This study presents an assessment of wildfire severity and its associated landslide risk on Ischia Island, Italy, integrating meteorological data, including precipitation and maximum temperature, with remote sensing datasets to assess the wildfire that occurred on 28thAugust 2023. Key findings include mapping wildfire extent using NASA FIRMS data and assessing severity through indices like Normalized Burn Ratio (NBR), differential NBR, Normalized Difference Vegetation Index (NDVI), and differential NDVI using Sentinel-2 images. High-resolution PlanetScope imagery enabled detailed change detection, while historical ground deformation data from Sentinel-1 (2018-2022) revealed significant susceptibility to landslides. The analysis indicated that the areas affected by the wildfire are also prone to landslides, with a mean deformation rate of -8 to -10 mm/year. Post-fire, the reduction in vegetation coverage and subsequent rainfall increased soil erosion and deformation, heightening the landslide risk. Strong correlations were observed between wildfire severity, land surface temperature (LST), and precipitation patterns. This integrated approach highlights the critical need for precise risk assessment and underscores the importance of multi-sensor data in post-fire landscape management and hazard mitigation.

Multi-modal Large Language Models (MLLMs) are currently an extremely active research topic for the multimedia and computer vision communities, and show a significant impact in visual analysis and text generation tasks. MLLM's are well-versed in integrated understanding, analysis of complex data from cross modalities (i.e. text-image) and text generation with chat abilities. Almost all MLLM's, focus on alignment of image features to textual features for downstream text generation tasks includes detailed image description, visual question answering, stories and poems generation, phrase grounding, etc.. However, when focusing on visual question answering, questions that are highly relevant to the context of an image may not be answered correctly with the existing MLLM's, contrary to questions that are related to visual aspects. Moreover, generating meta data (context) for an image using present day MLLM's is hard task due to hallucinating characteristic of underlying Large Language Models (LLM's), and adequate contextual information cannot be directly derived from an image based perspective.

In this paper, the mathematical derivation of the underlying probability distribution function for the normalized least-squares wavelet spectrogram is presented. The impact of empirical and statistical weights on the estimation of the spectral peaks and their significance are demonstrated from the statistical point of view both theoretically and practically. The simulation results show an improvement of approximately 0.02mm (RMSE) for annual signal estimation when statistical weights are considered in the least-squares wavelet analysis (LSWA). The weighted LSWA estimates the signals more accurately than the ordinary LSWA for different percentage amount of missing data. As a real-world application, Global Navigation Satellite Systems (GNSS) time series for a station in Rome, Italy are analyzed. The analyses of the GNSS time series provided by different agencies for the same station reveal statistically significant annual peaks, more significant in 2010 but less significant between 2018 and 2020, while the higher frequency components show different spectral patterns over time. A declining trend of approximately −0.42 mm/year since 2004 is estimated for the GNSS height time series, likely due to gradual land subsidence. The results not only highlight the advantages of LSWA but can also help to better understand the uncertainties involved in signal estimation.

Synthetic Aperture Radar (SAR) imagery is widely used for measuring Earth’s surface displacements over large areas. Persistent Scatterer Interferometry (PSI) [1] is a powerful A-DInSAR (Advanced Differential Synthetic Aperture Radar Interferometry) multitemporal technique that detects displacement measurements with sub-centimetric precision and monitors the temporal evolution of the processes. PSI data offer high spatial coverage and temporal repeatability, improving the understanding and evaluation of deformation processes for mitigating natural-related hazards. The capability to measure ground and structural deformation relies on the spatial pattern of the phenomenon and the available PSI density [2] , [3] . The spatial density provided by the PSI technique is related to the resolution of the considered sensor [4] and other site-specific characteristics (presence of vegetation or water bodies, foreshortening effects due to slopes, etc..). The wide availability of SAR satellite missions allows the development of a new data fusion approach for integrating multi-band SAR sensors (X, C, and L) to improve data coverage and information content. PSI data fusion combines multi-band products to exploit detailed and complementary information about the monitored surface.

We develop a mathematical algorithm that calculates discrete spectral derivative of the hyperspectral data of the satellite mission PRISMA in a Geographic Information System (GIS) environment. Vibrational fingerprits of minerals appear in the reflectance spectra as local "dips" centered at a resonance wavelength that characterizes some classes of minerals. The algorithm is then applied twice to each spectrum in each image datacube: once on the descending slope of a known vibrational feature, and once on the ascending slope. The sum of the absolute value of the two derivatives, obtained from a total of several tens of hyperspectral channels, is then combined in a vibrational line intensity score, which is then used to build mineral abundance maps. We apply the method to a mountain area, mapping freshly exposed gneiss under molten glaciers (vibrational resonance at 2222 nm), and to volcanic islands mapping rhyolite deposits on shores (resonance at 2104 nm).

This paper describes some of the data fusion methods developed by NHAZCA S.r.l. in the frame of the project “MUSAR”, funded by ASI, for the integration of data from multi-sensor/multiband satellite images. The aim of MUSAR is to extend the exploitation of EO data in the research area of natural hazard, with a specific focus on their interaction and interference with structures and infrastructures. The proposed methods are based on the post-processing of results achieved from InSAR and A-DInSAR analyses and by Photomonitoring techniques. Some preliminary results on real case studies are also presented.

Disaster Risk Financing (DRF) products strengthen the capacity of governments to take informed decisions on disaster risk finance based on sound actuarial exposure analysis in order to support stakeholders (national and local governments, homeowners, businesses, agricultural producers, and low-income populations) with better risk information and increase financial resilience. The objective of this work is to present the Landslide Hazard In formation System (LHIS) prototype aiming at responding to incipient likely landslide events (in Near-Real Time, NRT) and providing estimates of parameters suitable to inform parametric insurance calculations. LHIS concentrates on the development of databases, maps and understanding on pilot use cases in Morocco, making extensive use of EO data and advanced modelling capabilities.

A short survey of the use of neural networks and statistical discriminants in high energy physics for recognition of heavy flavor jets is presented. After illustrating the various neural and statistical classifiers currently used, some assessment of their comparative performance for top and bottom jets is made.