Climate change is associated with a variety of environmental phenomena, such as the melting of glaciers, the drying out of crops and soils, and an increase in the risk of avalanches. To detect and monitor these changes, satellite-based radar imaging is widely used, as it enables the large-scale mapping of the Earth's surface independent of weather and sunlight. An emerging field of research is to use radar systems mounted on unmanned aerial vehicles (UAVs) for this purpose. UAV-borne radar systems image smaller areas, but can capture them in greater detail, more flexibly in terms of flight trajectory, and at shorter repetition intervals. This article presents the capabilities of UAV-based radar systems and the associated challenges in terms of system design and imaging techniques for detecting and monitoring the effects of climate change. Current research results are shown. In addition, we give an outlook on using a swarm of cooperative UAVs, each carrying a radar. UAV swarms will enable a higher imaging quality with respect to resolution and detectability, but at the same time come with a number of additional challenges, which will be discussed in the paper as well.

Synthetic aperture radar interferometry (InSAR) is a technique that exploits the phase difference between two synthetic aperture radar (SAR) images to generate digital elevation models (DEMs) of the imaged terrain. InSAR has a well-established theoretical background, but approximations for narrowband signals are often made. To unwrap the interferometric phase, dual-baseline techniques are also commonly employed. Modern InSAR systems will use wider bandwidths to improve the accuracy and resolution of the DEMs. This letter generalizes the expressions of the baseline decorrelation and the critical baseline for InSAR systems with wide bandwidths and large baselines, where the spectral shrinkage is not negligible. It is also shown that radargrammetry offers high potential to unwrap the interferometric phase already on a pixel-by-pixel basis. The proposed approach can also exploit acquisitions with multiple baselines or frequency bands. The results show that a single wideband interferometric acquisition may suffice to perform phase unwrapping for high interferometric coherences. The concepts discussed in this paper will help to design and enhance the performance of future wideband and multi-band InSAR missions to deliver a new generation of DEMs with unprecedented resolution and accuracy.

Synthetic aperture radar (SAR) interferometry (InSAR) is a remote sensing technique that allows the generation of digital elevation models (DEMs) by combining two complex SAR images taken with a certain across-track separation. DEMs are of paramount importance for monitoring the Earth's surface and have traditionally been obtained using space- and airborne SAR systems, which are, however, characterized by stringent bandwidth regulations, constrained revisit times, and costly deployment and operation. Drone-borne InSAR enables cost effective surveillance of local areas with unprecedented accuracy and resolution through the exploitation of wider bandwidths and has therefore emerged as an attractive complementary system with short deployment times and flexible revisit intervals. This paper shows how to generate accurate, high-resolution DEMs through multiple acquisitions performed by a radar with ultrawide fractional bandwidth onboard a drone. We discuss the impact of the characteristics of the system and the acquisition parameters on the quality of the resulting DEMs. We also present a SAR processing scheme based on the omega-k algorithm, which allows fast processing of the drone-based raw SAR data and still yields comparable quality as back-projection algorithms, and an InSAR algorithm that takes into consideration the wide bandwidth of the system and exploits it to support phase unwrapping using radargrammetry. An experimental demonstration is conducted showing that DEMs with a height accuracy of a decimeter can be obtained at an independent posting of 25 cm × 25 cm. The system presented in this paper enables the generation of very high-quality DEMs with very short temporal baselines and also represents an important step towards the realization of single-pass distributed drone-based InSAR systems in the future.

Across-track synthetic aperture radar (SAR) interferometry is a well-established technique to obtain the topographic information by using a pair of SAR images, whose coherence influences the topographic height accuracy. Volume decorrelation occurs in presence of semi-transparent media, such as forests and vegetation, and is usually modelled using the vertical scatting profile, thus providing a link between the coherence and the physical parameters of the medium. This has been successfully exploited in polarimetric SAR interferometry and SAR tomography to retrieve three-dimensional structure information of the scene. Future SAR systems are expected to be equipped with larger bandwidths, which also allow for interferometric acquisitions with larger baselines and enhanced height accuracies. In these cases, a significant amount of co-registration decorrelation may occur within the volume, which should be considered in the model of volume decorrelation. This letter presents a novel volume decorrelation model, which accounts for co-registration effects and accurately predicts the coherence for interferometric acquisitions with large bandwidths and/or baselines. Simulation examples are shown for typical scenarios to prove the validity of the model, which is capable of predicting the phase of the complex coherence of a semi-transparent medium with sub-degree accuracy. The proposed model will be a crucial tool to analyze semi-transparent media in future spaceborne SAR missions and drone-borne SAR systems.

The proliferation of space debris in Low Earth Orbit (LEO) poses a growing threat to operational satellites, requiring robust surveillance and tracking systems. Radar is a primary technology for monitoring these objects; however, standard tracking algorithms often degrade when measurements are sparse, corrupted by non-Gaussian noise, or available only as short tracklets. Traditional methods, such as the Unscented Kalman Filter (UKF), rely on explicit physical and statistical models that may struggle to converge under highly nonlinear dynamics, uncertain initialization, and non-ideal sensor perturbations. In this study, we propose a hybrid deep learning framework for learned one-step tracklet filtering of radar measurements. The architecture consists of a stateful Gated Recurrent Unit (GRU) layer followed by one-dimensional Convolutional Neural Network (1D-CNN) layers, complemented by variable-specific preprocessing strategies, including residual learning for range and relative pivoting for azimuth, to handle the scale disparities and heterogeneous behavior of radar observables. This design combines the ability of GRUs to model temporal dependencies with the effectiveness of CNNs in extracting local features for signal denoising. The method is validated on synthetically generated LEO trajectories with realistic orbital perturbations and tunable radar noise profiles, including Gaussian noise, impulsive spikes, transient degradation, and state-dependent perturbations. Compared with EKF- and UKF-based analytical baselines, the proposed model achieves lower filtering error and improved robustness under severe non-Gaussian disturbances. Stress testing further shows that the network can reject non-physical sensor anomalies without requiring long initialization warm-up phases, making it suitable for sparse short-tracklet processing in synthetic SST benchmark scenarios.