Zambretti (zambretti_sager) Integration for Home Assistant — Local Weather Forecast Without Internet
Introduction
Most Home Assistant weather integrations rely on cloud services like OpenWeather or Met.no. But what if you want a fully автономous system that depends only on your own sensors?
That’s exactly what the Zambretti (zambretti_sager) integration delivers — a classic barometric weather forecasting algorithm running entirely inside Home Assistant.
Think of it as a vintage mechanical weather station reborn in software form.
What is the Zambretti Algorithm
The Zambretti algorithm dates back to 1915, created by Negretti & Zambra. It predicts weather based on:
- atmospheric pressure
- pressure trend over time
- wind direction
- seasonal adjustments
Unlike modern forecasting models:
no satellites- no internet
fully local- optimized for short-term forecasts (12–24 hours)
It’s simple, but surprisingly effective when fed with good data.
How the Integration Works
The integration analyzes data from your local sensors and tracks trends:
- pressure (primary signal)
- temperature and humidity
- wind speed and direction
- geographic location
It doesn’t just look at current values — it studies pressure history, which is the secret sauce.
Outputs include:
general forecast
fog likelihood
wind-related insights
weather alerts
Key Features
Fully Local Operation
No external API calls. Everything runs inside Home Assistant.
History-Based Analysis
Requires several hours of pressure history (typically 3–12 hours) to produce reliable results.
Flexible Configuration
You can create multiple sensors with different pressure analysis windows (3h, 6h, 12h).
Rich Attributes
Beyond a simple forecast, you also get:
- pressure trend (rising / falling / steady)
- rate of pressure change
- seasonal adjustments
- wind influence
Installation
Via HACS
- Add the repository
- Select Integration
- Install
- Restart Home Assistant
Manual Installation
- Copy the component into "custom_components"
- Restart Home Assistant
- Add integration via UI
Requirements
Minimum:
- pressure sensor
Recommended:
- temperature
- humidity
- wind speed and direction
More data = better accuracy.
Example Usage
The integration creates a sensor like:
sensor.zambretti_forecast
Attributes include:
- "forecast" — textual forecast
- "trend" — pressure trend
- "fog" — fog probability
- "alert" — weather warnings
Typically displayed in Lovelace using Markdown or Entities cards.
How It Compares to Standard Weather Integrations
Feature| Zambretti| Cloud APIs
Internet required| |
Local accuracy| 
|
Forecast range| 12–24h| up to 7 days
Dependency| |
Limitations
- Less accurate during rapid weather changes
- Highly dependent on sensor quality
- Originally tuned for specific regions (best in mid-latitudes)
- Not a full replacement for modern forecast services
Best Use Cases
- off-grid setups
- boats / RV systems
- backup weather forecasting
- local automations (e.g., close windows when pressure drops)
Final Thoughts
Zambretti isn’t “magic forecasting” — it’s:
«
smart pressure analysis with a century-old brain»
And that’s exactly why it shines:
- fast
- independent
- reliable in stable conditions
It works best as a companion to cloud-based weather integrations, not a replacement.
Conclusion
The zambretti_sager integration gives you:
- a classic algorithm in modern form
fully local weather forecasting- powerful automation potential
If you already have a pressure sensor, you’re just one step away from your own personal micro weather station.
Curious to hear how it compares to Met.no or OpenWeather in real setups 