Suggestion: check for previously downloaded files

[sjvrijn]

The run.get_extra_downloads options in the scenario files feel more mandatory than 'extra'.
It also feels error-prone to manually activating these options for the first run and deactivate them afterwards.

One simple alternative could be to always try to download the files, and simply skip the download if they already exist. This would automatically update the set of downloaded files if the date range ever changes. An example implementation:

def download_cds_weather_quantity(
      quantity, year, month, days, hours, area, download_folder):
    '''...'''

>   file_name = f'{download_folder}/{quantity}_{year}_{month}.grib'
>   if os.path.exists(file_name):
>       return  # skip file if already downloaded

    quantity_download = cdsapi.Client()
    quantity_download.retrieve(
        'reanalysis-era5-land',
        {
            'variable': quantity,
            'year': year,
            'month': month,
            'day': days,
            'time': hours,
            'area': area,
            'format': 'grib'
        },
        filename
    )

Note: This does not take changes in the area into account, so that would need to be checked and invalidated separately.

[Omar-Usmani]

The area check needs to be thought out