glacier_prepro.py

from oggm import cfg, utils, workflow, tasks, global_tasks
from oggm.shop import bedtopo, millan22, glathida
from MBsandbox.mbmod_daily_oneflowline import process_w5e5_data
import json, os

if __name__ == "__main__":

    working_dir = utils.gettempdir('ODINN_prepro')
    print("Working directory:", working_dir)
    base_url = 'https://cluster.klima.uni-bremen.de/~oggm/gdirs/oggm_v1.6/L1-L2_files/elev_bands/'

    cfg.initialize()

    # Settings
    cfg.PATHS['working_dir'] = working_dir
    cfg.PARAMS['use_multiprocessing'] = True # use all available CPUs
    cfg.PARAMS['border'] = 10
    cfg.PARAMS['hydro_month_nh'] = 1
    cfg.PARAMS['dl_verify'] = False
    cfg.PARAMS['continue_on_error'] = True

    # List of glaciers and/or regions to process
    rgi_ids = [
        'RGI60-11.00897',  # Hintereisferner, Alps
        'RGI60-11.01270',  # Gepatschferner, Alps
        'RGI60-11.01450',  # Pasterze, Alps
        'RGI60-11.01733',  # Vernagtferner, Alps
        'RGI60-11.02097',  # Kesselwandferner, Alps
        'RGI60-11.02346',  # Taschachferner, Alps
        'RGI60-08.00038',  # Nigardsbreen, Norway
        'RGI60-08.00087',  # Jostedalsbreen, Norway
        'RGI60-08.00147',  # Folgefonna, Norway
        'RGI60-08.00203',  # Hardangerjøkulen, Norway
        'RGI60-11.03638',  # Argentière, France
        'RGI60-07.00042',  # Bodleybreen, Svalbard
        'RGI60-07.00065',  # Rosenthalbreen, Svalbard
    ]

    # Now we initialize the glacier directories
    gdirs = workflow.init_glacier_directories(rgi_ids, 
                                              prepro_base_url=base_url,
                                              from_prepro_level=2)

    # We execute the entity tasks
    list_tasks = [tasks.gridded_attributes,
                  tasks.glacier_masks,
                  bedtopo.add_consensus_thickness,
                  millan22.thickness_to_gdir,
                  millan22.velocity_to_gdir,
                  glathida.glathida_to_gdir]

    for task in list_tasks:
        workflow.execute_entity_task(task, gdirs)

    ### Then we retrieve all the necessary climate data ###
    rgi_paths = {}
    for gdir in gdirs: # TODO: change to parallel processing by creating an entity task
        # We store all the paths for each RGI ID to be retrieved later on in ODINN
        rgi_paths[gdir.rgi_id] = gdir.dir.replace(working_dir+'/', '')
        process_w5e5_data(gdir, climate_type='W5E5', temporal_resol='daily') 

        print("dem path: " , gdir.get_filepath("dem"))

    with open(working_dir + '/rgi_paths.json', 'w') as f:
        json.dump(rgi_paths, f)

    # Verify that glaciers have no missing data
    task_log = global_tasks.compile_task_log(gdirs, 
                                            task_names=["gridded_attributes", "velocity_to_gdir", "thickness_to_gdir"])

    task_log.to_csv(os.path.join(working_dir, "task_log.csv"))