Step-by-Step guide to building preprocessed directories from scratch

Step-by-Step guide to building preprocessed directories from scratch#

This guide is for anyone who wants to learn how to start from scratch with building preprocessed directories, the same way OGGM does for its users. You might even use your own data. This tutorial assumes you are already familiar with OGGM basics. We won’t go into every little detail here, but you’ll find links for more information if you’re interested.

We’ve structured the guide into five main sections, each dedicated to a different level of preprocessing. At the beginning of each section, we’ll outline the tasks to be performed, the data we’ll use, and provide links to related tutorials. Additionally, at the end of each section, we’ll share a corresponding prepro_base_url. This URL allows you to start directly at that level with everything pre-setup, bypassing the need to complete the earlier steps. Plus, in the tutorial storing glacier directories for later use, we show you how to save your work. This way, you don’t have to redo everything from the beginning every time (many steps only need to be done once).

Tags: advanced, glacier-directory, workflow

Tip: There’s a lot to learn here. If you’re curious about a specific function and want to know more, just add a question mark (?) right after it, and you’ll see more details.

# Example: Getting help on the Python function 'print'
print?

Set-up#

First, let’s get everything ready to go. Here’s how we’ll do it:

  1. Import Functions: We’ll start by importing the functions we need. Depending on which preprocessed level you’re working with, you might not need all of them.

  2. Initialize OGGM: Next, we’ll set up OGGM and choose where we want to save our work (defining the working directory).

  3. Choose a Glacier: Lastly, we’ll pick one glacier to focus on as our example.

Remember, these steps are important no matter which level you’re starting from!

from oggm import cfg, utils, workflow, tasks, DEFAULT_BASE_URL

import geopandas as gpd
import numpy as np
import os

# we always need to initialzie and define a working directory
cfg.initialize(logging_level='WARNING')
cfg.PATHS['working_dir'] = utils.gettempdir(dirname='OGGM-full_prepro_elevation_bands', reset=True)

2024-08-25 21:07:33: oggm.cfg: Reading default parameters from the OGGM `params.cfg` configuration file.
2024-08-25 21:07:33: oggm.cfg: Multiprocessing switched OFF according to the parameter file.
2024-08-25 21:07:33: oggm.cfg: Multiprocessing: using all available processors (N=4)

# Our example glacier
rgi_ids = ['RGI60-11.00897']  # Hintereisferner
rgi_region = '11'  # this must fit to example glacier(s), if starting from level 0

# This section is only for future developments of the tutorial (e.g. updateing for new OGGM releases)
# Test if prepro_base_url valid for both flowline_type_to_use, see level 2.
# In total four complete executions of the notebook:
# (load_from_prepro_base_url=False/True and flowline_type_to_use = 'elevation_band'/'centerline')
load_from_prepro_base_url = False

Level 0#

Tasks:

  • Define the rgi_id for your glacier directory gdir.

  • Define the map projection of the glacier directory

  • Add an outline of the glacier.

  • Optionally add intersects to other outlines.

Data used:

  • Glacier outline

  • Optionally intersects

Related Tutorials:

CAUTION: When using your own glacier outlines, it's important to note that OGGM relies on the defined RGI_ID to fetch calibration data from global datasets, which are tailored to the RGI outlines. If your glacier's outline significantly deviates from its RGI counterpart, this could introduce errors, potentially large or small, into your model's results. Ideally, you should provide your own calibration data for custom outlines or, at the very least, be mindful of the discrepancies this might cause. 
# load all RGI outlines for our region and extract the example glaciers
rgidf = gpd.read_file(utils.get_rgi_region_file(rgi_region, version='62'))
rgidf = rgidf[np.isin(rgidf.RGIId, rgi_ids)]

# We also take care of intersects for this RGI version
cfg.set_intersects_db(utils.get_rgi_intersects_region_file(rgi_region, version='62'))

# set the used projection used for gdir, options 'tmerc' or 'utm'
cfg.PARAMS['map_proj'] = cfg.PARAMS['map_proj']  # default is 'tmerc'

gdirs = workflow.init_glacier_directories(rgidf, reset=True, force=True)

2024-08-25 21:07:40: oggm.workflow: Execute entity tasks [GlacierDirectory] on 1 glaciers

# Instruction for beginning with existing OGGM's preprocessed directories
if load_from_prepro_base_url:
    # to start from level 0 you can do
    prepro_base_url_L0 = 'https://cluster.klima.uni-bremen.de/~oggm/gdirs/oggm_v1.6/L1-L2_files/elev_bands/'
    gdirs = workflow.init_glacier_directories(rgi_ids,
                                              from_prepro_level=0,
                                              prepro_base_url=prepro_base_url_L0,
                                              prepro_border=80,  # could be 10, 80, 160 or 240
                                              reset=True,
                                              force=True,
                                             )

Level 1#

Tasks:

  • Define the border around the outline.

  • Define the local grid resolution, which will also set the resolution for the flowlines.

  • Add the digital elevation model DEM.

  • Set up a local grid for each gdir.

Data used:

  • DEM file

Related Tutorials:

  • dem_sources: Create local topography maps from different DEM sources with OGGM

  • rgitopo_rgi6: RGI-TOPO for RGI v6.0

Please note that registration may be required to access some of the DEM sources. For more information, refer to the dem_sources tutorial. 
# define the border, we keep the default here
cfg.PARAMS['border'] = cfg.PARAMS['border']

# set the method for determining the local grid resolution
cfg.PARAMS['grid_dx_method'] = cfg.PARAMS['grid_dx_method']  # The default method is 'square', which determines the grid spacing (dx) based on the glacier's outline area.
cfg.PARAMS['fixed_dx'] = cfg.PARAMS['fixed_dx']  # This allows setting a specific resolution in meters. It's applicable only when grid_dx_method is set to 'fixed'.

# set the DEM source to use
source = 'COPDEM90'  # we use COPDEM here

# this task adds the DEM and defines the local grid
workflow.execute_entity_task(tasks.define_glacier_region, 
                             gdirs,
                             source=source);

2024-08-25 21:07:40: oggm.workflow: Execute entity tasks [define_glacier_region] on 1 glaciers

# Instruction for beginning with existing OGGM's preprocessed directories
if load_from_prepro_base_url:
    # to start from level 1 you can do
    prepro_base_url_L1 = 'https://cluster.klima.uni-bremen.de/~oggm/gdirs/oggm_v1.6/L1-L2_files/elev_bands/'
    gdirs = workflow.init_glacier_directories(rgi_ids,
                                              from_prepro_level=1,
                                              prepro_base_url=prepro_base_url_L1,
                                              prepro_border=80,  # could be 10, 80, 160 or 240
                                              reset=True,
                                              force=True,
                                             )

Level 2#

Tasks:

  • Choose the type of flowline to use.

  • Create the flowlines surface structure, including surface height and width.

  • Create the downstream flowline, which starts from the glacier’s terminus and extends downstream.

  • Optionally you can bring in extra data from the OGGM-shop and bin it to the elevation band flowline.

Data used:

  • Outline

  • DEM

  • Optional: additional datasets

Related Tutorials:

Starting from this point, it's important to choose the prepro_base_url based on the type of flowline you're working with (see end of this chapter). 
flowline_type_to_use = 'elevation_band'  # you can also select 'centerline' here

if flowline_type_to_use == 'elevation_band':
    elevation_band_task_list = [
        tasks.simple_glacier_masks,
        tasks.elevation_band_flowline,
        tasks.fixed_dx_elevation_band_flowline,
        tasks.compute_downstream_line,
        tasks.compute_downstream_bedshape,
    ]

    for task in elevation_band_task_list:
        workflow.execute_entity_task(task, gdirs);

elif flowline_type_to_use == 'centerline':
    # for centerline we can use parabola downstream line
    cfg.PARAMS['downstream_line_shape'] = 'parabola'

    centerline_task_list = [
        tasks.glacier_masks,
        tasks.compute_centerlines,
        tasks.initialize_flowlines,
        tasks.catchment_area,
        tasks.catchment_intersections,
        tasks.catchment_width_geom,
        tasks.catchment_width_correction,
        tasks.compute_downstream_line,
        tasks.compute_downstream_bedshape,
    ]

    for task in centerline_task_list:
        workflow.execute_entity_task(task, gdirs);
    
else:
    raise ValueError(f"Unknown flowline type '{flowline_type_to_use}'! Select 'elevation_band' or 'centerline'!")

2024-08-25 21:07:41: oggm.workflow: Execute entity tasks [simple_glacier_masks] on 1 glaciers
2024-08-25 21:07:41: oggm.workflow: Execute entity tasks [elevation_band_flowline] on 1 glaciers
2024-08-25 21:07:41: oggm.workflow: Execute entity tasks [fixed_dx_elevation_band_flowline] on 1 glaciers
2024-08-25 21:07:41: oggm.workflow: Execute entity tasks [compute_downstream_line] on 1 glaciers
2024-08-25 21:07:41: oggm.workflow: Execute entity tasks [compute_downstream_bedshape] on 1 glaciers

# Instruction for beginning with existing OGGM's preprocessed directories
if load_from_prepro_base_url:
    # to start from level 2 we need to distinguish between the flowline types
    if flowline_type_to_use == 'elevation_band':
        prepro_base_url_L2 = 'https://cluster.klima.uni-bremen.de/~oggm/gdirs/oggm_v1.6/L1-L2_files/2023.2/elev_bands_w_data/'
    elif flowline_type_to_use == 'centerline':
        prepro_base_url_L2 = 'https://cluster.klima.uni-bremen.de/~oggm/gdirs/oggm_v1.6/L1-L2_files/centerlines/'
    else:
        raise ValueError(f"Unknown flowline type '{flowline_type_to_use}'! Select 'elevation_band' or 'centerline'!")

    gdirs = workflow.init_glacier_directories(rgi_ids,
                                              from_prepro_level=2,
                                              prepro_base_url=prepro_base_url_L2,
                                              prepro_border=80,  # could be 10, 80, 160 or 240
                                              reset=True,
                                              force=True,
                                             )

Level 3#

Tasks:

  • Add baseline climate data to gdir.

  • Calibrate the mass balance model statically (without considering glacier dynamics) using geodetic observations. This involves the calibration of melt_f, prcp_fac and temp_bias.

  • Conduct an inversion for the glacier’s bed topography. Including the calibration of glen_a and fs by matching to the total volume estimate.

  • Create the dynamic flowline for dynamic simulation runs.

Data used:

Related Tutorials:

For the inversion, we diverge from the standard preprocessed directories when focusing on individual glaciers instead of entire regions. This is because calibrate_inversion_from_consensus adjusts glacier volume based on total regional consensus estimates, not per glacier. Since volume estimates are model-based and not directly observed, they're less reliable for individual calibration. However, for our example, we'll calibrate using the consensus estimate for a single glacier, differing from the preprocessed approach. 
# define the climate data to use, we keep the default
cfg.PARAMS['baseline_climate'] = cfg.PARAMS['baseline_climate']

# add climate data to gdir
workflow.execute_entity_task(tasks.process_climate_data, gdirs);

# the default mb calibration
workflow.execute_entity_task(tasks.mb_calibration_from_geodetic_mb,
                             gdirs,
                             informed_threestep=True,  # only available for 'GSWP3_W5E5'
                            );

# glacier bed inversion
workflow.execute_entity_task(tasks.apparent_mb_from_any_mb, gdirs);
workflow.calibrate_inversion_from_consensus(
    gdirs,
    apply_fs_on_mismatch=True,
    error_on_mismatch=True,  # if you running many glaciers some might not work
    filter_inversion_output=True,  # this partly filters the overdeepening due to
    # the equilibrium assumption for retreating glaciers (see. Figure 5 of Maussion et al. 2019)
    volume_m3_reference=None,  # here you could provide your own total volume estimate in m3
);

# finally create the dynamic flowlines
workflow.execute_entity_task(tasks.init_present_time_glacier, gdirs);

2024-08-25 21:07:41: oggm.workflow: Execute entity tasks [process_climate_data] on 1 glaciers
2024-08-25 21:07:41: oggm.workflow: Execute entity tasks [mb_calibration_from_geodetic_mb] on 1 glaciers
2024-08-25 21:07:42: oggm.workflow: Execute entity tasks [apparent_mb_from_any_mb] on 1 glaciers
2024-08-25 21:07:42: oggm.workflow: Applying global task calibrate_inversion_from_consensus on 1 glaciers
2024-08-25 21:07:42: oggm.workflow: Consensus estimate optimisation with A factor: 0.1 and fs: 0
2024-08-25 21:07:42: oggm.workflow: Applying global task inversion_tasks on 1 glaciers
2024-08-25 21:07:42: oggm.workflow: Execute entity tasks [prepare_for_inversion] on 1 glaciers
2024-08-25 21:07:42: oggm.workflow: Execute entity tasks [mass_conservation_inversion] on 1 glaciers
2024-08-25 21:07:42: oggm.workflow: Execute entity tasks [filter_inversion_output] on 1 glaciers
2024-08-25 21:07:42: oggm.workflow: Execute entity tasks [get_inversion_volume] on 1 glaciers
2024-08-25 21:07:42: oggm.workflow: Consensus estimate optimisation with A factor: 10.0 and fs: 0
2024-08-25 21:07:42: oggm.workflow: Applying global task inversion_tasks on 1 glaciers
2024-08-25 21:07:42: oggm.workflow: Execute entity tasks [prepare_for_inversion] on 1 glaciers
2024-08-25 21:07:42: oggm.workflow: Execute entity tasks [mass_conservation_inversion] on 1 glaciers
2024-08-25 21:07:42: oggm.workflow: Execute entity tasks [filter_inversion_output] on 1 glaciers
2024-08-25 21:07:42: oggm.workflow: Execute entity tasks [get_inversion_volume] on 1 glaciers
2024-08-25 21:07:42: oggm.workflow: Consensus estimate optimisation with A factor: 7.876802795318141 and fs: 0
2024-08-25 21:07:42: oggm.workflow: Applying global task inversion_tasks on 1 glaciers
2024-08-25 21:07:42: oggm.workflow: Execute entity tasks [prepare_for_inversion] on 1 glaciers
2024-08-25 21:07:42: oggm.workflow: Execute entity tasks [mass_conservation_inversion] on 1 glaciers
2024-08-25 21:07:42: oggm.workflow: Execute entity tasks [filter_inversion_output] on 1 glaciers
2024-08-25 21:07:42: oggm.workflow: Execute entity tasks [get_inversion_volume] on 1 glaciers
2024-08-25 21:07:42: oggm.workflow: Consensus estimate optimisation with A factor: 3.98840139765907 and fs: 0
2024-08-25 21:07:42: oggm.workflow: Applying global task inversion_tasks on 1 glaciers
2024-08-25 21:07:42: oggm.workflow: Execute entity tasks [prepare_for_inversion] on 1 glaciers
2024-08-25 21:07:42: oggm.workflow: Execute entity tasks [mass_conservation_inversion] on 1 glaciers
2024-08-25 21:07:42: oggm.workflow: Execute entity tasks [filter_inversion_output] on 1 glaciers
2024-08-25 21:07:42: oggm.workflow: Execute entity tasks [get_inversion_volume] on 1 glaciers
2024-08-25 21:07:42: oggm.workflow: Consensus estimate optimisation with A factor: 2.044200698829535 and fs: 0
2024-08-25 21:07:42: oggm.workflow: Applying global task inversion_tasks on 1 glaciers
2024-08-25 21:07:42: oggm.workflow: Execute entity tasks [prepare_for_inversion] on 1 glaciers
2024-08-25 21:07:42: oggm.workflow: Execute entity tasks [mass_conservation_inversion] on 1 glaciers
2024-08-25 21:07:42: oggm.workflow: Execute entity tasks [filter_inversion_output] on 1 glaciers
2024-08-25 21:07:42: oggm.workflow: Execute entity tasks [get_inversion_volume] on 1 glaciers
2024-08-25 21:07:42: oggm.workflow: Consensus estimate optimisation with A factor: 2.869309409059597 and fs: 0
2024-08-25 21:07:42: oggm.workflow: Applying global task inversion_tasks on 1 glaciers
2024-08-25 21:07:42: oggm.workflow: Execute entity tasks [prepare_for_inversion] on 1 glaciers
2024-08-25 21:07:42: oggm.workflow: Execute entity tasks [mass_conservation_inversion] on 1 glaciers
2024-08-25 21:07:42: oggm.workflow: Execute entity tasks [filter_inversion_output] on 1 glaciers
2024-08-25 21:07:42: oggm.workflow: Execute entity tasks [get_inversion_volume] on 1 glaciers
2024-08-25 21:07:42: oggm.workflow: Consensus estimate optimisation with A factor: 2.7157333579518252 and fs: 0
2024-08-25 21:07:42: oggm.workflow: Applying global task inversion_tasks on 1 glaciers
2024-08-25 21:07:42: oggm.workflow: Execute entity tasks [prepare_for_inversion] on 1 glaciers
2024-08-25 21:07:42: oggm.workflow: Execute entity tasks [mass_conservation_inversion] on 1 glaciers
2024-08-25 21:07:42: oggm.workflow: Execute entity tasks [filter_inversion_output] on 1 glaciers
2024-08-25 21:07:42: oggm.workflow: Execute entity tasks [get_inversion_volume] on 1 glaciers
2024-08-25 21:07:42: oggm.workflow: Consensus estimate optimisation with A factor: 2.6897121901531937 and fs: 0
2024-08-25 21:07:42: oggm.workflow: Applying global task inversion_tasks on 1 glaciers
2024-08-25 21:07:42: oggm.workflow: Execute entity tasks [prepare_for_inversion] on 1 glaciers
2024-08-25 21:07:42: oggm.workflow: Execute entity tasks [mass_conservation_inversion] on 1 glaciers
2024-08-25 21:07:42: oggm.workflow: Execute entity tasks [filter_inversion_output] on 1 glaciers
2024-08-25 21:07:42: oggm.workflow: Execute entity tasks [get_inversion_volume] on 1 glaciers
2024-08-25 21:07:42: oggm.workflow: calibrate_inversion_from_consensus converged after 7 iterations and fs=0. The resulting Glen A factor is 2.6897121901531937.
2024-08-25 21:07:42: oggm.workflow: Applying global task inversion_tasks on 1 glaciers
2024-08-25 21:07:42: oggm.workflow: Execute entity tasks [prepare_for_inversion] on 1 glaciers
2024-08-25 21:07:42: oggm.workflow: Execute entity tasks [mass_conservation_inversion] on 1 glaciers
2024-08-25 21:07:42: oggm.workflow: Execute entity tasks [filter_inversion_output] on 1 glaciers
2024-08-25 21:07:42: oggm.workflow: Execute entity tasks [get_inversion_volume] on 1 glaciers
2024-08-25 21:07:42: oggm.workflow: Execute entity tasks [init_present_time_glacier] on 1 glaciers

Guidance on utilizing various baseline climates:

Currently, OGGM supports a variety of baseline climates, including ‘CRU’, ‘HISTALP’, ‘W5E5’, ‘GSWP3_W5E5’ (the default), ‘ERA5’, ‘ERA5L’, ‘CERA’, ‘ERA5dr’, and ‘ERA5L-HMA’. Although switching between these datasets is straightforward, calibrating the mass balance model according to each dataset is more complex. For instance, you’ll need to choose a default precipitation factor that suits both your selected climate dataset and your specific region. Additionally, you must determine the best method to calibrate the mass balance parameters. For a comprehensive guide on the available options, explanations, and how to incorporate your own geodetic observations, please refer to the tutorial massbalance_calibration.

Here’s an example of using the ERA5 dataset:

# define the baseline climate and add it
cfg.PARAMS['baseline_climate'] = 'ERA5'
workflow.execute_entity_task(tasks.process_climate_data, gdirs);

# define the default precipitation factor
cfg.PARAMS['prcp_fac'] = 1.6  # Note: This is not a universial value!
cfg.PARAMS['use_winter_prcp_fac'] = False  # This option is only available for 'GSWP3_W5E5'
cfg.PARAMS['use_temp_bias_from_file'] = False  # This option is only available for 'GSWP3_W5E5'

# an example of static calibration for mass balance, more options are available in the tutorial
workflow.execute_entity_task(tasks.mb_calibration_from_geodetic_mb,
                             gdirs,
                             calibrate_param1='melt_f',
                             calibrate_param2='prcp_fac',
                             calibrate_param3='temp_bias')

You can also utilize your own climate data. However, you will need to either convert your data into a specific format (for an example, see OGGM/oggm-sample-data ->test-files/histalp_merged_hef.nc) or create your own tasks.process_climate_data function. Here’s how you might do this:

cfg.PARAMS['baseline_climate'] = 'CUSTOM'
cfg.PATHS['climate_file'] = path_to_the_climate_file

workflow.execute_entity_task(tasks.process_climate_data, gdirs);

# proceed with defining the default precipitation factor and mass balance calibration as shown above

# Instruction for beginning with existing OGGM's preprocessed directories
if load_from_prepro_base_url:
    # to start from level 3 you can do
    if flowline_type_to_use == 'elevation_band':
        prepro_base_url_L3 = 'https://cluster.klima.uni-bremen.de/~oggm/gdirs/oggm_v1.6/L3-L5_files/2023.3/elev_bands/W5E5/'
    elif flowline_type_to_use == 'centerline':
        prepro_base_url_L3 = 'https://cluster.klima.uni-bremen.de/~oggm/gdirs/oggm_v1.6/L3-L5_files/2023.3/centerlines/W5E5/'
    else:
        raise ValueError(f"Unknown flowline type '{flowline_type_to_use}'! Select 'elevation_band' or 'centerline'!")

    gdirs = workflow.init_glacier_directories(rgi_ids,
                                              from_prepro_level=3,
                                              prepro_base_url=prepro_base_url_L3,
                                              prepro_border=80,  # could be 80 or 160
                                              reset=True,
                                              force=True,
                                             )

Level 4#

Tasks:

  • Initialize the current state of the glacier without a dynamic spinup. This method, default until version 1.6., is mainly for comparison purposes and can often be skipped.

  • Initialize the current glacier state with a dynamic spinup. This process includes a dynamic calibration of the mass balance. It’s important to note that this option isn’t available for centerlines in the current OGGM preprocessed directories, meaning it hasn’t been tested or analyzed.

Data used:

Related Tutorials:

# set the ice dynamic solver depending on the flowline-type
if flowline_type_to_use == 'elevation_band':
    cfg.PARAMS['evolution_model'] = 'SemiImplicit'
elif flowline_type_to_use == 'centerline':
    cfg.PARAMS['evolution_model'] = 'FluxBased'
else:
    raise ValueError(f"Unknown flowline type '{flowline_type_to_use}'! Select 'elevation_band' or 'centerline'!")

# get the start and end year of the selected baseline
y0 = gdirs[0].get_climate_info()['baseline_yr_0']
ye = gdirs[0].get_climate_info()['baseline_yr_1'] + 1  # run really to the end until 1.1.

# 'static' initialisation
workflow.execute_entity_task(tasks.run_from_climate_data, gdirs,
                             min_ys=y0, ye=ye,
                             fixed_geometry_spinup_yr=None,  # here you could add a static spinup if you want
                             output_filesuffix='_historical')

# 'dynamic' initialisation, including dynamic mb calibration
dynamic_spinup_start_year = 1979
minimise_for = 'area'  # other option would be 'volume'
workflow.execute_entity_task(
    tasks.run_dynamic_melt_f_calibration, gdirs,
    err_dmdtda_scaling_factor=0.2,  # by default we reduce the mass balance error for accounting for
    # corrleated uncertainties on a regional scale
    ys=dynamic_spinup_start_year, ye=ye,
    kwargs_run_function={'minimise_for': minimise_for},
    ignore_errors=True,
    kwargs_fallback_function={'minimise_for': minimise_for},
    output_filesuffix='_spinup_historical',
);

2024-08-25 21:07:42: oggm.workflow: Execute entity tasks [run_from_climate_data] on 1 glaciers
2024-08-25 21:07:42: oggm.workflow: Execute entity tasks [run_dynamic_melt_f_calibration] on 1 glaciers

# Instruction for beginning with existing OGGM's preprocessed directories
if load_from_prepro_base_url:
    # to start from level 4 you can do
    if flowline_type_to_use == 'elevation_band':
        prepro_base_url_L4 = DEFAULT_BASE_URL
    elif flowline_type_to_use == 'centerline':
        prepro_base_url_L4 = 'https://cluster.klima.uni-bremen.de/~oggm/gdirs/oggm_v1.6/L3-L5_files/2023.3/centerlines/W5E5/'
    else:
        raise ValueError(f"Unknown flowline type '{flowline_type_to_use}'! Select 'elevation_band' or 'centerline'!")
    gdirs = workflow.init_glacier_directories(rgi_ids,
                                              from_prepro_level=4,
                                              prepro_base_url=prepro_base_url_L4,
                                              prepro_border=80,  # could be 80 or 160
                                              reset=True,
                                              force=True,
                                             )

Level 5#

Tasks:

  • Retain only the data necessary for future projection runs to conserve disk space. At this stage, it’s not possible to revisit the preprocessing steps from earlier levels, but all required information for conducting future projection runs is preserved.

Data used:

  • No additional data is needed for this level.

Related Tutorials:

mini_base_dir = os.path.join(cfg.PATHS['working_dir'],
                             'mini_per_glacier')
mini_gdirs = workflow.execute_entity_task(tasks.copy_to_basedir, gdirs,
                                          base_dir=mini_base_dir,
                                          setup='run/spinup')

2024-08-25 21:07:45: oggm.workflow: Execute entity tasks [copy_to_basedir] on 1 glaciers
When you're ready to access your work later, you should first remove the 'per_glacier' folder from your working directory. Then, rename the 'mini_per_glacier' folder to 'per_glacier'. Remember, proceed with these steps only if you've completed setting up your glacier directories (gdirs) and are sure you won't need to make further changes! 
# Instruction for beginning with existing OGGM's preprocessed directories
if load_from_prepro_base_url:
    # to start from level 5 you can do
    if flowline_type_to_use == 'elevation_band':
        prepro_base_url_L5 = DEFAULT_BASE_URL
    elif flowline_type_to_use == 'centerline':
        prepro_base_url_L5 = 'https://cluster.klima.uni-bremen.de/~oggm/gdirs/oggm_v1.6/L3-L5_files/2023.3/centerlines/W5E5/'
    else:
        raise ValueError(f"Unknown flowline type '{flowline_type_to_use}'! Select 'elevation_band' or 'centerline'!")
    gdirs = workflow.init_glacier_directories(rgi_ids,
                                              from_prepro_level=5,
                                              prepro_base_url=prepro_base_url_L5,
                                              prepro_border=80,  # could be 80 or 160
                                              reset=True,
                                              force=True,
                                             )

And that’s it! We’ve successfully recreated all the preprocessed levels offered by OGGM. Remember, if you prefer, you can bypass all previous steps and jump straight into your future projections from Level 5. Happy modeling!

What’s next?#

Contents