teoaivalis/socio-ecological-regions

Socio-Environmental Regions Dataset

This dataset contains environmental and socioeconomic data organized by geographical regions and countries. The data is extracted from a multi-axis KG and covers 149 processed geographical regions and countries, mapping their physical terrain characteristics, local species, and historical flood profiles.

Environmental and Exposure Axes

Each region contains data structured across three primary categories:

• —Biological Axis: Records flora and fauna observations from the iNaturalist database. The data follows a taxonomy hierarchy (Kingdom $\rightarrow$ Species), with an average of 121.24 unique species per region.
• —Urban Morphology and Exposure Axis: Tracks the structural density and population exposure of the built environment using the Global Map of Local Climate Zones (LCZ) and GHSL data. This includes metrics for population density, mean building height, and total built volume.
• —Physical Land Cover Axis: Contains land-cover distributions from ESA LandCover data, mapping surface characteristics such as tree cover, wetlands, grasslands, and croplands for each region.

Historical Disasters & Connected Reports

For each of the 149 regions, the dataset lists the historical extreme flood events that occurred in that territory.

The historical_disasters field contains a list of unique event identifiers. The full textual reports, impact metrics (such as displacedPeople, housesCollapsed, powerOutage, and waterSupplyCut), and spatiotemporal anchors for these events are stored in the main registry dataset: 👉 [teoaivalis/extreme-floods-kg](https://huggingface.co/datasets/teoaivalis/extreme-floods-kg)

Schema Architecture

Each record in the dataset represents a single region and contains the following fields:

• —`region_name` : The name of the country or territory.
• —`location` : The exact coordinates containing lat and lon fields.
• —`other_properties` : Metadata metrics tracking terrain and population features (e.g., population_10km, built_volume_10km, building_density_10km, avg_building_height_10km, and object counts like treesCount, mountainsCount, springsCount, woodsCount).
• —`ecological_features` : Lists the categorical assignments (feature_type and feature_name) for the region's land cover classes, settlement models, and geomorphic zones.
• —`historical_disasters` : A list of extreme flood event text IDs associated with the region.
• —`regional_species` : Lists the local biodiversity profiles (kingdom, species_name, category, inaturalistId, wikipediaUrl, photoUrl).

Dataset Sample: Albania

Below is a structurally complete, ultra-clean instance highlighting the multi-layered capabilities of the region records:

{
  "region_name": "Albania",
  "location": {
    "lat": -28.38,
    "lon": 23.91
  },
  "other_properties": {
    "population_10km": 3298.921253539622,
    "cliffs": 8.0,
    "mountain_ranges": 1.0,
    "avg_building_height_10km": 0.006253737913390194,
    "woods": 25.0,
    "building_density_10km": 0.002535844774136404,
    "scrubs": 25.0,
    "grasslands": 3.0,
    "cave_entrances": 1.0,
    "peaks": 13.0,
    "built_volume_10km": 1171322.0
  },
  "ecological_features": [
    {
      "feature_type": "hasLandCover",
      "feature_name": "Permanent water bodies"
    },
    {
      "feature_type": "hasLandCover",
      "feature_name": "Built-up"
    },
    {
      "feature_type": "hasLandCover",
      "feature_name": "Tree cover"
    },
    {
      "feature_type": "hasSettlementType",
      "feature_name": "Very low density rural"
    }
  ],
  "historical_disasters": [
      "FL-2002-000607-ALB",
      "FL-2017-000174-ALB"
  ],
  "regional_species": [
    {
      "kingdom": "Insecta",
      "species_name": "Omocestus rufipes",
      "characteristics": {
        "category": "Insecta",
        "wikipediaUrl": "[http://en.wikipedia.org/wiki/Omocestus_rufipes](http://en.wikipedia.org/wiki/Omocestus_rufipes)",
        "inaturalistId": 362311,
        "photoUrl": "[https://inaturalist-open-data.s3.amazonaws.com/photos/248733151/medium.jpeg](https://inaturalist-open-data.s3.amazonaws.com/photos/248733151/medium.jpeg)"
      }
    },
    {
      "kingdom": "Plantae",
      "species_name": "Echium vulgare",
      "characteristics": {
        "category": "Plantae",
        "wikipediaUrl": "[http://en.wikipedia.org/wiki/Echium_vulgare](http://en.wikipedia.org/wiki/Echium_vulgare)",
        "inaturalistId": 56174,
        "photoUrl": "[https://inaturalist-open-data.s3.amazonaws.com/photos/130825766/medium.jpg](https://inaturalist-open-data.s3.amazonaws.com/photos/130825766/medium.jpg)"
      }
    },
    {
      "kingdom": "Mammalia",
      "species_name": "Vulpes vulpes",
      "characteristics": {
        "category": "Mammalia",
        "wikipediaUrl": "[http://en.wikipedia.org/wiki/Red_fox](http://en.wikipedia.org/wiki/Red_fox)",
        "inaturalistId": 42069,
        "photoUrl": "[https://static.inaturalist.org/photos/265916780/medium.jpg](https://static.inaturalist.org/photos/265916780/medium.jpg)"
      }
    }
  ]
}

⚡ How to Load

You can download this dataset using the Hugging Face datasets library:

from datasets import load_dataset

# Load the dataset splits directly from the hub
dataset = load_dataset("teoaivalis/socio-ecological-regions")

# Print dataset information
print(dataset)

# Preview the first spatial entry
first_record = dataset["train"][0]
print(first_record["region_name"])
print(first_record["location"])

📜 Citation

If you use this dataset, please cite us!