electricsheepafrica/africa-fao-eve-global-flood-monitoring-system

---
annotations_creators:
- no-annotation
language_creators:
- found
language:
- en
license: cc-by-4.0
multilinguality:
- monolingual
size_categories:
- 100K<n<1M
source_datasets:
- original
task_categories:
- tabular-classification
- other
task_ids: []
tags:
- africa
- humanitarian
- hdx
- electric-sheep-africa
- affected-area
- affected-population
- climate-hazards
- flooding
- afg
- ago
- bgd
- bfa
- bdi
pretty_name: "FAO EVE Global Flood Monitoring System"
dataset_info:
splits:
- name: train
num_examples: 155606
- name: test
num_examples: 38901
---

# FAO EVE Global Flood Monitoring System

**Publisher:** Food and Agriculture Organization (FAO) of the United Nations · **Source:** [HDX](https://data.humdata.org/dataset/fao-eve-global-flood-monitoring-system) · **License:** `cc-by` · **Updated:** 2026-04-03

---

## Abstract

The DIEM Events Visualization in Emergencies (EVE) system provides resources to enhance the understanding of flood events and their impact on different land cover types, with a particular focus on agricultural areas. EVE provides a flood persistence analysis as well as an estimation of the population exposed to such events.

EVE utilizes satellite-derived data from the NOAA Visible Infrared Imaging Radiometer Suite (VIIRS) at a 375-meter resolution, alongside land cover data from the European Space Agency’s WorldCover 10m 2021 dataset. Covering approximately 40 countries, the system offers daily and biweekly insights, providing a continuously updated view of flood dynamics and their effects. Access the [FAO EVE user guide](https://data-in-emergencies.fao.org/documents/3335ee769a4e45708e27d2ee25d13bef/about) to learn more.

The platform presents results through interactive maps, charts, and tables, supporting decision-making in disaster management, agricultural planning, and environmental monitoring. Most resources are publicly accessible, though downloading aggregated data at the admin2 level requires a [DIEM account](https://hqfao.maps.arcgis.com/sharing/rest/oauth2/signup?client_id=aEXLMtXxljlIrgPN&response_type=token&expiration=20160&showSocialLogins=true&locale=en-us&redirect_uri=https%3A%2F%2Fdata-in-emergencies.fao.org%2Ftorii-provider-arcgis%2Fhub-redirect.html).

EVE products are preliminary analyses and have not yet undergone field validation. Users are encouraged to provide ground feedback to the [FAO Data in Emergencies (DIEM) team](https://data-in-emergencies.fao.org/pages/contactus) to enhance the accuracy and utility of the data.

Each row in this dataset represents country-level aggregates. Temporal coverage is indicated by the `start_date`, `end_date` column(s). Geographic scope: **AFG, AGO, BGD, BFA, BDI, KHM, CMR, CAF, and 28 others**.

*Curated into ML-ready Parquet format by [Electric Sheep Africa](https://huggingface.co/electricsheepafrica).*

---

## Dataset Characteristics

| | |
|---|---|
| **Domain** | Climate and environment |
| **Unit of observation** | Country-level aggregates |
| **Rows (total)** | 194,508 |
| **Columns** | 19 (8 numeric, 9 categorical, 2 datetime) |
| **Train split** | 155,606 rows |
| **Test split** | 38,901 rows |
| **Geographic scope** | AFG, AGO, BGD, BFA, BDI, KHM, CMR, CAF, and 28 others |
| **Publisher** | Food and Agriculture Organization (FAO) of the United Nations |
| **HDX last updated** | 2026-04-03 |

---

## Variables

**Geographic** — `adm0_iso3` (THA, VNM, COL), `admin_level` (admin2), `pop_exposed` (range 0.0–1883766.0).

**Temporal** — `period_number` (range 13.0–54.0), `start_date`, `end_date`.

**Outcome / Measurement** — `total_area_flooded_sq_km` (range 0.01–5332.05), `total_area_flooded_ha` (range 1.0–533205.0), `perc_total_area_flooded` (range 0.0–99.3086).

**Identifier / Metadata** — `adm0_name` (Thailand, Viet Nam, Colombia), `adm1_pcode` (CO05, CO13, PK62022), `adm1_name` (Antioquia, Bolívar, Punjab), `adm2_pcode` (MMR001D001, CD7409, CD9202), `adm2_name` (Chau Thanh, Sucre, Bolívar) and 2 others.

**Other** — `cropland_flooded_sq_km` (range 0.0–2478.43), `cropland_flooded_ha` (range 0.0–247843.0), `perc_cropland_flooded` (range 0.0–100.0).

---

## Quick Start

```python
from datasets import load_dataset

ds = load_dataset("electricsheepafrica/africa-fao-eve-global-flood-monitoring-system")
train = ds["train"].to_pandas()
test = ds["test"].to_pandas()

print(train.shape)
train.head()
```

---

## Schema

| Column | Type | Null % | Range / Sample Values |
|---|---|---|---|
| `adm0_iso3` | object | 0.0% | THA, VNM, COL |
| `adm0_name` | object | 0.0% | Thailand, Viet Nam, Colombia |
| `admin_level` | object | 0.0% | admin2 |
| `adm1_pcode` | object | 0.0% | CO05, CO13, PK62022 |
| `adm1_name` | object | 0.0% | Antioquia, Bolívar, Punjab |
| `adm2_pcode` | object | 0.0% | MMR001D001, CD7409, CD9202 |
| `adm2_name` | object | 0.0% | Chau Thanh, Sucre, Bolívar |
| `period_number` | int64 | 0.0% | 13.0 – 54.0 (mean 35.1181) |
| `start_date` | datetime64[ns] | 0.0% | |
| `end_date` | datetime64[ns] | 0.0% | |
| `cropland_flooded_sq_km` | float64 | 0.0% | 0.0 – 2478.43 (mean 11.8787) |
| `cropland_flooded_ha` | int64 | 0.0% | 0.0 – 247843.0 (mean 1187.8734) |
| `total_area_flooded_sq_km` | float64 | 0.0% | 0.01 – 5332.05 (mean 54.5666) |
| `total_area_flooded_ha` | int64 | 0.0% | 1.0 – 533205.0 (mean 5456.6583) |
| `perc_cropland_flooded` | float64 | 0.0% | 0.0 – 100.0 (mean 3.7183) |
| `perc_total_area_flooded` | float64 | 0.0% | 0.0 – 99.3086 (mean 3.3393) |
| `pop_exposed` | int64 | 0.0% | 0.0 – 1883766.0 (mean 7721.7498) |
| `esa_source` | object | 0.0% | HDX |
| `esa_processed` | object | 0.0% | 2026-04-05 |

---

## Numeric Summary

| Column | Min | Max | Mean | Median |
|---|---|---|---|---|
| `period_number` | 13.0 | 54.0 | 35.1181 | 37.0 |
| `cropland_flooded_sq_km` | 0.0 | 2478.43 | 11.8787 | 0.2 |
| `cropland_flooded_ha` | 0.0 | 247843.0 | 1187.8734 | 20.0 |
| `total_area_flooded_sq_km` | 0.01 | 5332.05 | 54.5666 | 8.17 |
| `total_area_flooded_ha` | 1.0 | 533205.0 | 5456.6583 | 817.0 |
| `perc_cropland_flooded` | 0.0 | 100.0 | 3.7183 | 0.2217 |
| `perc_total_area_flooded` | 0.0 | 99.3086 | 3.3393 | 0.7009 |
| `pop_exposed` | 0.0 | 1883766.0 | 7721.7498 | 602.0 |

---

## Curation

Raw data was downloaded from HDX via the CKAN API and converted to Parquet. Column names were lowercased and standardised to snake_case. Common missing-value markers (`N/A`, `null`, `none`, `-`, `unknown`, `no data`, `#N/A`) were unified to `NaN`. 2 column(s) were cast from string to numeric or datetime based on parse-success rate (>85% threshold). The dataset was split 80/20 into train and test partitions using a fixed random seed (42) and saved as Snappy-compressed Parquet.

---

## Limitations

- Data originates from Food and Agriculture Organization (FAO) of the United Nations and has not been independently validated by ESA.
- Automated cleaning cannot correct for misreported values, definitional inconsistencies, or sampling bias in the original collection.
- This dataset spans 36 countries; geographic and methodological inconsistencies across national boundaries may affect cross-country comparability.
- Refer to the [original HDX dataset page](https://data.humdata.org/dataset/fao-eve-global-flood-monitoring-system) for the publisher's own methodology notes and caveats.

---

## Citation

```bibtex
@dataset{hdx_africa_fao_eve_global_flood_monitoring_system,
title = {FAO EVE Global Flood Monitoring System},
author = {Food and Agriculture Organization (FAO) of the United Nations},
year = {2026},
url = {https://data.humdata.org/dataset/fao-eve-global-flood-monitoring-system},
note = {Repackaged for machine learning by Electric Sheep Africa (https://huggingface.co/electricsheepafrica)}
}
```

---

*[Electric Sheep Africa](https://huggingface.co/electricsheepafrica) — Africa's ML dataset infrastructure. Lagos, Nigeria.*