Breathe London

Breathe London is a dense network of low-cost air quality sensors across Greater London, providing high spatial resolution monitoring data.

Overview

• Coverage: Greater London (~100+ sensors)
• Sensor type: Low-cost electrochemical and optical sensors
• Data quality: Indicative (calibrated against reference monitors)
• API key: Required
• Operator: Vodafone/Airly

Getting an API Key

1. Visit Breathe London Developers
2. Request API access through their portal
3. Set the environment variable:

export BL_API_KEY=your_key_here

Available Pollutants

• NO2 (nitrogen dioxide)
• PM2.5 (fine particulate matter)
• PM10 (coarse particulate matter)

Finding Sites

import aeolus

# Get all Breathe London sites
sites = aeolus.networks.get_metadata("BREATHE_LONDON")

# View site locations
print(sites[['site_code', 'site_name', 'latitude', 'longitude']])

Downloading Data

import aeolus
from datetime import datetime

data = aeolus.download(
    sources="BREATHE_LONDON",
    sites=["BL001", "BL002"],
    start_date=datetime(2024, 1, 1),
    end_date=datetime(2024, 1, 31)
)

Data Quality

Breathe London sensors are indicative rather than reference-grade:

• Calibrated using co-location with AURN reference monitors
• Subject to sensor drift and environmental interference
• Best used for spatial patterns rather than absolute values
• Data marked as ratification='Indicative'

Best Practices

1. Use for relative comparisons - Compare between sites rather than to absolute limits
2. Aggregate temporally - Daily/weekly means are more reliable than hourly
3. Cross-reference with AURN - Validate patterns against nearby reference monitors

Example: London Spatial Analysis

import aeolus
from datetime import datetime

# Download from multiple sensors
data = aeolus.download(
    sources="BREATHE_LONDON",
    sites=["BL001", "BL002", "BL003", "BL004"],
    start_date=datetime(2024, 1, 1),
    end_date=datetime(2024, 1, 31)
)

# Calculate site means
site_means = data.groupby(['site_code', 'measurand'])['value'].mean()
print(site_means.unstack())

Combining with AURN

For robust London analysis, combine Breathe London with AURN reference data:

data = aeolus.download(
    sources={
        "BREATHE_LONDON": ["BL001", "BL002"],
        "AURN": ["MY1", "KC1"]
    },
    start_date=datetime(2024, 1, 1),
    end_date=datetime(2024, 1, 31)
)

# Compare reference vs low-cost
data.groupby(['source_network', 'measurand'])['value'].mean()