PAGO uses MATLAB structures to store information regarding grid and outputs, which can be directly saved in .mat files:
MODEL_grid contains the limits of the subset of the grid that is analyzed, longitude, latitude and mask over this subset,
MODEL_sections contains all information for each section: grid characteristics, section indices and data along the section,
MODEL_area contains all information for each area: grid characteristics, area indices and data within area,
MODEL_indices contains information about the transport across each section,
MODEL_volumes contains information about the volume content and convergence for each area.
PAGO functions loaddata_* and indices_MODEL use a so-called structfile as argument. This .mat file contains variables MODEL_grid, MODEL_sections and eventually MODEL_area. This file is originally produced by PAGO function sections_MODEL, with NaN values instead of data output at the first time step. Data variables are then concatenated with actual data output by loaddata_*.
PAGO also offers the possibility to save the information in netCDF format.
data along sections (create_netcdf_sec) ie temperature, salinity and normal velocity, as well as some grid information: length, depth and area of each cell. This file is suitable for visual browsers of netCDF files such as ncview.
indices describing the transport across sections (create_netcdf_ind), over 30 time series. See comments of the netCDF file and description of index calculation for full description of the time series.
indices describing the volume content and convergence (create_netcdf_vol), nearly 30 time series. See comments of the netCDF file and description of volume calculation for full description of the time series.
When the package is used to diagnose monthly data, the user may want to calculate yearly averages of the time series. Depending on the convention used to produce the data, each month may or may not have the same number of days. PAGO offers a possibility to compute yearly averages of monthly indices, taking into account leap years, 360-day-years and 365-day-years (yrave_volumes, yrave_indices).
