Estimates of Global Coastal Losses Under Multiple Sea Level Rise Scenarios

Results from the Python Coastal Impacts and Adaptation Model (pyCIAM), along with the inputs and source code necessary to replicate these outputs and the results presented in Depsky et al. 2023 (under review).

All zipped Zarr stores can be downloaded and accessed locally or can be directly accessed via code similar to the following:

from fsspec.implementations.zip import ZipFileSystem
import xarray as xr
xr.open_zarr(ZipFileSystem(url_of_file_in_record}}).get_mapper())

File Inventory

Products

• pyCIAM_outputs.zarr.zip: Outputs of the pyCIAM model, using the SLIIDERS dataset to define socioeconomic and extreme sea level characteristics of coastal regions and the 17th, 50th, and 83rd quantiles of local sea level rise as projected by various modeling frameworks (LocalizeSL and FACTS) and for multiple emissions scenarios and ice sheet models.
• diaz2016_outputs.zarr.zip: A replication of the results from Diaz 2016 - the model upon which pyCIAM was built, using an identical configuration to that of the original model.
• suboptimal_capital_by_movefactor.zarr.zip: An analysis of the observed present-day allocation of capital compared to a "rational" allocation, as a function of the magnitude of non-market costs of relocation assumed in the model. See Depsky et al. 2023 for further details.

Inputs

• ar5-msl-rel-2005-quantiles.zarr.zip: Quantiles of projected local sea level rise as projected from the LocalizeSL model, using a variety of temperature scenarios and ice sheet models developed in Kopp 2014, Bamber 2019, DeConto 2021, IPCC SROCC. The results contained in pyCIAM_outputs.zarr.zip cover a broader (and newer) range of SLR projections from a more recent projection framework (FACTS); however, these data are more easily obtained from the appropriate Zenodo records and thus are not hosted in this one.
• diaz2016_inputs_raw.zarr.zip: The coastal inputs used in Diaz 2016, obtained from GitHub and formatted for use in the python-based pyCIAM. These are based on the Dynamic Integrated Vulnerability Assessment (DIVA) dataset.
• surge-lookup-v1.1-seg(_adm).zarr.zip: Pre-computed lookup tables estimating average annual losses from extreme sea levels due to mortality and capital stock damage. This is an intermediate output of pyCIAM and not necessary to replicate the model results. However, it is more time consuming to produce than the rest of the model and is provided for users who may wish to start from the pre-computed dataset. Two versions are provided - the first contains estimates for each unique intersection of ~50km coastal segment and state/province-level administrative unit (admin-1). This is derived from the characteristics in SLIIDERS. The second is simply estimated on a version of SLIIDERS collapsed over administrative unit to vary only over coastal segments. Both are used in the process of running pyCIAM.
• ypk_2000_2100_20221122.zarr.zip: An intermediate output in the creation of SLIIDERS that contains country-level projections of GDP, capital stock, and population, based on the Shared Socioeconomic Pathways (SSPs). This is only used in normalizing costs estimated in pyCIAM by country and global GDP for the purposes of reporting in Depsky et al. 2023. It is not used in the execution of pyCIAM but is provided for the purpose of replicating results reported in the manuscript.

Source Code

• pyCIAM-1.1.2.zip: Contains the python-CIAM package as well as a notebook-based workflow to replicate the results presented in Depsky et al. 2023. It also contains two master shell scripts (run_example.sh and run_full_replication.sh) to assist in executing a small sample of the pyCIAM model or in fully executing the workflow of Depsky et al. 2023, respectively. This code is consistent with release 1.1.2 in the pyCIAM GitHub repository and is available as version 1.1.2 of the python-CIAM package on PyPI.