Data

Data such as exposure and hazard must be provided from external sources.

Exposure

Exposure describes the set of assets, people, livelihoods, infrastructures, etc. within an area of interest that to be analyzed.

• In ClimateRisk, the baseline of Exposure must be provided in the format of shapefile. Some examples are provided in etc/data

• In addition to the baseline, the asset value must be provided. The value can be either provided as a fixed value, or it can be estimated from LitPop or GDP.

  The options of asset value should be defined in the configuration file (Details can be obtained from Exposure value). LitPop or GDP value can be obtained either from CLIMADA API, or a preprocessing procedure:

  • LitPop (in process/exposure.py):

    from climada.util.api_client import Client
    
    client = Client()
    
    if isinstance(country, str):
        litpop_obj = client.get_litpop(country=country)
    

  • GDP (in etc/data/get_gdp):

    GDP data must be firstly downloaded from The Inter-Sectoral Impact Model Intercomparison Project. Then the downloaded netcdf data can be processed by the CLIMADA API as:

    from climada_petals.entity.exposures.gdp_asset import GDP2Asset
    
    gdpa = GDP2Asset()
    gdpa.set_countries(countries = ['NZL'], ref_year = 2020, path="/Users/zhans/Downloads/gdp_2005soc_0p5deg_annual_2006-2099.nc4")
    gdpa_gdf = gdpa.gdf
    

    Where the path is the downloaded data from ISIMPI. Then the GDP data can be assigned to the exposure baseline data.

Hazard

Currently there are three types of hazards (Tropical cyclone, landslide and flood) can be applied:

• Tropical cyclone:

  Tropical cyclone (TC) contains two types of dataset: track and wind speed. In ClimateRisk at the moment, track is used in the impact or supply-chain calculation, while wind speed is used for estimating cost-benefit.

  There are two ways to obtain the TC data (defined in process/hazard.py):

  • The data can be obtained from the downloaded ibtracs netcdf file (via CLIMADA API):

    from climada.hazard.tc_tracks import TCTracks
    
    hazard_hist = TCTracks.from_ibtracs_netcdf(
        provider=tc_data_cfg["track"]["provider"],
        year_range=str2list_for_year(tc_data_cfg["track"]["year_range"]),
        estimate_missing=True)
    

  • The data also can be directly obtained from CLIMADA API as:

    from climada.util.api_client import Client
    
    client = Client()
    
    client.get_hazard(
        "tropical_cyclone",
        properties={
            "country_name": country,
            "climate_scenario": "historical",
            "nb_synth_tracks": "10"})
    

• Landslide:

Landslide data are obtained from NASA global landslide catalog (points). The CLIMADA-PETALS call Landslide* then can be used to decode the data:

from process.climada_petals.landslide import Landslide

landslide = Landslide.from_hist(bbox=domain, input_gdf=LANDSLIDE_DATA, res=res)

The above gives landslide data globally, while for New Zealand, there are a total of 164 events recorded spanning from 1979 to 2018.

• Flood:

Global flood data can be downloaded from The Inter-Sectoral Impact Model Intercomparison Project. The netcdf file can be obtained from Here.

• flood depth: flddph_150arcsec_clm40_gswp3_0.nc

• flood fraction: fldfrc_150arcsec_clm40_gswp3_0.nc

The downloaded netCDF file can be decoded using :

from climada_petals.hazard.river_flood import RiverFloods
years = array(range(1979, 2010, 1)).tolist()
rf = RiverFlood.from_nc(countries = ["NZL"], years=years, dph_path=dph_path, frc_path=frc_path)

where NZL means we only process the data for New Zealand.