I am Driss BARI and I earned, in 2015, my PhD in Meteorology (option: Ocean, Atmosphere and Continental Surface) from Paul Sabatier University in Toulouse, France. My PhD, which takes part from the FOG project undertaken by the  Moroccan Meteorological Directorate (Maroc Météo), involved numerical weather prediction of fog and understanding of the interaction between the physical processes during its whole life cycle over the coastal regions, particularly over the Grand Casablanca region in Morocco. Prior to that, I completed a Master degree in Meteorology at the same university and also an engineer graduate in Meteorology from the National School of Meteorology in Toulouse, in 2003. From 2004 to 2009, I worked at the regional department of climatology of Maroc Météo, where I was involved in many climatological projects and also in training of observers. Since 2010, I worked as an engineer of Research & Development at the National Center of Meteorological Research (CNRM). My main research field is the data assimilation for a 1D boundary layer model and micro-physical parametrization for fog and low clouds. I am also interested in detection and prediction of fog and low clouds using data mining methods. Since 2004, I have been a partial time lecturer at EHTP school (Ecole Hassania des Travaux Publics) in Casablanca, Morocco where I have taught dynamic meteorology, applied statistics and climatology for future meteorological engineer and Data analysis for other engineering fields.  I also have participated in many e-learning and training activities.

My research focuses on  fog forecasting modeling, particularly in coastal areas, using numerical weather prediction models and/or advanced artificial intelligence techniques. Coastal fogs form in complex regions and can depend, among other factors, on topography, land-sea spatial heterogeneity and the shape of the coast. These fogs are therefore influenced by several spatio-temporal scales and are difficult to predict.

In my first research line, I levarage numerical weather prediction models to improve the accuracy of fog forecasts and to better understand the physical mechanisms that govern the fog life cycle, with a focus on areas coastal. Using detailed meteorological observations, advanced numerical simulations and digital terrain models, I tried to elucidate the specific factors that contribute to the formation of coastal fog, particularly over the northwest coasts of Morocco (Bari et al ., JAMC, 2016). This in-depth understanding of physical processes is essential to improve the prediction capacity of models and enable better management of fog risks in these sensitive regions (Bari et al., QJRMS 2015; Bari et al., AAQR 2018). The initial conditions are also essential to the quality of the forecast, the wind data (zonal and southern components) from the synoptic station (at 10 m) and those from the AMDARs were assimilated for the first time within the framework of a one-dimensional model dedicated to  fog and low clouds forecasting (Bari, Atmos. 2019).

In my second  research line, I levarage artificial intelligence techniques to improve the accuracy of fog forecasting. My research studies focus on AI applications targeting the prediction of low visibility conditions through two approaches:

(1) The inclusion of AI techniques in numerical weather forecasting models, for example by replacing sub-grid parameterizations. Thus, the horizontal visibility at 2m was diagnosed from the main meteorological parameters of the atmospheric boundary layer from the AROME operational model using three supervised machine learning regression techniques ( decision trees – based Ensemble , neural networks and generalized linear methods) (Bari and Ouagabi, SNAS, 2020). In this context, the sensitivity of the model developed by ML to the platform (Weka, H2O, Keras, Scikit Learn) and the algorithm (XGBoost, RF, GBM, DFFN) used was evaluated (Bari et al., MORGEO Conf ., 2020).

(2) The use of AI techniques to determine fog predictability. Numerical models and observations are an essential element for AI learning. Both of these methods have been explored. In certain methods, learning is driven solely by observational data (Bari et al., Atmos. 2023; Bari and Lekhlifi, IJBAS 2015). In other methods such as analogue ensemble forecasting, learning optimally combines the two data sources (NWP model and observations) (Alaoui et al., JMR 2022; Alaoui et al., Atmos. 2023; Alaoui et al., MedGU Conf. 2022).

Publications in Indexed Journals (Scopus/Web of Science)

  • 2023 : Bari, D.; Lasri, N.; Souri, R.; Lguensat, R. Machine Learning for Fog-and-Low-Stratus Nowcasting from Meteosat SEVIRI Satellite Images. Atmosphere2023, 14, 953.
  • 2022 : Alaoui, B., Bari, D., and , Ghabbar, Y.. Surface weather parameters forecasting using analog ensemble method over the main airports of Morocco. J. Meteor. Res., 36(6), 1–17, 2022.
  • 2022 : Alaoui, B.; Bari, D.; Bergot, T.; Ghabbar, Y. Analog Ensemble Forecasting System for Low-Visibility Conditions over the Main Airports of Morocco. Atmosphere 2022, 13, 1704.
  • 2020 : de Vos, M. G., Hazeleger, W., Bari, D., Behrens, J., Bendoukha, S., Garcia-Marti, I., van Haren, R., Haupt, S. E., Hut, R., Jansson, F., Mueller, A., Neilley, P., van den Oord, G., Pelupessy, I., Ruti, P., Schultz, M. G., and Walton, J.: Open weather and climate science in the digital era, Geosci. Commun., 3, 191–201,, 2020.
  • 2020 : Bari, D., Ameksa, M., and Ouagabi, A. A comparison of datamining tools for geo-spatial estimation of visibility from AROME-Morocco model outputs in regression framework, 2020 IEEE International conference of Moroccan Geomatics (Morgeo), Casablanca, Morocco, 2020, pp. 1-7,
  • 2020 : Bari, D., and Ouagabi, A. Machine-learning regression applied to diagnose horizontal visibility from mesoscale NWP model forecasts. Springer Nature Applied Sciences. 2, 556.
  • 2019 : Bari, D. A Preliminary Impact Study of Wind on Assimilation and Forecast Systems into the One-Dimensional Fog Forecasting Model COBEL-ISBA over Morocco. Atmosphere, 10, 615.
  • 2018 : Bari, D. Visibility Prediction Based on Kilometric NWP Model Outputs Using Machine-Learning Regression, 2018 IEEE 14th International Conference on e-Science (e-Science), Amsterdam, 2018, pp. 278-278,
  • 2018 : Bari, D. and Bergot, T. Influence of Environmental Conditions on Forecasting of an Advection-Radiation Fog: A Case Study from the Casablanca Region, Morocco. Aerosol and Air Quality Research, 18(1), pp.62-78.
  • 2016 : Bari, D., Bergot, T., and EL Khlifi, M. Local Meteorological and Large-Scale Weather Characteristics of Fog over the Grand Casablanca Region, Morocco. Journal of Applied Meteorology and Climatology. 55(8), 1731-1745.
  • 2015 : Bari, D., Bergot, T., and EL Khlifi, M. Numerical Study of a Coastal Fog Event over Casablanca, Morocco. Quarterly Journal of the Royal Meteorological Society. 141 (690), 1894-1905.
  • 2015 : Bari, D. and EL Khlifi, M. LVP Conditions at Mohamed V Airport, Morocco: Local Characteristics and Prediction using Neural Networks. International Journal of basic and applied sciences. 4(4), 354.

Supervision of internships (Levels : Master / Engineer Graduate)

2020 : Nabila LASRI et Rania SOURI. Application du Deep Learning pour la prévision immédiate des brouillards et/ou des nuages bas à partir des images satellitaires. Ecole Hassania des Travaux Publics (EHTP), Casablanca. Cycle d’Ingénieur en Météorologie. Niveau : Ingénieur.

2019 : Mohamed AMEKSA. Benchmark de l’impact de la plateforme et la technique Machine learning sur la performance du modèle développé : Cas d’un problème de régression. Ecole Nationale Supérieure des Arts et Métiers (ENSAM). Université Hassan II. Casablanca. Niveau : Master.

2018 : Abdelali OUAGABI. Estimation de la visibilité horizontale à partir des prévisions du modèle AROME grâce au Datamining sur la partie Nord du Maroc. Département Informatique. Faculté des Sciences. Université Ibn Tofail. Kénitra. Niveau : Master.

2017 : Amina BERGHOUT et Afaf IRKMANE. Optimisation du schéma d’assimilation dans COBEL-ISBA et évaluation de son impact sur la prévision du brouillard à L’aéroport de Nouasseur. Ecole Hassania des Travaux Publics (EHTP), Casablanca. Cycle d’Ingénieur en Météorologie. Niveau : Ingénieur.

2016 : Karima MOUTACHAOUIQ et Mouna KABBOUNE. Evaluation et Validation des simulations du brouillard sur le Maroc issues du modèle AROME cycle 38t1. Ecole Hassania des Travaux Publics (EHTP), Casablanca. Cycle d’Ingénieur en Météorologie. Niveau : Ingénieur.

2013 : Siham BENOUALIDI et Jamila RHILMANE. Brouillard à Nouasseur : Évaluation de la Prévision Humaine et Apport de la Modélisation à l’Aide d’un Modèle 1D forcé par des champs méso-échelle. Ecole Hassania des Travaux Publics (EHTP), Casablanca. Cycle d’Ingénieur en Météorologie. Niveau : Ingénieur.

2011 : Zineb NAIT SAID et Ghizlane CHARIFI. Brouillard à Nouasseur : Climatologie et Modélisation Statistique par les Réseaux de Neurones. Ecole Hassania des Travaux Publics (EHTP), Casablanca. Cycle d’Ingénieur en Météorologie. Niveau : Ingénieur.