Skip to content

HANZE historical flood impacts — introduction#

HANZE (Historical Analysis of Natural Hazards in Europe), compiled by Paprotny et al., is the reference record of observed European flood events and their impacts: real floods since 1870, each with its fatalities, persons affected, area flooded and economic losses. It is the observed hazard → loss side of the flood chain — the yardstick a modelled event set is validated against. Companion to the global emdat backend.

earthlens ships a single hanze backend. This page orients it. For the hands-on walkthrough see Usage; the rendered API is the Reference page.

What you get#

OUTPUT_KIND is per instance here:

  • the default → tabular, a pandas.DataFrame. One row per historical flood, carrying HANZE's documented columns: Country code, Year, Type, Regions affected (NUTS 3), Area affected, Fatalities, Persons affected, and both nominal and inflation-adjusted (2025 €) losses.
  • with_geometry=True → vector, a pyramids FeatureCollection of the affected NUTS-3 regions. Each event's semicolon-separated Regions affected (NUTS 3) list is joined to the NUTS-3 boundary polygons, and every affected region comes back once with an n_events count — ready to draw as a choropleth.

Know these three things before you use it#

1. It is a pinned Zenodo release, and it is a beta. earthlens pins the exact version record 20478847 — HANZE v3.0.1-beta, 1870–2025, 42 countries — rather than the moving concept DOI, so a request is reproducible. The -beta label is the authors'; treat the figures as provisional until a stable v3 lands. The files are small individual objects (a 618 KB events CSV, a 2.4 MB region shapefile zip), so earthlens downloads them directly and caches them under your output directory.

2. The flood-type vocabulary has four values, and Compound is not one of them. The Type column is one of River, Flash, Coastal, or River/Coastal — the combined river-and-coastal case is spelled River/Coastal, not Compound. Select with flood_type="River" (case-insensitive); an unknown type is rejected with a did-you-mean hint.

3. Losses are reported, not exposure-normalised. HANZE ships nominal losses and inflation-adjusted real losses (2025 €). Comparing impacts across eras fairly needs exposure normalisation — accounting for how much more there is to damage today — which is the paper's derived method, not a raw column. earthlens returns the reported columns and does not fabricate a normalised one.

Selection#

The backend is selected purely by facet keyword arguments — it declares no variables:

  • country= — one ISO2 code or a list ("DE", ["DE", "NL"]).
  • region= — one NUTS-3 code or a list, matched against each event's affected-region list.
  • flood_type= — one of the four types or a list.
  • start= / end= — a date window, applied on the event Year.
  • lat_lim= / lon_lim= (or aoi=) — a bbox, resolved through the region geometry.

Licence and citation#

HANZE is CC-BY-4.0 — redistributable and cacheable with attribution. Cite:

What this backend deliberately does not do#

  • The gridded exposure layers and the ~15,000 modelled flood footprints. HANZE also publishes population / land-use exposure grids and a modelled event set (1950–2020). The observed events + impacts table is the deliverable here; the exposure grids and modelled footprints are a follow-up, not part of this backend.
  • Exposure-normalised losses. See point 3 above — the reported columns are returned as-is.