Monitoring particulate matter (PM) is of critical importance due to its significant impact on human health. Ground stations provide highly accurate measurements of various pollutants on a local scale. However, the limited distribution of these stations makes achieving global coverage challenging. To address this limitation, satellite imagery serves as a valuable resource, offering wide-area PM estimates in near real-time through abundant data and frequent revisit intervals. In contrast to other studies, this work introduces deep learning (DL) models to estimate ground-level PM concentration maps over Europe. These models rely exclusively on radiance data from the Sentinel-5P satellite, forgoing auxiliary information, such as meteorological data, which are commonly incorporated in similar studies. The proposed approach has demonstrated both robust estimation accuracy and effective generalization capabilities. Furthermore, the estimated PM concentration maps have been validated against ground-based measurements, showing superior performance with respect to widely used models and datasets that consider meteorological inputs. The dataset and the code are available here:https://github.com/antoniomazza88/PMUnet.

This work deals with the estimation of the tropospheric vertical column density of nitrogen dioxide from Sentinel-5P radiance data using convolutional neural networks. The current processing chain to retrieve this information from Sentinel-5P data requires a complex, computationally demanding, physical modeling that involves the use of additional side information such as meteorological variables, which are not always available. Therefore, in this proof-of-concept study, we explored the feasibility of an estimation exclusively using radiance data from Sentinel-5P, leveraging on the powerful representational capacity of deep neural networks. Preliminary results are very promising encouraging further investigation.

The recent paradigm shift from model-based to data-driven approaches has involved a growing number of data-fusion tasks. Specifically for pansharpening, unsupervised deep learning methods have been recently explored with the goal of overcoming the generalization limits shown by early pansharpening convolutional neural networks based on supervised training schemes. Furthermore, some of these exploit the target-adaptive modality to face the scarcity of training data. On the downside, combining usupervised training and target adaptivity causes a non-negligible increase of the computational cost. This work presents a new target adaptive scheme that allows to keep limited the computational cost at inference time while preserving accuracy.

Optical remote sensing images are subject to cloud phenomena that can cause information loss in Earth observation. The main alternative is represented by the synthetic aperture radar images. However, many Earth monitoring applications exploit specific spectral features defined for multispectral data only. In this work, we propose a method that aims to recover several spectral features through deep learning-based data fusion of Sentinel-1 and Sentinel-2 time-series. The proposed approach has been experimentally validated for radiometric indexes such as the normalized difference vegetation index, the normalized difference water index, the soil-adjusted vegetation index and the atmospherically resistant vegetation index. Both numerical and visual results show that the proposed solution outperforms consistently the compared methods.

In this work is presented a new adversarial training framework for deep learning neural networks for super-resolution of Sentinel 2 images, exploiting the data fusion techniques on 10 and 20 meters bands. The proposed scheme is fully convolutional and tries to answer the need for generalization in scale, producing realistic and detailed accurate images. Furthermore, the presence of a$\mathcal{L}_{1}$loss limits the instability of GAN training, limiting possible problems of spectral dis-tortion. In our preliminary experiments, the GAN training scheme has shown comparable results in comparison with the baseline approach.

In recent years, there has been a growing interest in deep learning-based pansharpening. Thus far, research has mainly focused on architectures. Nonetheless, model training is an equally important issue. A first problem is the absence of ground truths, unavoidable in pansharpening. This is often addressed by training networks in a reduced-resolution domain and using the original data as ground truth, relying on an implicit scale invariance assumption. However, on full-resolution images, results are often disappointing, suggesting such invariance not to hold. A further problem is the scarcity of training data, which causes a limited generalization ability and a poor performance on off-training-test images. In this article, we propose a full-resolution training framework for deep learning-based pansharpening. The framework is fully general and can be used for any deep learning-based pansharpening model. Training takes place in the high-resolution domain, relying only on the original data, thus avoiding any loss of information. To ensure spectral and spatial fidelity, a suitable two-component loss is defined. The spectral component enforces consistency between the pansharpened output and the low-resolution multispectral input. The spatial component, computed at high resolution, maximizes the local correlation between each pansharpened band and the panchromatic input. At testing time, the target-adaptive operating modality is adopted, achieving good generalization with a limited computational overhead. Experiments carried out on WorldView-3, WorldView-2, and GeoEye-1 images show that methods trained with the proposed framework guarantee a pretty good performance in terms of both full-resolution numerical indexes and visual quality.

Cloud segmentation of remotely sensed multispectral images is an important topic not only for weather forecast but, more in general, for establishing when the sensed data actually relate to the soil so that can be reliably used for some monitoring purpose. In this work, leveraging on the capability of convolutional neural networks to accurately approximate complex relationships between raw data and higher-level products, we propose a U-Net-like solution conceived for Sentinel-2 images. In order to face the scarsity of training data, a proper domain adaptation strategy has been pursued, which resorts to a labeled Landsat-8 dataset. Preliminary results show a consistent improvement over standard tools.

The rise of deep learning has impacted profoundly all aspects of image processing and remote sensing. Following this trend, in the last few years, a large number of data-driven methods have been proposed also for SAR image despeckling. However, in spite of this large effort, only limited performance gains have been observed. We believe this is mostly due to the use of training sets that are only partially fit to the task, and sometimes plain wrong. In this work we assess experimentally the impact of training set design on the performance of SAR image despeckling with the goal of highlighting solid guidelines for sensible training.

In the last few years, deep learning (DL) has been successfully and massively employed in computer vision for discriminative tasks, such as image classification or object detection. This kind of problems are core to many remote sensing (RS) applications as well, though with domain-specific peculiarities. Therefore, there is a growing interest on the use of DL methods for RS tasks. Here, we consider the forest/non-forest classification problem with TanDEM-X data, and test two state-of-the-art DL models, suitably adapting them to the specific task. Our experiments confirm the great potential of DL methods for RS applications.

Sentinel-2 data represent a rich source of information for the community due to the free access and to the temporal-spatial coverage assured. However, some of the spectral bands are sensed at reduced resolution due to a compromise between technological limitations and Copernicus program's objectives. For this reason in this work we present a new super-resolution method based on Convolutional Neural Networks (CNNs) to rise the resolution of the short wave infra-red (SWIR) band from 20 to 10 meters, that is the highest resolution provided. This is accomplished by fusing the target band with the finer-resolution ones. The proposed solution compares favourably against several alternative methods according to different quality indexes. In addition we have also tested the use of the super-resolved band from an applicative perspective by detecting water basins through the Modified Normalized Difference Water Index (MNDWI).