Quick start

  1. Prepare the environment:
    • Python 3.6.x
    • Anaconda 3 ANACONDA-download
  2. Install libraries (if not installed):
    • pandas
    • scipy
    • numpy
    • configparser
  3. Get the gfortran compiler:
    GFORTRAN GFORTRAN
  4. Download SELENE to your file system:
    SELENE SELENE Software
  5. Prepare FORTRAN foreman script:
    • Go to /SELENE/foreman and compile predicc.f with gfortran and options -fno-automatic, -ffixed-line-lenght-132 and -O2
    • Give execution permissions to predicc.e (+755)
  6. Configure SELENE:
    • Edit /SELENE/configuration/config.ini to indicate where is foreman and outputs folder
    • Edit /SELENE/configuration/stations.json to add stations and related parameters
  7. Prepare original sea level data series:
    Copy sea level time series files to data files folder (these files are configured in stations.json)
  8. Run SELENE:
    python selene.py [station_id]
  9. Visualize results with SELENE:
    python selenevis.py -station [station_id]

Display of the output files of SELENE software applied to Melilla Tide Gauge (REDMAR network) data in July 2018: top: original data (1-min) with QC flags (in red the bad data detected by the software); center: two by-products generated by SELENE: 5-min interpolated time series and hourly values (orange); bottom: final by-product generated if harmonic constants are available, the hourly residual for the analyzed period