Search

Important changes in the Earth’s water cycle can be assessed by analysing sea surface salinity, as this variable on average reflects the balance between precipitation and evaporation over ocean, being the upper layers the most sensitive to atmosphere-ocean interactions. In situ measurements of salinity are relatively scarce, reduced to a limited number of field campaigns, buoys, and drifters, and typically acquired some meters below the sea surface. Thus, they may not necessarily well represent ocean-atmosphere exchanges. Satellite measurements, on the contrary, are synoptic, repetitive and represent the uppermost surface of the ocean. In this work, we show that the dynamics captured by satellite-derived sea surface salinity (SSS) measurements differ from the dynamics shown by in situ near SSS (NSS) measurements. We compare a temporal series of 8 years of the Soil Moisture and Ocean Salinity (SMOS) SSS maps with the output of an ocean model that assimilates in situ salinity measurements, which includes salinity in the first meters of the surface (NSS), as well as the mixed layer depth (MLD) and the sea surface temperature (SST). On the one hand, the satellite SSS measurements present a clear intensification of the water cycle which is somewhat less evident in the NSS. The water cycle is expected to intensify in the context of global warming, according to the Clausius-Clapeyron (CC) relation, which states that the saturation of the water vapor pressure increases at a rate of 7% per degree Celsius of warming. During the analysed 8 years, we observe a positive SST trend ranging between 0.2ºC/year and 0.1ºC/year (depending on the region) and a SSS trend ranging between 0.008 psu/year and 0.015 psu/year, which is consistent with the CC law. On the other hand, we observe that the largest positive differences between the satellite SSS and the NSS trends are in regions that simultaneously present a large positive SST trend, and a negative MLD trend. This suggests that globa

The Soil Moisture and Ocean Salinity mission (SMOS) is the European Space Agency (ESA) second Earth Explorer. It was launched on 2nd November 2009, and it continues to provide L-band Brightness Temperature (BT) measurements from which a number of applications are derived. More notoriously, soil moisture measurements, sea surface salinity, sea-ice thickness and high wind speeds. Its payload, the Microwave Imaging Radiometer with Aperture Synthesis (MIRAS) continues to be the first and only 2D radiometer interferometer ever flown to space for Earth Observation. 12 years after launch, SMOS is still in good health, and the SMOS calibration team continues to further improve the quality of the data. Most recently, the 3rd Mission Reprocessing level 1 data set was released, after a number of improvements performed in calibration and the image reconstruction process. Among the changes introduced in calibration for the 3rd Mission Reprocessing release, there was the change in strategy for the Noise Injection Radiometer calibration, that became fixed after the team realised that this particular receiver is far more stable than what can be measured in calibration; the update of the Power Measurement System (PMS) thermal sensitivity and NIR antenna losses characterization values, and the introduction of a thermal latency parameter for the thermal sensor in the NIR antenna. Among the changes in the image reconstruction process, the most relevant were the refinement of the Gibbs correction (called Gibbs-2) to account for differences in the Sea and Land BT, the introduction of the super-sampled Sun BT correction, to account for inhomogeneities of the L-band Sun BT signal within the Sun disk, the correction of the Sun BT signal even when the Sun is in the back of the instrument, which is observed through the side-lobes, and the addition of new Radio Frequency Interference (RFI) flags to alert users of a degradation of certain BT measurements. All these changes introduced a clear po

The Faraday rotation (FR) is a nonnegligible effect at the L-band, which is the operation frequency of the Soil Moisture and Ocean Salinity (SMOS) mission. This effect introduces a rotation in the electromagnetic field polarization when propagating through the ionosphere that must be compensated. Recently, a methodology was developed in order to retrieve the vertical total electron content (VTEC) from SMOS radiometric data with the aim to better correct the FR effect [1] . In that work, systematic patterns in the retrieved FR angle (FRA) were detected. In this article, these systematic patterns are characterized and corrected to improve the quality of the retrieved VTEC maps. These maps can be then reused in the SMOS level 2 processor for the correction of the FRA in the mission. The impact of using the SMOS-derived VTEC maps instead of the VTEC data from global positioning system (GPS) measurements on the ocean brightness temperatures (TB) measurement has also been analyzed. Results of this analysis show that the usage of those maps allows a significant enhancement in the quality of the TB, which will lead to an improvement on salinity retrievals

Changes in the Earth’s water cycle can be estimated by analyzing sea surface salinity. This variable reflects the balance between precipitation and evaporation over the ocean, since the upper layers of the ocean are the most sensitive to atmosphere–ocean interactions. In situ measurements lack spatial and temporal synopticity and are typically acquired at few meters below the surface. Satellite measurements, on the contrary, are synoptic, repetitive and acquired at the surface. Here we show that the satellite-derived sea surface salinity measurements evidence an intensification of the water cycle (the freshest waters become fresher and vice-versa) which is not observed at the in-situ near-surface salinity measurements. The largest positive differences between surface and near-surface salinity trends are located over regions characterized by a decrease in the mixed layer depth and the sea surface wind speed, and an increase in sea surface temperature, which is consistent with an increased stratification of the water column due to global warming. These results highlight the crucial importance of using satellites to unveil critical changes on ocean–atmosphere fluxes

Es presenta el disseny i la implementació d’un sistema de detecció de cares amb i sense mascareta, preparat per poder ser exportat en diferents plataformes. S’ha desenvolupat amb TensorFlow i amb la llibreria OpenCV mitjançant els llenguatges de programació Python i C++. El model prèviament entrenat SSD MobileNet V2 FPNLite 320x320 ha estat la base de l’entrenament de la xarxa neuronal. L’entrenament s’ha dut a terme a partir d’una base d’imatges àmplia i diversa, construïda gràcies a la creació d’un procés automatitzat que, donades imatges de cares sense mascareta, es generen cares amb mascareta de diferents tipus, colors i contrasts. A més a més, s’ha augmentat aquesta base d’imatges amb altres processos automatitzats, com el canvi aleatori de la brillantor de la imatge o el redimensionament i col·locació de les cares amb mascareta generades sobre fons aleatoris. Aquestes accions permeten aconseguir més robustesa per a imatges amb múltiples il·luminacions, així com per a cares de diverses mides i col·locades en diferents parts de la imatge. La situació de la pandèmia de la COVID-19 fa que l’eina pugui tenir diverses aplicacions mentre l’ús de les mascaretes sigui obligatori en espais tancats. Des de control de l’entrada en una porta automàtica a aportar informació a una persona amb visibilitat reduïda., Se presenta el diseño y la implementación de un sistema de detección de caras con y sin mascarilla, preparado para poder ser exportado en diferentes plataformas. Se ha desarrollado con TensorFlow y con la librería OpenCV mediante los lenguajes de programación Python y C++. El modelo previamente entrenado SSD MobileNet V2 FPNLite 320x320 ha sido la base del entrenamiento de la red neuronal. El entrenamiento se ha llevado a cabo a partir de una base de imágenes amplia y diversa, construida gracias a la creación de un proceso automatizado que, dadas imágenes de caras sin mascarilla, se generan caras con mascarilla de diferentes tipos, colores y contrastes. Además, se ha aumentado esta base de imágenes con otros procesos automatizados como el cambio aleatorio del brillo de la imagen o el redimensionamiento y la colocación de las caras con mascarilla generadas sobre fondos aleatorios. Estas acciones permiten conseguir mas robustez para imágenes con múltiples iluminaciones, así como para caras de varios tamaños y colocadas en diferentes partes de la imagen. La situación de la pandemia de la COVID-19 hace que la herramienta pueda tener varias aplicaciones mientras el uso de las mascarillas sea obligatorio en espacios cerrados. Desde control de la entrada en una puerta automática a aportar información a una persona con visibilidad reducida., The design and implementation of a face wearing and not wearing a face mask detection system are presented, prepared to be exported for different platforms. It has been developed with TensorFlow and with the OpenCV library using the Python and C++ program languages. The pre-trained model SSD MobileNet V2 FPNLite 320x320 has been the basis of neural network training. The model has been trained from a wide and diverse dataset, built by the creation of an automated process which, given images of faces not wearing a face mask, generate faces wearing a face mask of different types, colors, and contrasts. In addition, this dataset has been increased with other automated processes, such as random change in image brightness or resizing and placing faces with a mask on random backgrounds. These actions provide more robustness for images with multiple illuminations, as well as for faces of various sizes and placed in different parts of the image. The situation of the COVID-19 pandemic makes the tool able to have several applications while the use of masks is mandatory in enclosed spaces. From input control on an automatic door to provide information to a person with reduced visibility., Objectius de Desenvolupament Sostenible::3 - Salut i Benestar

ESA's Soil Moisture and Ocean Salinity (SMOS) mission [1] has been in orbit for over 11 years, and its Microwave Imaging Radiometer with Aperture Synthesis (MIRAS) in two dimensions keeps being fully operational. This II-year long lifetime of SMOS, so far, has enabled the calibration and Level-1 processor team to improve the calibration procedures and the image reconstruction resulting in a new version of the Level-1 data processor, v724. To present the main performance features of this new version and the improvement in the calibration procedures constitute the main objective and content of this presentation

In this work, a methodology to correct the Faraday Rotation Angle (FRA) is presented. It consists of calculating a systematic error pattern introduced by MIRAS, which is calculated in zones where the FRA tends to zero. Once calculated, this error is subtracted in the rest of the measurements. In both cases, the FRA is calculated following a process of minimization of the equation that relates the SMOS full polarization radiometric measurements to that parameter

The quality of the Soil Moisture and Ocean Salinity (SMOS) sea surface salinity (SSS) measurements has been noticeably improved in the past years. However, for some applications, there are still some limitations in the use of the Level-2 ocean salinity product. First, the SSS measurements are still affected by a latitudinal and seasonal bias. Second, the high standard deviation of the SSS error could significantly degrade part of the SSS signal. Finally, the coverage of the Level-2 salinity measurements is significantly reduced after applying filtering criteria to discard the poor-quality retrievals. In this work, we apply nodal sampling to the SMOS brightness temperatures (TBs), which effectively reduces the standard deviation of the TB error; then, we use debiased non-Bayesian retrieval for the mitigation of systematic biases on SSS and the statistical filtering criteria of the degraded salinity retrievals; and finally, we comprehensively characterize the residual latitudinal and seasonal biases and derive a correction for the retrieved SSS. We generate three years of an enhanced SMOS Level-2 Ocean Salinity product and we compare its performances with the ones corresponding to the European Space Agency SMOS Level-2 Ocean Salinity product (v662)

In this work, a new methodology is proposed in order to derive vertical total electron content (VTEC) maps from the radiometric measurements of the Soil Moisture and Ocean Salinity (SMOS) mission as an alternative approach to those based on external databases and models. This approach uses spatiotemporal filtering techniques with optimized filters to be robust against the thermal noise and image reconstruction artifacts present in SMOS images. It is also possible to retrieve the Faraday rotation angle from the recovered VTEC maps in order to correct the effect that it causes in the SMOS brightness temperatures

After more than 10 years in orbit, the SMOS team has started a new reprocessing campaign for the SMOS measurements, which includes the changes in calibration and image reconstruction that have been made to the Level 1 Operational Processor (L1OP) during the past few years. The current L1 processor, version v620, was used for the second mission reprocessing in 2014. The new version, v724, is the one run in the third mission reprocessing and will become the new operational processor. The present paper explains the major changes applied and analyses the quality of the data with different metrics. The results have been obtained with numerous individual tests that have confirmed the benefits of the evolutions and an end-to-end processing campaign involving three years of data used to assess the improvements of the SMOS measurements quantitatively

A method for internally calibrating microwave total power radiometers by using only one level of noise injection is presented. It is based on having a previous accurate characterization of the receiver noise temperature, which used de facto as a second calibration standard. The method proves to be at least equivalent to the classical two level, as demonstrated through their intercomparison using the data provided by the Microwave Imaging Radiometer using Aperture Synthesis (MIRAS) on board the European Space Agency Soil Moisture and Ocean Salinity (SMOS) Satellite. The long-term stability in terms of retrieved brightness temperature using both methods has similar trends with a small advantage for the one-point approach proposed her

After 9 years of operation, the ESA’s SMOS (Soil Moisture and Ocean Salinity) mission continues providing good quality full polarimetric Brightness Temperature (TB) data to generate frequent and global maps of soil moisture over landmasses and surface salinity over the ocean thanks to its unique payload, MIRAS (Microwave Imaging Radiometer using Aperture Synthesis). At its operating frequency (1.41 GHz), there is a non-negligible effect that must be compensated, called Faraday Rotation which rotates the electromagnetic field components coming from the Earth microwave radiation when propagating through the ionosphere. The Faraday rotation angle (FRA) magnitude depends on the frequency, geomagnetic data, and the total electron content on the ionosphere [1] [2]. Currently, the FRA is theoretically estimated in SMOS by using a formulation that depends on two external sources where the first one provides geomagnetic field data and the second one, ionosphere data read from a global VTEC database with an interval of two hours. In order to get improved geophysical parameter retrievals, the FRA must be directly recovered from the SMOS full-pol TB data in a continuous way. Latest advances in image reconstruction led to improving third and fourth Stokes parameters [3] making possible the instantaneous retrieval of the FRA with SMOS full-pol brightness temperature [4]. However, due to the large thermal noise, spatial bias, and image reconstruction artifacts, FRA retrievals for a single snapshot present high errors. A previous work showed that FRA could be directly retrieved at boresight from SMOS full-pol TB with good accuracy by using a smart spatio-temporal filtering strategy [5]. However, averaged boresight FRA estimations are not representative across the complete SMOS field of view (FoV), that is, if the averaged boresight FRA is assigned to all pixels in the FoV, a large systematic bias appears across the FoV [6]. In this work, a new methodology is presented in order to r, REFERENCES [1] D. M. Le Vine and S. Abraham, “The effect of the ionosphere on remote sensing of sea surface salinity from space: Absorption and emission at L band,” IEEE Transactions on Geoscience and Remote Sensing, vol. 40, no. 4, pp. 771–782, April 2002. [2] S. H. Yueh, “Estimates of Faraday rotation with passive Microwave polarimetry for microwave remote sensing of earth surfaces,” IEEE Transactions on Geoscience and Remote Sensing, vol. 38, no. 5, pp. 2434–2438, September 2000. [3] L. Wu et al. , “Radiometric performance of SMOS full polarimetric imaging,” IEEE Geoscience and Remote Sensing Letters, vol. 10, no. 6, pp. 1454–1458, November 2013. [4] S. H. Yueh, “Estimates of Faraday rotation with passive Microwave polarimetry for microwave remote sensing of earth surfaces,” IEEE Transactions on Geoscience and Remote. [5] I. Corbella, L. Wu, F. Torres, N. Duffo and M. Martin-Neira. “Faraday Rotation Retrieval Using SMOS Radiometric Data”. IEEE Geoscience & Remote Sensing Letters, Vol.12, iss. 3, pp. 458- 461. 2015. [6] R. Rubino et al., "Direct faraday rotation angle retrieval in SMOS field of view," 2017 IEEE International Geoscience and Remote Sensing Symposium (IGARSS), pp. 697-698, 2017, doi: 10.1109/IGARSS.2017.8127047. [7] Schaer, S., Gurtner, W., & Feltens, J. (1998, February). IONEX: The ionosphere map exchange format version 1. In Proceedings of the IGS AC workshop, Darmstadt, Germany (Vol. 9, No. 11). [8] Vergely, J.-L., P. Waldteufel, J. Boutin, X. Yin, P. Spurgeon, and S. Delwart (2014), New total electron content retrieval improves SMOS sea surface salinity, J. Geophys. Res. Oceans, 119, 7295–7307, doi:10.1002/2014JC010150

In this work, a refined methodology to correct the Faraday Rotation angle (FRA) for the SMOS mission is proposed. The method is based on calculating the FRA using the SMOS full-pol radiometric data to obtain VTEC maps applying spatiotemporal filtering techniques, to then making possible the compensation of the FRA effect. By this way, the FRA would be corrected from the data that is measuring the satellite and not using an external databas

ESA's Soil Moisture and Ocean Salinity (SMOS) mission [1] has been in orbit for over 9 years, and its Microwave Imaging Radiometer with Aperture Synthesis (MIRAS) in two dimensions is working well. The data products are generated using version v620 of the Level-1 operational processor, a version which entered into operation in Spring 2015. During last year a comprehensive data set was processed using a new processor version v720 and the assessment of the results is expected to be completed by mid 2019. In parallel to this evaluation of v720, the following version v730 of the Level-1 processor of SMOS has been already produced. This latter version is intended for investigating the capability to reduce Radio Frequency Interferences (RFI) by applying image processing techniques. This paper describes the major features and status of the two mentioned versions of the SMOS Level-1 processor, and importantly, aims at updating the remote sensing community on those aspects of the SMOS mission

Temporal stability of the radiometer MIRAS, on board the SMOS satellite, is analyzed using long series of data spanning for about 6 years. Brightness temperature retrieved from one orbit per day over the Pacific Ocean is compared against a forward model and the difference between both is analyzed. Plots of the average in central latitudes as a function of time show small seasonal ripples (~±0.5 K) that depend on the calibration strategy chosen. An extremely low drift of only several mK/year is found and in some cases disappears. Hovmoller plots for the same data, consisting of brightness temperature biases as a function of latitude and time, confirm the small ripples and show some latitude dependent structures, probably coming from geophysical signatures

The quality of the Soil Moisture and Ocean Salinity (SMOS) Sea Surface Salinity (SSS) maps has been noticeably improved in the last two years, in particular those produced at the Barcelona Expert Center (BEC). However, the BEC SSS maps are still affected by a latitudinal and seasonal bias. In this work, we comprehensively characterize the residual latitudinal and seasonal biases, which are used to correct de retrieved SSS, leading to a new generation of higher-quality SSS maps. The shape and regularity of this bias suggests that the effect, which produces this error, is not a poor characterization of the galaxy, some residual Total Electron Content (TEC) effect, or a poor characterization of the systematic Sea Surface Temperature (SST) effects on the SSS retrieval. It appears to be related to a geometrical effect associated to the relative position between the SMOS antenna, the Sun and the Earth

In microwave interferometric radiometers with a large field of view, as for example the Microwave Imaging Radiometer with Aperture Synthesis (MIRAS) onboard the Soil Moisture and Ocean Salinity (SMOS) satellite, one of the major causes of reconstruction error is the contribution to the visibility of the brightness temperature outside the fundamental period, defined on the basis of reciprocal grids. A mitigation method consisting of estimating this contribution through the application of a brightness temperature model outside the fundamental period is proposed. The main advantage is that it does not require any a posteriori addition of artificial scenes to the reconstructed image. Additionally, a method to avoid the sophisticated matrix regularization and inversion techniques usually applied in microwave interferometry is presented. Image reconstruction algorithms are implemented on a minimum grid size in order to maximize their numerical efficiency. An improved method to apply an apodization window to the reconstructed image for reducing Gibbs oscillations is also proposed. All procedures are generally described considering the single polarization case and successively implemented applying the MIRAS layout in both its single polarization and full polarimetric modes. Results show similar performance of the proposed algorithm with respect to the nominal one applied by SMOS. All algorithms are implemented in the MIRAS Testing Software and have been successfully used for scientific studies by other teams

The microwave imaging radiometer with aperture synthesis (MIRAS) is formed by 69 total power radiometers, of which three are the noise-injection type. Their calibration is reviewed on the basis of the data gathered during more than eight years of operation. Internally calibrated gain and offset corrections with improved temporal stability are presented. New front-end loss characterization with lower seasonal dependence originated from external temperature swings is also proposed. Finally, a methodology to validate the external calibrations, with the instrument pointing to the cold sky, is developed. It seems to indicate that the change of orientation of the instrument, with associated thermal variations, may induce small changes in the radiometer front-end losses, thus introducing calibration errors

In this work, a new methodology to estimate the Faraday Rotation Angle (FRA) from SMOS full-pol brightness temperature is presented. This approach is focused on retrieving the Vertical Total Electron Content (VTEC) and then, the FRA. for most of the SMOS overpass, with the final aim of improving geophysical retrievals

It has recently been demonstrated that boresight averaged Faraday rotation angle (FRA) can be retrieved directly from SMOS full-pol radiometric data. However, in order to extend FRA retrievals to the full Alias-Free Field of View (AF-FoV), SMOS relatively poor pixel radiometric sensitivity and accuracy must be compensated by spatial and temporal averaging. This requires some kind of tradeoff to constrain systematic FRA estimation bias both within SMOS AF-FoV and along the orbit. This work presents the first results given by a SMOS end-to-end FRA simulator, currently under development, that is used to trim and assess the performance of several FRA retrieval approaches

It has recently been found that the visibility denormalization process introduces a spatial error distribution due to small sporadic offset jumps in the PMS detectors. The radiometric impact of this error at system level is very low. However, due to the good performance of the SMOS instrument, a study has recently been conducted to evaluate the amplitude of such visibility errors and develop a mitigation technique. The main results of this study are summarized in this presentation

Since each of the individual elements of the MIRAS array is a total power radiometer, the zero-spacing visibility can be obtained by the average of all the corresponding antenna temperatures. The main advantage of this option with respect to using the NIR measurements is that amplitude calibration is more consistent between zero-spacing visibility and the rest. On the other hand, total power radiometers are not usually as stable as noise injection radiometers, so a small loose of stability could be expected. Preliminary results show, however, similar performance

[EN] This article summarizes some of the activities in which Jordi Font, research professor and head of the Department of Physical and Technological Oceanography, Institut de Ciències del Mar (CSIC, Spanish National Research Council) in Barcelona, has been involved as co-Principal Investigator for Ocean Salinity of the European Space Agency Soil Moisture and Ocean Salinity (SMOS) Earth Explorer Mission from the perspective of the Remote Sensing Lab at the Universitat Politècnica de Catalunya. We have probably left out some of his many contributions to salinity remote sensing, but we hope that this review will give an idea of the importance of his work. We focus on the following issues: 1) the new accurate measurements of the sea water dielectric constant, 2) the WISE and EuroSTARRS field experiments that helped to define the geophysical model function relating brightness temperature to sea state, 3) the FROG 2003 field experiment that helped to understand the emission of sea foam, 4) GNSS-R techniques for improving sea surface salinity retrieval, 5) instrument characterization campaigns, and 6) the operational implementation of the Processing Centre of Levels 3 and 4 at the SMOS Barcelona Expert Centre, [ES] Este artículo resume algunas de las actividades en las que Jordi Font, profesor de investigación y jefe del Departamento de Física y Tecnología Oceanográfica, del Institut de Ciències del Mar (CSIC) en Barcelona, ha estado desarrollando como co-Investigador Principal de la parte de la misión SMOS de la ESA, una misión Earth Explorer, desde la perspectiva del Remote Sensing Lab, de la Universitat Politècnica de Catalunya. Seguramente, estamos olvidando algunas de sus muchas contribuciones a la teledetección de la salinidad, pero esperamos que esta revisión dé una idea de la importancia de su trabajo. Este artículo se focaliza en los siguientes puntos: 1) las medidas de alta calidad de la constante dieléctrica del agua marina, 2) las campañas de medidas WISE y EuroSTARRS que ayudaron a la definición del modelo geofísico relacionando la temperatura de brillo con el estado del mar, 3) la campaña de medidas FROG 2003 que ayudó a entender la emisión de la espuma marina 4) presentación de las técnicas de GNSS-R para la mejora de la recuperación de la salinidad superficial 5) campañas para la caracterización del instrumento y 6) la implantación del centro de procesado operacional de niveles 3 y 4 en el SMOS Barcelona Expert Centre

Soil moisture and ocean salinity (SMOS) brightness temperature (TB) images and calibrated visibilities are related by the so-called G-matrix. Due to the incomplete sampling at some spatial frequencies, sharp transitions in the TB scenes generate a Gibbs-like contamination ringing and spread sidelobes. In the current SMOS image reconstruction strategy, a Blackman window is applied to the Fourier components of the TBs to diminish the amplitude of artifacts such as ripples, as well as other Gibbs-like effects. In this paper, a novel image reconstruction algorithm focused on the reduction of Gibbs-like contamination in TB images is proposed. It is based on sampling the TB images at the nodal points, that is, at those points at which the oscillating interference causes the minimum distortion to the geophysical signal. Results show a significant reduction of ripples and sidelobes in strongly radio-frequency interference contaminated images. This technique has been thoroughly validated using snapshots over the ocean, by comparing TBs reconstructed in the standard way or using the nodal sampling (NS) with modeled TBs. Tests have revealed that the standard deviation of the difference between the measurement and the model is reduced around 1 K over clean and stable zones when using NS technique with respect to the SMOS image reconstruction baseline. The reduction is approximately 0.7 K when considering the global ocean. This represents a crucial improvement in TB quality, which will translate in an enhancement of the retrieved geophysical parameters, particularly the sea surface salinity