Maria Soledad Fernandez Carvelo

I studied a bachelor's degree in physics as well as a master's degree in Geophisics and Meteorology at the University of Granada. I was working as Optical Engineer (signaling department) for a few years. Thereafter, I joined to the Color Imaging Lab group at the department of Optics of the University of Granada on the project “Automatic enhancement of images degraded by the atmosphere with multispectral techniques in the visible and infrared”. Nowadays, I am pursuing a PhD in the Atmospheric Physics group (GFAT) at the department of Applied Physic of University of Granada which focus on advanced remote sensing techniques for the study of atmospheric aerosol (LIDAR Technique) which is financed by a national grant for research initiation.

EDUCATION

- 2013: BSc. & MSc. Physics
- 2019: MSc. Meteorology (Physics of the Atmosphere)

RESEARCH INTERESTS

My research interests involve atmospheric optic, color, spectral imaging, dehazing, atmospheric aerosol and clouds interaction, remote sensing, fluorescence and LIDAR technique.

Scopus Publications

Hybrid methodology for optimised water vapour mixing ratio profiles from Raman lidar measurements
Arlett Díaz-Zurita, Daniel Pérez-Ramírez, David N. Whiteman, Onel Rodríguez-Navarro, Víctor M. Naval-Hernández, Jorge Muñiz-Rosado, Soledad Fernández-Carvelo, Jesús Abril-Gago, Ana del Águila, Pablo Ortiz-Amezcua, Juan Antonio Bravo-Aranda, María José Granados-Muñoz, Juan Luis Guerrero-Rascado, Manuel Antón, Javier Vaquero-Martínez, Inmaculada Foyo-Moreno, Jose Antonio Benavent-Oltra, Lucas Alados-Arboledas, Francisco Navas-Guzmán
Atmospheric Measurement Techniques, 2026
This study presents a hybrid methodology to obtain high temporal resolution calibration constants for water vapour Raman lidar measurements, and posteriorly retrieve high-accuracy water vapour mixing ratio profiles. The hybrid method combines correlative measurements of collocated precipitable water vapour and Numerical Weather Prediction data to reconstruct the profile within the incomplete overlap region. The hybrid methodology is applied to the Raman lidar system, which operated at the EARLINET/ACTRIS station of the University of Granada, Spain, for the period 2009–2022. The system has been continuously updated to meet EARLINET/ACTRIS requirements for aerosol measurements, but the hybrid method has allowed tracking the impact of these changes on calibration constants for water vapour retrievals, and consequently to exploit water vapour mixing ratio profiles that were previously unavailable. The hybrid method was optimised for the Granada station by selecting Global Navigation Satellite System precipitable water vapour data as the most appropriate due to its better agreement with collocated and simultaneous radiosonde data (coefficient of determination of 0.95). Furthermore, the ERA5 reanalysis model was selected as the most appropriate because of its better temporal and spatial resolution and its accuracy when evaluated against radiosonde data. The advantages of the hybrid methodology were evaluated in comparison to traditional calibration methods such as those based on radiosondes or precipitable water vapour data assuming a constant water vapour mixing ratio in the incomplete overlap region. Although all methods generally provided good calibration constants, the hybrid method presented the best assessments under conditions where atmospheric layers were not well-mixed. Comparison with radiosonde data revealed excellent agreement, with a mean bias of −0.1 ± 0.3 g kg−1, a standard deviation of 1.0 ± 0.4 g kg−1 and a coefficient of determination of 0.87 across the entire period and vertical range (0–6 kma.g.l.). The most important result of this study is the ability to continuously evaluate calibration constants in a system that its configuration has been changing over 14 years of operation. This new methodology expanded the dataset from 31 initial cases using collocated radiosondes to more than 2000 values through the hybrid methodology. The posterior application of the hybrid methodology to all Raman lidar measurements enabled the generation of a comprehensive database of water vapour mixing ratio profiles for the entire period 2009–2022. Illustrative cases under different atmospheric conditions are presented to showcase the potential of Raman lidar measurements in monitoring water vapour and to investigate its role in climate dynamics and weather prediction.
Aerosol type classification with machine learning techniques applied to multiwavelength lidar data from EARLINET
Ana del Águila, Pablo Ortiz-Amezcua, Siham Tabik, Juan Antonio Bravo-Aranda, Sol Fernández-Carvelo, Lucas Alados-Arboledas
Atmospheric Chemistry and Physics, 2025
Aerosol typing is essential for understanding atmospheric composition and its impact on the climate. Lidar-based aerosol typing has been often addressed with manual classification using optical property ranges. However, few works addressed it using automated classification with machine learning (ML) mainly due to the lack of annotated datasets. In this study, a high-vertical-resolution dataset is generated and annotated for the University of Granada (UGR) station in Southeastern Spain, which belongs to the European Aerosol Research Lidar Network (EARLINET), identifying five major aerosol types: Continental Polluted, Dust, Mixed, Smoke and Unknown. Six ML models – Decision Tree, Random Forest, Gradient Boosting, XGBoost, LightGBM and Neural Network- were applied to classify aerosol types using multiwavelength lidar data from EARLINET, for two system configurations: with and without depolarization data. LightGBM achieved the best performance, with precision, recall, and F1-Score above 90 % (with depolarization) and close to 87 % (without depolarization). The performance for each aerosol type was evaluated and dust classification improved by ∼ 30 % with depolarization, highlighting its critical role in distinguishing aerosol types. Validation against independent datasets, including a smoke case and a Saharan dust event, confirmed robust classification under real and extreme conditions. Compared to NATALI, a neural network-based EARLINET algorithm, the approach presented in this work shows improved aerosol classification accuracy, which emphasize the benefits of using high-resolution multiwavelength lidar data from real measurements. This highlights the potential of ML-based methods for robust and accurate aerosol typing, establishing a benchmark for future studies using multiwavelength lidar at high-resolution data from EARLINET.
Intercomparison of WRF-chem aerosol schemes during a dry Saharan dust outbreak in Southern Iberian Peninsula
Miguel Pino-Carmona, José A. Ruiz-Arias, Sol Fernández-Carvelo, Juan A. Bravo-Aranda, Lucas Alados-Arboledas
Atmospheric Environment, 2024
The Iberian Peninsula (IP), where this study is conducted, has experienced an increase of the frequency and intensity of Saharan aerosol dust outbreaks over the latest decades, which may have an impact on its regional climate. The Weather Research and Forecasting model coupled with chemistry (WRF-chem) has been used worldwide to simulate dust outbreaks and can support the analysis of such potential impacts. However, it includes multiple alternative aerosol parameterization choices that have not been conveniently evaluated in the study region yet. Here, three of the most popular WRF-chem aerosol parameterization schemes, namely, the Goddard Chemistry Aerosol Radiation and Transport (GOCART), the Model for Simulating Aerosol Interactions and Chemistry (MOSAIC) and the Modal Aerosol Dynamics Model for Europe (MADE) schemes, are inter-compared during a strong and dry dust outbreak on July 2021 in southern IP. The results show that the three schemes predict qualitatively similar dust intrusion patterns that are consistent with ground observations and have inter-model dust loading differences smaller than 4%. However, their average dust size distributions differ notably. While GOCART is reasonably consistent with observations, MOSAIC underpredicts the amount of dust particles with sub-micron diameters and overpredicts that of large particles and MADE does the opposite. This is found to have a strong detrimental impact on the prediction performance of dust optical properties in MOSAIC and MADE, which is related, at least partially, with issues in the required inter-sectional redistribution of dust parameters during the dust emission and calculation of optical properties. Overall, GOCART generally appears a better choice for strong and dry dust outbreak events in southern IP. It remains to be evaluated during wet dust outbreaks, which is a work underway. • WRF-chem aerosol schemes yield distinct results simulating a dry dust outbreak in southern Spain on July 2021. • Results with GOCART are more consistent with observations of dust optical properties and size distribution. • MOSAIC and MADE present issues related with the representation of mineral dust size distribution.
Synergy between Short-Range Lidar and In Situ Instruments for Determining the Atmospheric Boundary Layer Lidar Ratio
Andres Esteban Bedoya-Velásquez, Romain Ceolato, Gloria Titos, Juan Antonio Bravo-Aranda, Andrea Casans, Diego Patrón, Sol Fernández-Carvelo, Juan Luis Guerrero-Rascado, Lucas Alados-Arboledas
Remote Sensing, 2024
Short-range elastic backscatter lidar (SR-EBL) systems are remote sensing instruments for studying low atmospheric boundary layer processes. This work presents a field campaign oriented to filling the gap between the near-surface aerosol processes regarding aerosol radiative properties and connecting them with the atmospheric boundary layer (ABL), centering attention on the residual layer and the ABL transition periods. A Colibri Aerosol Lidar (CAL) instrument, based on the short-range lidar with high spatio-temporal resolution, was used for the first time in the ACTRIS AGORA facility (Andalusian Global Observatory of the Atmosphere) in Granada (Spain). This study showed the possibility of combining lidar and in situ measurements in the lowermost 150 m. The results address, on the one hand, the characterization of the short-range lidar for developing a method to find the calibration constant of the system and to correct the incomplete overlap to further data exploitation. On the other hand, relevant radiative properties such as the temporal series of the aerosol lidar ratio and extinction coefficient were quantified. The campaign was divided in three different periods based on the vehicular emission peak in the early mornings, namely, before, during, and after the emission peak. For before and after the emission peak data classification, aerosol properties presented closer values; however, large variability was obtained after the emission peak reaching the maximum values of extinction and a lidar ratio up to 51.5 ± 11.9 (Mm)−1 and 36.0 ± 10.5 sr, respectively. During the emission peaks, the values reached for extinction and lidar ratio were up to 136.8 ± 26.5 (Mm)−1 and 119.0 ± 22.7 sr, respectively.
Four Years of Atmospheric Boundary Layer Height Retrievals Using COSMIC-2 Satellite Data
Ginés Garnés-Morales, Maria João Costa, Juan Antonio Bravo-Aranda, María José Granados-Muñoz, Vanda Salgueiro, Jesús Abril-Gago, Sol Fernández-Carvelo, Juana Andújar-Maqueda, Antonio Valenzuela, Inmaculada Foyo-Moreno, Francisco Navas-Guzmán, Lucas Alados-Arboledas, Daniele Bortoli, Juan Luis Guerrero-Rascado
Remote Sensing, 2024
This work aimed to study the atmospheric boundary layer height (ABLH) from COSMIC-2 refractivity data, endeavoring to refine existing ABLH detection algorithms and scrutinize the resulting spatial and seasonal distributions. Through validation analyses involving different ground-based methodologies (involving data from lidar, ceilometer, microwave radiometers, and radiosondes), the optimal ABLH determination relied on identifying the lowest refractivity gradient negative peak with a magnitude at least τ% times the minimum refractivity gradient magnitude, where τ is a fitting parameter representing the minimum peak strength relative to the absolute minimum refractivity gradient. Different τ values were derived accounting for the moment of the day (daytime, nighttime, or sunrise/sunset) and the underlying surface (land or sea). Results show discernible relations between ABLH and various features, notably, the land cover and latitude. On average, ABLH is higher over oceans (≈1.5 km), but extreme values (maximums > 2.5 km, and minimums < 1 km) are reached over intertropical lands. Variability is generally subtle over oceans, whereas seasonality and daily evolution are pronounced over continents, with higher ABLHs during daytime and local wintertime (summertime) in intertropical (middle) latitudes.
Band selection for dehazing algorithms applied to hyperspectral images in the visible range
Sol Fernández-Carvelo, Miguel Ángel Martínez-Domingo, Eva M. Valero, Javier Romero, Juan Luis Nieves, Javier Hernández-Andrés
Sensors, 2021
Images captured under bad weather conditions (e.g., fog, haze, mist, dust, etc.), suffer from poor contrast and visibility, and color distortions. The severity of this degradation depends on the distance, the density of the atmospheric particles and the wavelength. We analyzed eight single image dehazing algorithms representative of different strategies and originally developed for RGB images, over a database of hazy spectral images in the visible range. We carried out a brute force search to find the optimum three wavelengths according to a new combined image quality metric. The optimal triplet of monochromatic bands depends on the dehazing algorithm used and, in most cases, the different bands are quite close to each other. According to our proposed combined metric, the best method is the artificial multiple exposure image fusion (AMEF). If all wavelengths within the range 450–720 nm are used to build a sRGB renderization of the imagaes, the two best-performing methods are AMEF and the contrast limited adaptive histogram equalization (CLAHE), with very similar quality of the dehazed images. Our results show that the performance of the algorithms critically depends on the signal balance and the information present in the three channels of the input image. The capture time can be considerably shortened, and the capture device simplified by using a triplet of bands instead of the full wavelength range for dehazing purposes, although the selection of the bands must be performed specifically for a given algorithm.