@article{fdi:010088149,
  title = {{M}achine learning for fog-and-low-stratus nowcasting from {M}eteosat {SEVIRI} satellite images},
  author = {{B}ari, {D}. and {L}asri, {N}. and {S}ouri, {R}. and {L}guensat, {R}edouane},
  editor = {},
  language = {{ENG}},
  abstract = {{F}og and low stratus ({FLS}) are meteorological phenomena that have a significant impact on all ways of transportation and public safety. {D}ue to their similarity, they are often grouped together as a single category when viewed from a satellite perspective. {T}he early detection of these phenomena is crucial to reduce the negative effects that they can cause. {T}his paper presents an image-based approach for the short-term nighttime forecasting of {FLS} during the next 5 h over {M}orocco, based on geostationary satellite observations ({M}eteosat {SEVIRI}). {T}o achieve this, a dataset of hourly night microphysics {RGB} product was generated from native files covering the nighttime cold season ({O}ctober to {A}pril) of the 5-year period (2016-2020). {T}wo optical flow techniques (sparse and dense) and three deep learning techniques ({CNN}, {U}net and {C}onv{LSTM}) were used, and the performance of the developed models was assessed using mean squared error ({MSE}) and structural similarity index measure ({SSIM}) metrics. {H}ourly observations from {M}eteorological {A}viation {R}outine {W}eather {R}eports ({METAR}) over {M}orocco were used to qualitatively compare the {FLS} existence in {METAR}, where it is also shown by the {RGB} product. {R}esults analysis show that deep learning techniques outperform the traditional optical flow method with {SSIM} and {MSE} of about 0.6 and 0.3, respectively. {D}eep learning techniques show promising results during the first three hours. {H}owever, their performance is highly dependent on the number of filters and the computing resources, while sparse optical flow is found to be very sensitive to mask definition on the target phenomenon.},
  keywords = {{FLS} ; deep learning ; optical flow ; {M}eteosat {SEVIRI} ; fog ; {MAROC}},
  booktitle = {},
  journal = {{A}tmosphere},
  volume = {14},
  numero = {6},
  pages = {953 [25 p.]},
  year = {2023},
  DOI = {10.3390/atmos14060953},
  URL = {https://www.documentation.ird.fr/hor/fdi:010088149},
}