GeoAfrica SDK

One pip install. Every geospatial dataset. Any country on Earth.

Specialising in African communities, powered entirely by open data.

---

Overview

GeoAfrica is a unified Python SDK for geospatial data access, built and maintained by Merczcord Technologies Ltd.
It aggregates 10+ authoritative open-data sources into a single, developer-friendly interface, no complex API wrangling, no data format juggling.

Whether you're mapping health facilities in Nigeria, tracking active wildfires in Ethiopia, analysing population density in Kenya, or querying road networks across West Africa, one import is all you need.

import geoafrica

# That's it. Every dataset in Africa and the world, at your fingertips.

---

Features

Module	What it does	Data Source
geoafrica.boundaries	Admin boundaries at any level (0–3)	GADM, HDX
geoafrica.population	Population grids & statistics	WorldPop, SEDAC
geoafrica.osm	Any OSM feature — hospitals, roads, schools	OpenStreetMap / Overpass
geoafrica.satellite	Sentinel-2, Landsat, Digital Earth Africa STAC	DE Africa, Earth Search
geoafrica.climate	Rainfall & temperature rasters	CHIRPS, ERA5
geoafrica.health	Health facilities worldwide	HealthSites.io, WHO
geoafrica.fire	Near real-time & historical fire detections	NASA FIRMS
geoafrica.elevation	DEM — SRTM, Copernicus, AW3D30, NASADEM	OpenTopography
geoafrica.humanitarian	Crisis, displacement, flood datasets	UN OCHA HDX
geoafrica.roads	Road networks & density stats	OpenStreetMap
geoafrica.viz	Interactive Folium maps, choropleths	-
geoafrica.analysis	Clip, buffer, zonal stats, proximity	-

---

Installation

pip install geoafrica

For satellite imagery support:

pip install geoafrica[satellite]

For climate ERA5 support:

pip install geoafrica[climate]

Full install (everything):

pip install geoafrica[full]

Python 3.9+ required. Works on Windows, macOS, and Linux.

---

API Keys

Most modules work without any API key. A few providers require a free registration:

Provider	Required For	Register
NASA FIRMS	geoafrica.fire	firms.modaps.eosdis.nasa.gov
OpenTopography	geoafrica.elevation (large areas)	portal.opentopography.org
Copernicus CDS	ERA5 climate data	cds.climate.copernicus.eu

Set your keys via CLI:

geoafrica config set GEOAFRICA_FIRMS_KEY your_key_here

Or via environment variables:

export GEOAFRICA_FIRMS_KEY=your_key_here

Or in Python:

import geoafrica
geoafrica.configure(nasa_firms_key="your_key_here")

---

Usage Examples

Administrative Boundaries

import matplotlib.pyplot as plt
from geoafrica import boundaries, viz

# 1. Get Country Outline
nigeria = boundaries.get_country("Nigeria")

# Print the first few rows
print("--- NIGERIA BOUNDARY DATA ---")
print(nigeria.head())

# Static Plot (Matplotlib)
nigeria.plot(color='lightgreen', edgecolor='black', figsize=(6, 6))
plt.title("Nigeria - Country Outline")
plt.show()

# 2. Get States / Provinces
states = boundaries.get_admin("Nigeria", level=1)
print(f"\nTotal states found: {len(states)}")

# 3. Interactive Web Maps (Folium)
# Generate a pan/zoomable interactive map
interactive_map = viz.quick_map(states)
# interactive_map.save("states_map.html")
# In Jupyter/Colab, simply place 'interactive_map' on the last line to view inline:
# interactive_map

# 4. Districts & Bounding Boxes
districts = boundaries.get_admin("Kenya", level=2)
ghana = boundaries.get_by_iso("GH", level=1)
bbox = boundaries.get_bbox("Ethiopia")  # [33.0, 3.4, 47.9, 14.9]

# 5. List supported countries
all_countries = boundaries.list_countries()
africa_only = boundaries.list_countries(region="Africa")

Population Data

from geoafrica import population
import matplotlib.pyplot as plt

# 1km population grid (WorldPop)
pop_grid = population.get_grid("Nigeria", year=2020, resolution=1000)
print(f"Total pop: {int(pop_grid.where(pop_grid > 0).sum().values):,}")

# View the population density raster
pop_grid.plot(cmap="hot", figsize=(7, 7))
plt.title("Nigeria - Population Density")
plt.show()

# Admin-level statistics
stats = population.get_stats("Lagos", country="Nigeria", level=1)
# Returns DataFrame: admin_name, population, area_km2, density_per_km2

# Available data years
years = population.available_years("Ghana")

OpenStreetMap Features

from geoafrica import osm, viz
import matplotlib.pyplot as plt

# Hospitals in Lagos
hospitals = osm.get_amenity("Lagos, Nigeria", amenity="hospital")

# Print top 5 results
print(hospitals[['name', 'amenity', 'geometry']].head())

# View interactively (in Jupyter)
# viz.quick_map(hospitals, color="blue")

# Schools across Ghana
schools = osm.get_amenity("Ghana", amenity="school")

# Primary roads in Nairobi
roads = osm.get_roads("Nairobi", road_type="primary")

# Buildings in a city
buildings = osm.get_buildings("Kigali")

# Custom tags — any OSM feature
water_sources = osm.get_features(
    "Ethiopia",
    tags={"amenity": "drinking_water"}
)

# Search within a bounding box
markets = osm.get_features_bbox(
    bbox=[3.0, 6.3, 3.5, 6.7],
    tags={"amenity": "marketplace"}
)

Climate Data (CHIRPS Rainfall)

from geoafrica import climate
import matplotlib.pyplot as plt

# Annual rainfall for Ethiopia
rain_2023 = climate.get_rainfall("Ethiopia", year=2023)

# View the climate raster
rain_2023.plot(cmap="Blues", figsize=(8, 6))
plt.title("Ethiopia - 2023 Rainfall (mm)")
plt.show()

# Monthly rainfall (July 2022)
rain_july = climate.get_rainfall("Nigeria", year=2022, month=7)

# Monthly time series as DataFrame
monthly = climate.monthly_series("Ghana", year=2023)
# month | month_name | mean_rainfall_mm | max_rainfall_mm

# Rainfall anomaly vs 1981–2010 baseline
anomaly = climate.rainfall_anomaly("Senegal", year=2022)

Active Fires (NASA FIRMS)

from geoafrica import fire, viz
import matplotlib.pyplot as plt

# Active fires in Nigeria (last 7 days)
fires = fire.get_country("Nigeria", days=7)
print(f"Total active fires: {len(fires)}")

# Plot static map
fires.plot(color='red', markersize=5, figsize=(6,6))
plt.title("Active Fires (Last 7 Days)")
plt.show()

# View interactive Heatmap
# heatmap = viz.fire_map(fires, country="Nigeria")

# Fires in West Africa bounding box
west_africa = fire.get_active(bbox=[-18, 4, 16, 20], days=3)

# Historical fire data
hist = fire.get_historical(
    "Democratic Republic of Congo",
    start="2023-01-01",
    end="2023-12-31"
)

# Quick summary
summary = fire.summary(fires, by="acq_date")

Elevation & Terrain

from geoafrica import elevation
import matplotlib.pyplot as plt

# SRTM 30m DEM for Rwanda
dem = elevation.get_dem("Rwanda", source="SRTMGL1")

# Visualize terrain raster
dem.plot(cmap="terrain", figsize=(8, 8))
plt.title("Rwanda - Digital Elevation Model")
plt.show()

# Copernicus 30m DEM
dem_cop = elevation.get_dem("Kenya", source="COP30")

# Elevation profile between two points
profile = elevation.terrain_profile(
    start=(36.8, 3.8),   # (lon, lat) near Addis Ababa
    end=(38.7, 6.1),
    num_points=200,
)
profile.plot(x="distance_km", y="elevation_m")

# Slope and aspect
slope, aspect = elevation.compute_slope_aspect(dem)

# List all available DEM sources
elevation.list_sources()

Health Facilities

from geoafrica import health, viz

# All hospitals in Nigeria
hospitals = health.get_facilities("Nigeria", facility_type="hospital")
print(hospitals[['name', 'facility_type']].head())

# Interactively view hospital locations
# map_hospitals = viz.quick_map(hospitals, color="red")

# Clinics in Ghana
clinics = health.get_facilities("Ghana", facility_type="clinic")

# Nearest 3 hospitals to a point (e.g. Lagos Island)
nearest = health.nearest_to(lat=6.45, lon=3.39, country="Nigeria", n=3)

# Facility count per state
counts = health.count_by_admin("Kenya", level=1)

Humanitarian Data (HDX)

from geoafrica import humanitarian

# Search datasets
results = humanitarian.search("Nigeria flood 2024", rows=10)

# Get all resources in a dataset
resources = humanitarian.get_dataset("cod-ab-nga")

# Download shapefiles
files = humanitarian.download_dataset("cod-ab-nga", resource_format="SHP")

# Country-specific data
nigeria_data = humanitarian.get_country_datasets("Nigeria", data_type="displacement")

# Load as GeoDataFrame directly
gdf = humanitarian.load_geospatial("cod-ab-nga")

Satellite Imagery

from geoafrica import satellite

# Search Sentinel-2 scenes
items = satellite.search(
    collection="sentinel-2-l2a",
    country="Rwanda",
    date_range="2023-06-01/2023-09-30",
    max_cloud_cover=15,
    limit=5,
)
print(f"Found {len(items)} scenes")

# List all collections from Digital Earth Africa
deafrica_cols = satellite.deafrica_products()

# Load RGB composite
import matplotlib.pyplot as plt
rgb = satellite.load_rgb(items[0])

# Plot the RGB composite satellite image
rgb.plot.imshow(figsize=(8, 8))
plt.title("Sentinel-2 RGB Composite")
plt.show()

Analysis Tools

from geoafrica.analysis import spatial, zonal_stats, proximity

# Clip hospitals to Lagos state
lagos = boundaries.get_admin("Lagos", country="Nigeria", level=1)
lagos_hospitals = spatial.clip(hospitals, lagos)

# Buffer schools by 2km (service area)
school_zones = spatial.buffer_km(schools, km=2)

# Population within admin units (zonal stats)
pop_stats = zonal_stats.compute(pop_grid, states, stats=["mean", "sum", "max"])

# Find nearest facility for each village
result = proximity.nearest_facility(villages, hospitals)

# Service coverage — % of population within 5km of any hospital
coverage = proximity.service_coverage(hospitals, pop_grid, radius_km=5)
print(f"Coverage: {coverage['coverage_pct']}%")

Visualization

from geoafrica import viz

# Quick interactive map
m = viz.quick_map(hospitals, color="#E91E63", tooltip_cols=["name", "facility_type"])
m.save("hospitals.html")

# Choropleth map
m = viz.choropleth(
    states.merge(pop_stats, on="NAME_1"),
    column="sum",
    title="Nigeria — Population by State",
)
m.save("population_choropleth.html")

# Multi-layer map
m = viz.quick_map(states)
viz.add_layer(m, hospitals, name="Hospitals", color="#E53935")
viz.add_layer(m, schools, name="Schools", color="#43A047")
m.save("multi_layer.html")

# Fire heatmap
fires_map = viz.fire_map(fires, country="Nigeria")
fires_map.save("fires.html")

I/O

from geoafrica.io import readers, writers

# Read any geospatial file
gdf = readers.read("data/nigeria.geojson")
gdf = readers.read("s3://my-bucket/data.gpkg", layer="roads")

# Read CSV with lat/lon columns
gdf = readers.read_csv_geo("hospitals.csv", lat_col="lat", lon_col="lng")

# Write to multiple formats
writers.to_geojson(hospitals, "hospitals.geojson")
writers.to_shapefile(hospitals, "hospitals.shp")
writers.to_geopackage(hospitals, "data.gpkg", layer="hospitals")
writers.to_csv(hospitals, "hospitals.csv")
writers.to_geoparquet(hospitals, "hospitals.parquet")

---

CLI Usage

# Show version and API key status
geoafrica info

# Download boundaries
geoafrica boundaries nigeria --level 1 --output states.geojson
geoafrica boundaries ghana --level 2 --source hdx --output gh_districts.gpkg

# OpenStreetMap queries
geoafrica osm facilities --location "Kenya" --type hospital --output ke_hospitals.geojson
geoafrica osm roads --location "Lagos, Nigeria" --type primary

# Fire detections
geoafrica fire active --country Nigeria --days 7 --output fires.geojson
geoafrica fire active --bbox "-18,4,30,20" --days 3

# Elevation
geoafrica elevation dem --country Rwanda --source COP30 --output rwanda.tif
geoafrica elevation sources

# Configuration
geoafrica config set GEOAFRICA_FIRMS_KEY your_nasa_key
geoafrica config show

# List countries
geoafrica countries --region africa

---

Supported Countries

GeoAfrica has built-in support for all 54 African nations plus major countries worldwide. Use any of the following identifier formats:

# These all work
boundaries.get_country("Nigeria")
boundaries.get_country("NG")       # ISO-2
boundaries.get_country("NGA")      # ISO-3
boundaries.get_country("DRC")      # Common abbreviation

---

Architecture

geoafrica/
├── core/          # Config, HTTP session, exceptions
├── datasets/      # 10 data modules (boundaries, population, osm, ...)
├── analysis/      # Spatial ops, zonal stats, proximity
├── io/            # Universal readers & writers
└── viz/           # Folium interactive maps

Design principles:

• Plug-and-play — zero config for most modules
• Smart caching — downloads cached locally to avoid re-fetching
• Retry logic — automatic retry with exponential backoff
• Rate limiting — per-host rate limiting to respect API policies
• Standard outputs — all vector data returns GeoDataFrame, rasters return xarray.DataArray

---

Data Attribution

GeoAfrica aggregates open data. Please attribute the original providers:

Source	License	Citation
GADM	CC BY 4.0 (non-commercial)	gadm.org
WorldPop	CC BY 4.0	worldpop.org
OpenStreetMap	ODbL	© OpenStreetMap contributors
NASA FIRMS	Open	NASA EOSDIS FIRMS
CHIRPS	Open	Funk et al., 2015
HealthSites.io	CC BY 4.0	healthsites.io
HDX / UN OCHA	CC BY-IGO 3.0	data.humdata.org
OpenTopography	CC BY 4.0	opentopography.org

---

Contributing

Contributions are welcome! To add a new data module:

1. Create src/geoafrica/datasets/your_module.py
2. Follow the existing module pattern (returns GeoDataFrame or DataArray)
3. Add unit tests in tests/test_your_module.py
4. Add to src/geoafrica/datasets/__init__.py
5. Submit a PR

---

License

MIT License — see LICENSE.

---

About Merczcord Technologies Ltd.

GeoAfrica is a product of Merczcord Technologies Ltd., a technology company focused on building data infrastructure and geospatial intelligence tools for African communities and beyond.

• Website: merczcord.com
• Contact: info@merczcord.com
• GitHub: github.com/Mercy14846/Geo-API

---

_{Made with ❤️ for Africa and the world by Merczcord Technologies Ltd.}