With the rapid development of remote sensing technology, efficiently extracting useful information from large-scale remote sensing image data has become one of the core challenges in the field. However, traditional target recognition methods face difficulties when dealing with complex backgrounds, dynamic scenes, multi-angle targets, and so on.

This study proposes an innovative approach based on topological data analysis (TDA), introducing persistent homology and mapper analysis to model and analyze infrared remote sensing images at multiple scales. The remote sensing image is mapped to a point cloud structure in a high-dimensional topological space, and then persistent homology analysis or mapper analysis is used to extract the simplified topological structure of the point cloud data, forming a persistence diagram or a complex network structure. The network exhibits strong global properties, robustness, and excellent visualization characteristics, and can effectively capture the geometric and topological structures in the data while highlighting subtle differences. In previous experiments conducted on public datasets for airplane and automobile classification, the proposed method achieved a 1.5% improvement in accuracy compared to using convolutional neural networks (CNNs). And for ship classification, the method successfully categorized four types of ships, which shows strong applicability in target recognition and classification tasks.

The rotational invariance of topological data analysis effectively addresses the challenges in target recognition caused by changes in observation angles, thereby improving recognition and classification accuracy. It provides a new approach for infrared target recognition and classification， and also shows potential in the large-scale processing and pattern recognition of remote sensing data. Additionally, the method holds significant importance and practical value in fields such as military applications, urban management, climate change research, and disease detection.

Andean communities in central Peru play a key role in the conservation of vicuñas (Vicugna vicugna), a protected species that depends on puna grass and flooded vegetation for food and access to water throughout the year. This study focuses on seven communities of Lucanas in Ayacucho, a dry mountainous region of Peru, emphasizing the need for accurate information to monitor resources in a context of climate change and support community decision-making. In this research, based on Google Earth Engine (GEE), we evaluated the performance of classification algorithms using Sentinel-1 (S1) and Sentinel-2 (S2) image data for rangeland classification. The process used ground-based and image-based points to train and validate the models, a filter to minimize spatial autocorrelation between training and validation sets, and spectral separability measurements using the Jeffries–Matusita (JM) distance. All of these steps allowed for adequate discrimination and representation of the classes. Additionally, we used 64 feature variables (including vegetation, texture, topographic, snow, water, mineral, and radar features) and applied Cloud Score+, a quality assessment (QA) processor for S2 image collection, to improve classification accuracy. The Random Forest (RF) algorithm achieved an overall accuracy (OA) of 92% and a Kappa coefficient of 0.908, outperforming the Support Vector Machine (SVM) algorithm, which obtained an OA of 90.9% and a Kappa coefficient of 0.895. The results show that, in the semi-captivity sectors, 1,777.5 hectares of puna grass and 319.1 hectares of flooded vegetation were identified, while in wild management areas, 5,431.1 hectares of puna grass and 843.8 hectares of flooded vegetation were recorded. These findings highlight the importance of integrating remote sensing tools and machine learning algorithms to generate key information in the management of natural resources in communities.

Short-rotation coppices (SRCs) play an increasingly important role in sustainable fast-growing wood biomass production, offering rapid returns while contributing to climate change mitigation and reducing pressure on natural forests. Traditional field measurements of tree parameters in SRC plantations are time-consuming and labour-intensive, creating a bottleneck in efficient plantation management. While various remote sensing technologies exist, UAV-mounted LiDAR systems offer unique advantages for SRC monitoring through high precision and operational flexibility, yet their application in SRC contexts remains understudied.

We explore the potential of UAV LiDAR for tree diameter at breast height (DBH) estimation in SRC plantations by systematically investigating point cloud feature extraction methods and comparing predictive models. Working on a 1-hectare plot at the Leibniz Institute for Agricultural Engineering and Bioeconomy in Potsdam, Germany, we deployed a RIEGL miniVUX-1UAV scanner mounted on a DJI M600 platform to collect high-density point cloud data (2227 pt/m²). Manual DBH measurements and geolocation data from 500 trees were collected to validate our estimation models. Our results demonstrate the most suitable LiDAR metric combinations and model architectures for tree diameter estimation and contribute to understanding the applicability of UAV LiDAR technology in rapid SRC inventory assessment.

Decarbonization strategies aim at increasing Renewable Energy Source (RES) capacity, including new photovoltaic (PV) systems. Utility-scale PV installations are often placed in agricultural areas, resulting in a reduction of agricultural land, and affecting the environment. To balance agricultural and energy policies, PV development should not limit agricultural purposes, allowing for sustainable exploitation under specific technological and environmental conditions, and particularly in areas of actual or potential abandonment.

Studying agricultural abandonment is a complex due to its multifaceted nature, lack of a clear definition, and challenges in acquiring cartographic data. This study introduces and compares two methodologies to identify abandoned agricultural areas, aiming to delineate macro-areas of potential abandonment and examine patterns in conversion to energy use, with a focus on Toscana, a region (NUTS 2) in central Italy which has experienced cropland reduction unrelated to urbanization.

The first simplified approach analyzes land cover changes from 2000 to 2018, while the second method provides a more detailed abandonment detection by means of medium spatial resolution satellite imagery from the Harmonized Landsat and Sentinel-2 dataset. A Random Forest classifier combined with Object-Based Image Analysis (OBIA) is applied to satellite data to map annual active/non-active croplands. Annual maps are then validated with a trajectory-based approach to detect agricultural land abandonment. This second methodology can help in providing spatially and timely estimates of abandoned agricultural areas to be recovered for energy purposes, and promote the sustainable growth of PV systems.

The expansion of photovoltaic (PV) installations is crucial for global energy transition, but detailed information on their spatial distribution remains scarce, posing challenges for effective energy planning. This study presents a methodology for the automatic recognition of ground-mounted PV systems in Italy, using semantic segmentation and Sentinel-2 10-meter-resolution RGB images. The proposed methodology aims at accurately detecting both locations and sizes of plants, estimating capacity and ensuring regular map updates, to support energy planning strategies.

The segmentation model, based on a U-Net architecture, is trained on a dataset from 2019 and tested on two distinct cases, involving different imagery dates and areas. We propose a multi-temporal approach, applying the model to a series of images captured throughout the year and aggregating outputs to create a PV detection probability map. Users can adjust probability thresholds to optimize accuracy: lower thresholds enhance Producer Accuracy, ensuring continuous area detection to estimate capacity, while higher thresholds improve User Accuracy by minimizing false positives. Post-processing methods, such as plastic-covered greenhouse filtering, help reduce detection errors. Nonetheless, model generalizability across diverse landscapes needs improvement, requiring retraining with images from various environmental contexts.

The Salar de Atacama is a high-altitude salt flat desert ecosystem. This unique environment is rich in lithium. However, the extraction of this resource poses potential risks to the delicate ecological balance of the region. To assess the long-term impact of mining and climate change on this fragile ecosystem, this study aims to quantify changes in vegetation cover and water bodies within the Salar de Atacama.

Using Sentinel-2 satellite imagery (Level 2A) from July 2015 to June 2024, the Normalized Difference Vegetation Index (NDVI) and Normalized Difference Water Index (NDWI) were calculated using the Google Earth Engine. These indices were classified in Python using thresholds of 0.2 and 0.3, respectively, for the assessment of temporal variability.

To validate the results, the data were compared with the water level and meteorological and vegetation monitoring data from the “Sociedad Química y Minera de Chile” mining company.

Vegetation growth appears to be more strongly correlated with temperature than precipitation, with a two-month lag. An indirect relationship was observed between vegetation patch area and surface water levels.

The natural vegetation patch on the east side of the salt flat exhibits distinct behavior compared to the cultivated areas near San Pedro de Atacama. The natural vegetation patch reaches a maximum NDVI of 0.3, while crops exhibit a wider range of NDVI values, from 0.2 to 0.8. Additionally, the natural vegetation experiences a five-month dormant period (June to November), whereas crops have a shorter two-month dormant period (June to August). Analysis of water level data in the east border of the salt flat revealed no correlation with the near vegetation.

Due to the relatively small size of the lagoons (maximum of 1 km²), monitoring salinity dynamics using satellite imagery alone is challenging. Future studies could incorporate radar imagery to enhance the analysis of surface soil moisture and ground deformation.

In recent decades, Interferometric Synthetic Aperture Radar (InSAR), using satellite imagery, has emerged as a powerful tool for ground-deformation monitoring. It brings an advanced, sustainable, and cost-effective solution for long-term monitoring projects on wide areas. Additionally, InSAR can provide historical measurement data, adding further context about past ground-deformation and motion patterns. When using high-resolution sensors, the measurements can achieve millimetric-range deformation precision (1-2 mm) and metric-range geolocation precision (1-2 m) with densities over tens of thousands of measurement points per km². In urban areas, measuring surface motion and ground stability is key, because vulnerability is high due to large population densities and complex land conditions and uses.

The aim of this work is to showcase the applications of InSAR in urban areas at different spatial scales derived using the ATLAS interferometric processing chain, developed by SIXENSE around the core software GAMMA. In this study, the technology will be detailed and several cities over the globe will be presented as examples. Use cases will be provided at three different scopes: the overall city, a district, and a single building. The results will show how a wider view of the metropolitan area allows for the exploration of general motion patterns at a large scale, leveraging the high density and wide coverage of measurements. At a district or neighbourhood scale, areas of active deformation can be identified and monitored, like groups of buildings and roads experiencing subsidence after tunnelling or nearby construction. For the analysis of singular buildings, deformation can be tracked by analysing a time series of point measurements, and surface changes can be identified by loss of phase coherence. Advanced analytics allow us to calculate acceleration, risk of failure, and provide an estimation of the collapse date of structures, which serves as an alert system to anticipate critical events.

Ground-based remote sensing techniques are crucial in earthquake geology, as they provide high-resolution data for analyzing fault dynamics and earthquake-induced landscape changes. Hyperspectral imagery, in particular, has proven valuable for improving the recording and analysis of paleoseismological trenches. This research aims to enhance workflows that integrate hyperspectral data with high-resolution 3D terrain models. The study focuses on refining earthquake chronologies by improving the mapping and detection of prehistoric earthquake event horizons and structural features in paleoseismological trenches, using the Alhama de Murcia fault in the Eastern Betics Cordillera, Spain as a pilot area—one of the most seismically active regions in the western Mediterranean. Three paleoseismological trenches were excavated and analyzed using logging, georeferencing, and traditional interpretation techniques. Hyperspectral data were captured with a SPECIM Aisa Fenix camera, covering visible to shortwave infrared spectral ranges. High-resolution terrain models with millimeter accuracy were generated using LiDAR and digital photogrammetry. These datasets were processed and corrected using open-source Python tools and integrated into point-cloud data for spectral and dimensional reduction analyses, enabling 3D semiautomatic outcrop mapping. We present the integration of hyperspectral and digital terrain data into point clouds for the three trenches, comparing traditional 2D field logging with 3D digital mapping. At least three paleoearthquake events over the past 34 ka were identified, consistent with previous studies. Additionally, newly identified features reveal further evidence of earthquake imprints. Structural features linked to these events’ surface deformation were detected, including details not visible to the naked eye. Specific image-band ratios also revealed variations in relative mineral abundance along deformation structures, offering new insights into fault damage-zone characterization and fault kinematics. This study highlights the value of hyperspectral data in paleoseismology by reducing uncertainties, validating findings, and revealing previously hidden features through advanced 3D digital mapping techniques.

The high resolution of the Synthetic Aperture Radar (SAR) imagery, in addition to its capability to see through clouds and rain, makes it a crucial remote sensing technique. However, SAR images are very sensitive to the radar parameters, the observation geometry and the scene characteristics. Therefore, SAR simulators capable of generating raw SAR signals under specific acquisition conditions are essential for developing and evaluating SAR systems. The work proposed in this communication aims to develop a methodology to generate raw SAR data and SAR images for a complex scene of interest. More specifically, complex scenes of interest with targets located on a rough surface are considered in the scope of this study. In this context, electromagnetic (EM) phenomena are complex, and a composite scattering field between the targets and the environment occurs and needs to be modelled. This consists of multiple reflections changing the radar signature of the elements of the scene and causing defocus in the SAR image. One of the objectives of the presented work is to use the generated SAR images to extract information about the coupling between the objects and the environment.

With this prospect, the entire radar acquisition chain is considered: the sensor parameters, the atmospheric attenuation, the interactions between the incident EM field and the scene, and the SAR image formation.

Finally, simulations are performed considering different scenarios with a rough dielectric soil and canonical targets considered as Perfect Electric Conductors (PECs) to generate the raw SAR signal and the SAR image with the stripmap and spotlight acquisition modes. Thus, the impact of the different parameters of the radar link on the quality of the reconstructed SAR image is analyzed by distinguishing the parameters of the radar, the observed surface and the objects present on the surface.

Phenological parameters extracted from the time series (TS) of spectral indices are crucial for the characterization of crops. The accuracy of these parameters can be affected by missing data in the TS. The Copernicus Land Monitoring Service (CLMS) provides these parameters and information on their quality in the period of occurrence. The aim of this work was to assess, for extensive herbaceous crops, the impact of missingvegetation index time data on the phenological parameters provided by the CLMS at the time of occurrence. A methodology was proposed using the package developed in this research, TSGenerator, for the download, processing, and analysis of the data applied to evaluate the parameters at the start and end of the season and the day of maximum development (SOS, EOS, and MAX) between the years 2018 and 2023. It used 252 images of the BIOPAR-VI module products, 6 images of phenology parameters, and 2025 barley and maize plots from Monegros and Zaidín irrigated lands in Spain. In barley, SOS and MAX were the most affected, with an average of 42.9% and 40.9 % of missing data according to the Copernicus HR-VPP product evaluation parameter (±21 days). However, for maize, the most affected parameters were SOS and EOS, with 36.6% and 41.0 % of missing data. The correlation between the QFLAG-VPP quality parameter provided by Copernicus and the one proposed in this study, and the average percentage of non-missing data in the two most affected parameters, was r = 0.89 for barley and r = 0.74 for maize. The vegetation index series for barley presented about a 50 % probability of missing data in SOS and MAX according to a generalized additive fit model. A similar percentage was given for maize at SOS and EOS time points. This work represents an advance in the knowledge of the effect of missing data at the specific times of SOS, EOS, and MAX.