI) and short-wave infrared (SWIR) Airborne Spectrographic Imager (SASI). The continuous spectral signature over the SWIR and the visible and near-infrared (VNIR) channels will be used to perform a binary classification by means of a one-dimensional convolutional neural network. The test case with 3 tomato fields and 4 corn fields is sited near Gros-seto, in Tuscany, Italy. Results will show the potentialities offered by the PRISMA mission for remote sensing applications.

This article aims to explore the potential of current approaches for quantum image classification in the context of remote sensing. After a brief outline of quantum computers and an analysis of the current bottlenecks, it shows for the first time experiments with quantum neural networks on a reference Earth observation (EO) dataset: EuroSAT. Moreover, it establishes the proof of concept of quantum computing for EO: the models trained and run on a quantum simulator are on par with classical ones. We make the open-source code available for further developments11QNN4EO repository: https://github.com/ESA-PhiLab/QNN4EO..