Hyperspectral sensors are increasing in popularity for Earth Observation applications due to their ability to gather data over multiple spectral bands. However, the processing of such amount of information is difficult to handle for the current computing capabilities of small satellites. Several Band Selection methodologies have been developed in the last years; although, some of them demand very low computational resources, they use, at least, Level 1 data products. Therefore, the Level 0 data needs to be processed and the spectral bands coregistered. Artificial Intelligence has shown its potential to reduce the computational burden while achieving high accuracies in EO applications. In this study, a Neural Network-based methodology is proposed to select a spectral band set directly using non coregistered data captured by hyperspectral sensors.

Today hyperspectral imagers are capable to capture hundreds of spectral bands simultaneously, while they are getting smaller, fitting in CubeSats. The huge amount of information provided by these sensors is used for numerous studies, and it can be used to infer multiple variables. However, the huge data volume generated is difficult to be managed by large satellites, and it is almost impossible in the smaller ones. This work presents a low resource consuming unsupervised band selection algorithm that uses only the information of the captured image, and no further auxiliary information. It can be used to reduce the amount of data to be downloaded, while controlling the performance degradation. © 2022 IEEE.

FSSCat, the “Federated Satellite Systems/3Cat-5” mission was the winner of the 2017 ESA S^3 (Sentinel Small Satellite) Challenge and overall winner of the Copernicus Masters competition. FSSCat consists of two 6 unit cubesats carrying on board UPC's Flexible Microwave Payload - 2 (FMPL-2), an L-band microwave radiometer and GNSS-Reflectometer implemented in a software defined radio, and Cosine's HyperScout-2 visible and near infrared + thermal infrared hyperspectral imager, enhanced with PhiSat-1, a on board Artificial intelligence experiment for cloud detection. Both spacecrafts include optical and UHF inter-satellite links technology demonstrators, provided by Golbriak Space and UPC, respectively. This paper describes the mission, and the main scientific results of the FMPL-2 obtained during the first three months of the mission, notably the sea ice concentration and thickness, and the downscaled soil moisture products over the Northern hemisphere.

Knowledge about sea ice concentration and extent in polar regions is of great interest both for economic interests, and as a proxy of the climate change. Retrieved maps are based on data from microwave radiometers, which are currently provided by large satellite missions. Nowadays, CubeSats have proven to be a cost-effective alternative. Due to their low cost, they can be launched in large constellations to obtain high spatial coverage and daily revisit. This study presents a neural network approach to generate sea ice concentration and sea ice extension maps using the L-band microwave radiometer, and the GNSS-Reflectometer data from the FMPL-2 instrument onboard3Cat-5/A, one of the two CubeSats of the FSSCat mission. The results obtained during the first 2 months of the mission are presented.

This work presents the first scientific results over land from the Flexible Microwave Payload −2 (FMPL-2), onboard the FSSCat mission. FMPL-2 is composed of an L-band microwave radiometer and a Global Navigation Satellite System - Reflectometer (GNSS-R). Two separate ANNs models are trained using the first three months of collected data of both observations, with the objective to retrieve global soil moisture maps. The first network addresses the coarsely-resolved FMPL-2 antenna footprint in a downscaling approach. Predicted values resulted in good agreement with those obtain from the SMAP mission, with an error smaller than 9.6%, and a bias smaller than 0.001 m3/m3. The second network is implemented to estimate soil moisture exclusively on GNSS-R data. In this second case, the combination of multiple GNSS-R measurements in a single track allows to retrieve soil moisture data with an error standard deviation with respect to SMAP lower than 0.056 m3/m3, with a bias smaller than 0.0007 m3/m3.

Attitude determination and control systems (ADCS) are critical in many satellite missions, especially when high pointing accuracy is needed, such as in optical imagers. Current actuators used to control the platform's attitude, such as reaction wheels or magnetorquers, have some limitations, inducing some jitter, which is crucial in small satellites. This jitter induces the image blurring. This work analyzes the impact of actuator's jitter, and then a deblurring technique to minimize and correct the distortion produced is studied. Two points are addressed to restore the blurred images. First, the data coming from the inertial sensors embarked on the satellite to infer the movement of the camera is used. Secondly, a deblurring technique for non-constant blurring over an image is applied.

Global Navigation Satellite System Reflectometry and L-band microwave radiometry have been used for soil moisture, biomass, and cryosphere studies. Combining both technologies in a low-power and cost-effective solution could largely improve current Earth observations.3Cat-4 mission is a 1-Unit CubeSat technology demonstrator of the Flexible Microwave Payload - 1, a reduced size payload that combines these two technologies. This work presents the objectives of the3Cat-4 mission and the details of the spacecraft architecture and performance. Each spacecraft subsystem is detailed at hardware and software levels. In addition, the presented results indicate that the current spacecraft design will survive the flight conditions (i.e. thermal and structural ones).

Spaceborne constellations composed by several heterogeneous platforms are an actual solution to undertake Earth Observation missions. However, designing such missions present problems due to the heterogeneity and the multiple design levels that must be considered. In this paper, a high-level methodology to address the design of a spaceborne Earth Observation constellation is outlined. In addition to the framework, this document presents an application of this methodology to a specific use-case, the Agriculture Hydric Stress for a worldwide coverage, and describes its results.