Hydrological drought is one of the most severe consequences of climate change, leading to a reduction in available water resources. This study presents a workflow utilizing SAR data from COSMO-SkyMed STRIPMAP imagery to monitor water levels and surface extents as an alternative to traditional gauge stations. The methodology involves a radiometric algorithm to segment water surfaces by analyzing pixel backscatter differences [1]. Image quality was enhanced using histogram equalization and bilateral filtering, while classification techniques were employed to segment water, land, and border areas. The workflow was applied to Albano Lake in Italy as a first case study. Validation with manually digitized reference masks showed high accuracy, with F1 scores of 0.997 for Otsu’s method and 0.996 for k-means clustering. A stereo-SAR technique was employed for water level estimation, leveraging DATE software processes [2] and ascending and descending images acquired close in time. These images were first projected in ground range onto a plane at mean elevation. The actual elevation was then determined by iteratively refining it until the correlation between the images was maximized. Although validated on a single case study, the workflow demonstrates significant potential for broader application to diverse water bodies and SAR datasets.

This research is performed in the framework of the GRAW project, funded by the Italian Space Agency (ASI), Agreement n. 2023-1-HB.0, as part of the ASI’s program “Innovation for Downstream Preparation for Science” (I4DP_SCIENCE).

[1] Li, J.; Ma, R.; Cao, Z.; Xue, K.; Xiong, J.; Hu, M.; Feng, X. Satellite Detection of Surface Water Extent: A Review of Methodology. Water 2022, 14, 1148.

[2] Di Rita, M., Nascetti, A., & Crespi, M. (2017). Open-source tool for DSMs generation from high resolution optical satellite imagery: Development and testing of an OSSIM plug-in. International Journal of Remote Sensing, 38(7), 1788–1808.

Inland water bodies are vital freshwater sources, requiring effective monitoring to assess climate change and human impact. Advances in remote sensing technologies now enable cost-effective, long-term surface-water-level tracking. This study refines continuous water-level time series using satellite altimetry data from the Global Ecosystem Dynamics Investigation (GEDI) and Surface Water and Ocean Topography (SWOT) missions, focusing on their integration to enhance accuracy, precision, and revisit frequency.

GEDI, a spaceborne LiDAR altimeter aboard the International Space Station, provides high-resolution measurements (25m footprint) between 51.6°N and 51.6°S. Data from March 2019 to March 2023, available on Google Earth Engine (GEE), were evaluated for Italian lakes (2019–2022)]. An outlier detection workflow, incorporating GEDI metadata and a 3NMAD-based test, improved measurement precision. For Northern Italian lakes with gauge data, the intrinsic precision was 0.11m, while GEDI achieved sub-10cm precision for smaller ungauged lakes in Lazio.

SWOT, operational since April 2023, uses a Ka-band Radar Interferometer to monitor 86% of Earth's surface with a 100m pixel size and a 21-day revisit time. Over Northern Italian and Swiss lakes, SWOT achieved a 92% correlation with gauge measurements and a precision of ~0.06m. For Central Italian ungauged lakes, spatial NMAD was under 10cm, with minimal outliers. The complementary integration of GEDI and SWOT data offers a robust solution for monitoring inland water levels globally.

Acknowledgments

This research is performed in the framework of the GRAW project, funded by the Italian Space Agency (ASI), Agreement n. 2023-1-HB.0, as part of the ASI’s program “Innovation for Downstream Preparation for Science” (I4DP_SCIENCE).

Very-high-resolution soil moisture estimation from remote-sensed SAR data has been under investigation in the literature. At this spatial scale, hypotheses that can be fulfilled at lower spatial scales are in fact not supported. For example, when analyzing agricultural areas at the field scale, it is essential to account for the seasonal evolution of vegetation and variations in soil roughness due to agricultural practices.
Similarly to what was previously conducted with Sentinel-1 C-band data [1], this work investigates the retrieval of superficial soil moisture from SAOCOM L-band data at the field scale. The L-band data are representative of a thicker layer of soil (20 cm) with respect to the C-band (6 cm) and are characterized by a spatial resolution of around 10 m. In situ soil moisture data on the study area are available, where the vegetation types are derived from the EUCROPMAP for the year 2022, and changes in roughness conditions are taken into account viaanomaly detection [2].

Acknowledgements
This research was performed under the framework of the GRAW project, funded by the Italian Space Agency (ASI), agreement no. 2023-1-HB.0, as part of the ASI’s program “Innovation for Downstream Preparation for Science” (I4DP_SCIENCE).

References
[1] Graldi, G.; Zardi, D.; Vitti, A. Retrieving Soil Moisture at the Field Scale from Sentinel-1 Data over a Semi-Arid Mediterranean Agricultural Area. Remote Sens. 2023, 15, 2997. https://doi.org/10.3390/rs15122997
[2] Zhu, L. et al. (2019). ‘Roughness and vegetation change detection: A pre-processing for soil moisture retrieval from multi-temporal SAR imagery’. In: Remote Sensing of Environment 225, pp. 93–106. DOI: 10.1016/j.rse.2019.02.027.

With the ever-increasing demand for efficient and accurate detection technology in the fields of mineral exploration and mining, the remote Laser-induced Breakdown Spectroscopy (LIBS) scanning imaging technology has demonstrated unique advantages and broad application prospects in the identification of ore types. Traditional analytical methods usually require a large amount of manpower, material resources, and time, and it is difficult to conduct rapid and comprehensive surveys of large areas. The remote LIBS technology can break through this limitation. By emitting high-energy laser beams, it can stimulate the surface of ores at a distance of several meters or even dozens of meters, causing the atoms or ions in the ores to undergo energy-level transitions and generate plasma emission spectra. By using spectrometers to collect and analyze these emission spectra, the characteristic spectral line information of various elements in the ores can be obtained, thus determining the composition of the ores. Moreover, by scanning the ores, the spectral data at different positions can also be collected to achieve scanning imaging of the ore elements, and then obtain the concentration distribution information of the sample elements, which provides a more intuitive, comprehensive, and accurate basis for the identification and evaluation of ores, and fervently encourages the mineral industry to forge ahead in the direction of intelligence and high efficiency.

Over the past three years, agricultural drought has repeatedly affected Italy, resulting in unpredictable yield losses. This phenomenon is influenced not only by meteorological factors but also by vegetation phenology and soil moisture conditions, making effective monitoring essential to assess its severity and impacts.

Hyperspectral satellite imagery offers a promising tool for monitoring drought effects on vegetation due to its sensitivity to stress-related parameters. However, its effectiveness is often limited by the availability of reliable in situ data for validation.

This study utilized PRISMA, the Hyperspectral Precursor of the Application Mission of the Italian Space Agency (ASI), to analyze spectral signatures of drought-affected crop fields in comparison with in situ data provided by the Institute of Services for Agricultural and Food Market (ISMEA). PRISMA offers high spectral (240 bands, 400–2500 nm) and spatial (30 m) resolution. The in situ dataset includes standardized estimates of drought damage percentages (0–100%) at the field scale.

First, a terrain-independent correction approach was applied to address PRISMA’s geolocation error. Then, median spectral signatures were extracted for fields with varying damage levels, excluding bands affected by atmospheric absorption. One case study focused on durum wheat fields in Foggia (Southern Italy) during the 2022 season. Two cloud-free PRISMA images (29/04/2022 and 14/06/2022) were analyzed, revealing differences in the 700–1100 nm spectral range between fields with 0% and 36% damage. These bands are associated with leaf structure and water content, parameters influenced by drought stress.

Despite the limited dataset, the results demonstrate PRISMA’s potential for agricultural drought monitoring. Future efforts will expand the dataset to enhance analysis reliability.

Hyperspectral image classification (HSI classification) aims to assign land cover categories to pixels based on their spectral characteristics and has wide applications in environmental monitoring, smart agriculture, military operations, and other fields. However, in practical scenarios, the data distribution of the source and the target domain may be significantly different, that is, domain shift. Unsupervised domain adaptation (UDA) for HSI classification attempts to leverage the representation ability of source-trained models on the target domain. Various factors can cause a domain shift in HSI images. For example, different sensor characteristics and environmental factors can cause discrepancies in the data distribution, which can lead to different characteristics of the same object in different domains, making it difficult for a source-trained model to generalize well on target domains. To address this issue, huge progress has been made in existing methods. However, there are still two main problems: Firstly, existing methods usually only utilize adversarial learning mechanisms for feature alignment, ignoring the effectiveness of combining multiple learning strategies. Secondly, adversarial learning still has limitations in mitigating the representation tendency of source domain samples. To solve these two problems, we propose a hybrid training-driven unsupervised domain adaptation network (HT-UDANet) for HSI classification to improve adaptation performance on the target domain. Specifically, we first incorporate a self-training mechanism alongside adversarial learning to improve the model's compatibility from two different perspectives. Then, we further design a module-separated strategy in the self-training mechanism. With this strategy, the optimization process becomes more flexible and the high-quality pseudo-labels on target images can better suppress any possible overfitting on annotated source samples. Extensive experiments on multiple HSI datasets, including Pavia University, Pavia Center, Houston2013, and Houston2018 demonstrate the effectiveness of our proposed HT-UDANet for UDA HSI classification. A comparison with existing methods shows that HT-UDANet performs better in terms of classification accuracy.

Urban areas are increasingly vulnerable to the impacts of climate change, including elevated temperatures caused by the urban heat island effect. This study evaluated the effectiveness of Nature-Based Solutions (NBSs) in mitigating urban surface temperatures in Guimarães, Portugal, a city recognized for its sustainable urban planning initiatives. Integrating remote sensing data, machine learning models, and spatiotemporal analysis, it examined the evolution of land use and land cover (LULC) and land surface temperature (LST) from 2013 to 2023, with predictions for 2028.

This study employed Landsat 8 imagery to derive LST, NDVI, and NDBI indices and classified LULC using the Random Forest algorithm. A progressive increase in vegetated areas from 154.76 km² in 2013 to 172.22 km² in 2023 reflects the influence of urban sustainability initiatives such as the "Guimarães Mais Floresta" program. Machine learning models, including XGBoost, Bagging, and AdaBoost, were used to predict LST for 2028. XGBoost outperformed others, achieving an R² of 0.9543 and providing precise predictions for urban planners.

The results highlight that NBS implementations, such as green roofs and urban gardens, reduced local temperatures by up to 2.49 °C. However, projections for 2028 indicate a slight reduction in vegetated areas, underscoring the need for stronger environmental policies. This study also identified thermal hotspots, predominantly in built-up areas, where temperatures are expected to exceed 37 °C in 2028. These findings emphasize the importance of targeted NBS interventions in urban planning to mitigate climate risks.

This research advances spatiotemporal methodologies by combining multitemporal remote sensing data, machine learning predictions, and local validation. The approach offers a replicable framework for assessing the effectiveness of NBSs and provides recommendations for sustainable urban development. Future studies could expand the temporal dataset with additional satellite imagery and test this methodology in cities with diverse climatic and urbanization patterns, offering broader insights into the effectiveness of NBSs.

Roof material classification is a key task for applications such as urban heat island studies, energy modeling, and resource management. This study focuses on leveraging machine learning (ML) algorithms to achieve accurate roof material classification, despite the challenges posed by noisy training data. BD TOPO, a widely used database providing roof material information at the building footprint level, is known for its inaccuracies and incompleteness. Recognizing the potential of ML models to produce reliable classifications even when trained on imperfect data, we designed a framework to improve upon BD TOPO’s baseline accuracy.

The methodology includes advanced preprocessing steps to enhance input data quality. For each building footprint in BD TOPO, statistical metrics—mean, median, and standard deviation—were extracted from multi-seasonal Pleiades imagery (winter and summer). These features were corrected for the satellite's off-nadir acquisition angle to ensure alignment with building boundaries. Additional spectral indices (e.g., NDVI and GNDVI) and band ratios were derived to capture material-specific properties and seasonal variability.

To address the inherent noise in the BD TOPO labels, we employed XGBoost and Random Forest, two algorithms known for their robustness, to label inaccuracies. Our hypothesis is that training these models on noisy yet extensive datasets will result in classifications that are more accurate and consistent than the original BD TOPO labels.

While results are pending, we expect this approach to effectively distinguish roof materials and enhance the accuracy of urban mapping. The proposed framework demonstrates the potential of combining multi-seasonal data and ML techniques for scalable and reliable roof material classification, with broader applications in sustainability and urban planning

Rural landscapes, shaped by centuries of interaction between humans and nature, represent vital cultural heritage and are among the most significant expressions of a nation’s identity.

In the Mediterranean, olive groves and rural areas contribute significantly to biodiversity, soil protection, and ecosystem resilience. However, these landscapes face threats from the abandonment of traditional farming practices and the intensification of agriculture. To address these challenges, land cover classification and change detection techniques offer valuable tools for assessing, preserving, and restoring olive groves.

This study focuses on identifying rural landscapes of historical significance in the Lazio region (Italy) and assessing their conservation status. Specifically, the olive landscape of Cures, located in the historic province of Sabina, was analysed using a multi-temporal approach that combined historical literature and cartographic data, including orthophotos from 1954.

The research employed the VASA (Historical Environmental Assessment) methodology, a technique designed to evaluate the evolution of landscapes, providing insights into changes in agricultural practices and land use. Software and tools such as Collect Earth and Google Earth were utilised to examine high-resolution satellite images from different periods, allowing for the photointerpretation of land-use changes. The spatial distribution of identified land units was estimated to capture the landscape’s configuration.

By overlaying land-use polygons from 1954 and 2022, the study created a merged database that revealed the evolutionary dynamics of the landscape. This analysis identified key elements of the Cures landscape, highlighted areas at risk of degradation, and pinpointed landscape emergencies—land uses that have experienced significant reductions in area. The findings underscore the importance of these methodologies for enhancing the understanding of rural landscapes, providing essential data to support sustainable agricultural practices and the preservation of cultural traditions. By implementing these strategies, it is possible to better manage and protect these landscapes, ensuring the environmental balance of agricultural lands is maintained.

As a hazardous heavy metal element, cadmium can migrate in the soil and be absorbed by crops, which may seriously jeopardize human kidneys and other systems through the food chain, posing a great threat to human health. In recent years, with the rapid development of industry and the frequent use of agrochemicals, soil heavy metal pollution has become increasingly serious, forcing the degradation of arable land and affecting the safety of food quality. Therefore, rapid, real-time, and accurate monitoring of Cd contamination in soil is crucial. In this regard, we propose a dual-pulse-based remote laser-induced breakdown spectroscopy (LIBS ) combined with chemometrics for the quantitative detection of soil Cd content, which can be used for remote elemental analysis at a distance of 50 meters.

In this study, in order to enhance the intensity and sensitivity of the obtained spectra, a high-energy double-pulse laser was used in our experiments, and the spectral signals of Cd were significantly enhanced by optimizing the pulse interval, spectrograph delay, and integration time for the best matching parameter. To obtain accurate quantitative analysis results, feature spectral screening algorithms such as LAR, PCA, and SPA combined with PLS, SVM, ANN, and CNN models were tested to establish several quantitative analysis models to realize the accurate prediction of Cd elements in soil. The best correlation coefficient for the prediction set reached R²=97.75%.

In this work, the horizontal and vertical migration patterns of Cd in soil were revealed by the quantitative detection of Cd, which provides a reliable solution for the detection of cadmium pollution in soil at a long distance and large scale. This technique provides a valuable method for the detection of heavy metals in soil in areas with serious heavy metal pollution such as smelters and landfills.