WEATHER STATION

Weather station in Home Assistant: wind, rain, temperature and automations

A weather forecast in Home Assistant tells you what is likely to happen in your area. A weather station on your own house tells you what is happening right here, right now: how hard the wind is blowing at the awning, whether it has just started to rain and how much sun is hitting the front of the house.

This guide shows what data a local station can provide, how to connect it to Home Assistant without the cloud, and how to use its readings for heating, blinds and awnings, ventilation, irrigation, solar PV and energy management.

It is not an installation guide for a particular model. It is about what is possible, making good choices when buying and mounting a station, and knowing where weather data alone stops being enough.

MEASUREMENTS

What data a local weather station provides

A typical home weather station measures several things at once. Each of them is useful for something different:

Temperature and humidity

The basis for heating, ventilation and frost warnings. Together they also give you the dew point, which tells you whether outdoor air would bring moisture into the house.

Air pressure

On its own it rarely controls anything, but its trend helps you spot a change in the weather. Stations usually measure absolute pressure, so it has to be converted to sea level before you compare it with a forecast.

Wind speed and gusts

Average wind speed and gusts are two different things. An awning or pergola is far more at risk from a short, strong gust than from a steady breeze, so both readings matter.

Wind direction

Lets you react only when the wind comes from a particular side, for example towards a west facing patio. Stations report it in degrees, and a label such as “north west” is created in Home Assistant.

Rain

Whether it is raining right now, and how much has fallen over time. The first is useful for quick reactions such as retracting an awning, the second for irrigation and statistics.

UV and light level

Shows how much sun is actually reaching the house. For blinds, awnings and overheating protection this is a better signal than the time of day or the forecast cloud cover.

MEASURED OR FORECAST

Your own measurements versus an online forecast

A forecast, for example from the default Met.no integration, describes the expected weather for an area. It is essential for planning: whether tomorrow will be sunny, whether frost is coming tonight, how much energy your solar panels might produce.

A station measures conditions at one specific spot. It cannot predict the future, but it knows what is happening now and does not depend on the internet. Local differences can be large: a house on a hill, next to woodland or near water can see very different wind from what the local forecast suggests.

The two sources work best together:

the forecast plans ahead, for example preheating the house before a freezing night or skipping irrigation when rain is expected;
the station reacts, for example retracting the awning in a gust or stopping irrigation when it has just started to rain;
comparing them shows how well the forecast matches the conditions at your house.

INTEGRATION

How to connect a weather station to Home Assistant: Zigbee, WiFi and Bluetooth

The most important decision is how the station communicates. Ideally the data reaches Home Assistant locally, without going through the manufacturer’s cloud. These are the sensible options:

For help choosing between ZHA and Zigbee2MQTT, see ZHA or Zigbee2MQTT. For more on what ESPHome can do, see ESPHome in Home Assistant.

Popular Tuya stations come in a WiFi version, which usually depends on the cloud, and a Zigbee version, which can often be connected locally. Tuya in Home Assistant explains the difference.

Zigbee

The station then works like any other sensor on your Zigbee network, through ZHA or Zigbee2MQTT. This is convenient if you already have a Zigbee network. Before buying, check that your chosen integration supports the model and which readings it exposes.

Bluetooth

Some stations broadcast their data over Bluetooth, for example in the open BTHome format. Home Assistant receives it through a Bluetooth adapter or, if the station is far from the server, through an ESPHome Bluetooth proxy placed closer to it.

WiFi and a local gateway

Many WiFi stations only send their data to the manufacturer’s app. The ones that work locally have their own gateway or let you set a custom server to receive the data. Classic Ecowitt kits, for example, can send data to Home Assistant on your local network. The Ecowitt integration receives it straight from the gateway. In the same way, the WeatherFlow integration receives data from a Tempest station locally.

433 or 868 MHz receiver

Many off the shelf stations transmit on 433 or 868 MHz to their own display. These signals can be picked up with an SDR receiver and rtl_433. It is an option for people who want to use a station they already own, but it takes more configuration.

ESPHome and your own sensors

An anemometer, rain gauge or light sensor can be connected to an ESPHome module. This gives you full control over the data, but the enclosure, power and weatherproofing are then up to you.

Cloud as a last resort

Some stations only make their data available through the manufacturer’s online service. The Netatmo integration is one example. That data arrives with a delay and disappears when the internet does, so it is not suitable for protecting an awning or any other quick reaction.

EXAMPLE DEVICE

Example: Ecowitt WS90 7-in-1 Powered by Shelly

A good example of a station designed with smart home systems in mind is the Ecowitt WS90 7-in-1. Shelly also offers the WS90 as a Powered by Shelly device, with Zigbee and Bluetooth built in. It is a single compact unit mounted on a pole, with no separate anemometer or rain gauge.

It measures temperature, humidity, air pressure, wind speed and gusts, wind direction, rain, UV index and light level. Wind is measured ultrasonically and rain by a sensor that detects the impact of raindrops, so there are no moving parts to jam. It is powered by a solar panel with lithium AA batteries as a backup, and an optional 12 V powered heater helps in winter with snow and ice.

The Powered by Shelly version sends its data over both at the same time. In Home Assistant it can be connected through Zigbee2MQTT, which as well as the basic readings also exposes calculated values such as dew point, rain rate and pressure trend. Users also read it over Bluetooth as a BTHome device when the station is within range of an adapter or Bluetooth proxy.

Bear in mind that its connectivity differs from the classic WS90, which works with an Ecowitt gateway. It does not work with Ecowitt gateways or displays, and which readings are available in Home Assistant depends on the station firmware and the integration version. Check this against current versions before deciding. I describe it here as an example of a solution, not as the only right choice.

USES

How to use weather station data in your home

Weather data on its own changes nothing. The value appears when Home Assistant combines it with the devices in your home:

Heating

The outdoor temperature measured at your house, wind strength and sunshine all affect how much heat you need. Rooms on the sunny side can be heated less on a bright day, and heating can start earlier in strong, freezing wind. For choosing the heat source, see Home Assistant and heating.

Blinds and awnings

An awning can retract in gusts and rain, and blinds can protect against overheating when light and temperature pass a set level. On freezing days it is worth blocking blinds that could freeze in place. The Blinds project shows an example panel.

Ventilation

Outdoor dew point and humidity help decide whether airing the house will dry it out or bring moisture in. In summer, cooler night air can be used to cool the house through the heat recovery unit’s bypass. See Ventilation in Home Assistant for details.

Irrigation

Rainfall over the last 24 hours and the forecast for the next few hours let you skip watering when the ground is already wet. A rain sensor can also stop watering that is already running. Soil moisture sensors give the most accurate picture when combined with this.

Solar PV

Light level shows the current cloud cover, while a production forecast, for example from the Forecast.Solar integration, plans the day. Together they help decide when to run larger loads. See PV surplus in Home Assistant.

Energy management

With dynamic tariffs, weather data helps decide when to heat the house or hot water with cheaper energy before a cold night or a cloudy day arrives. Energy in Home Assistant covers this in more depth.

AUTOMATIONS

Example automations with a weather station

The examples below show typical scenarios. The thresholds are only a guide and always need to be set for the specific house, devices and manufacturer recommendations:

Good weather automations have hysteresis, delays and clear behaviour when data is missing. How to build and troubleshoot such rules is covered in Home Assistant automations.

Awning in gusts

When gusts pass a set threshold, the awning retracts immediately. It can only extend again after a quarter of an hour or so of calmer wind, so it does not move with every gust.

Rain and roof windows

When the station detects rain, Home Assistant closes the roof windows, retracts the awning and sends a message to your phone saying which windows were open.

Overheating protection

On a summer day with high light levels, an outdoor temperature above a set value and the sun on a particular side of the house, the blinds on that side lower part of the way.

Sensible irrigation

Scheduled watering checks the rainfall over the last 24 hours and the forecast. If there has been enough rain or rain is certain, the cycle is skipped and you get a short message explaining why.

Frost warning

When the temperature at the house drops below a set value, Home Assistant reminds you about plants and the outdoor tap, and the blinds switch to a limited movement mode.

Night cooling

When it is clearly cooler outside than inside at night and the dew point is not too high, ventilation switches to cooling with fresh air and returns to its normal mode in the morning.

CHOOSING AND MOUNTING

What to consider when choosing and mounting a station

Even the best station gives poor data if it is badly chosen or badly mounted. Before buying and mounting one, think about these areas:

Local communication

Check that the data reaches Home Assistant without the cloud and that your chosen integration supports exactly this model. The same product name can cover versions with different connectivity.

Range

A station usually sits outdoors, often on a roof or pole. Zigbee needs routers nearby, and Bluetooth needs an adapter or proxy close to where the station is mounted. Test the range before you drill any holes.

Power and winter

A solar panel with a battery is convenient, but winter days are short. Batteries need to cope with frost, and where snow is common, heated wind and rain sensors are useful.

Mounting position

Wind is best measured high up and away from obstacles, and temperature in a well ventilated spot, not next to a warm wall or roof. An all in one station is always a compromise, so choose the position for the reading that matters most to your automations.

Sensors without moving parts

Ultrasonic wind measurement and a raindrop impact sensor do not jam and have no mechanism to clean. An impact sensor can, however, be less accurate in very light rain than a classic tipping bucket gauge.

Service access

A station occasionally needs cleaning, new batteries or an update. Mounting it somewhere you can reach safely saves trouble later.

LIMITATIONS

When weather data alone is not enough

A weather station greatly extends what your automations can do, but it has limits worth keeping in mind:

an awning or pergola should have protection that does not depend on Home Assistant, such as the manufacturer’s own wind sensor, because the server, the network or the station itself may be unavailable for a while;
when the station stops sending data, the automation needs to know what to do, ideally move to a safe state rather than assume the weather has not changed;
a station measures one spot, so the wind at a patio behind the building may differ from the wind at the pole;
planning for the next hours and days needs a forecast, because the station only sees current conditions;
soil moisture sensors are the best basis for irrigation and the inverter for PV production, with weather data as a supplement;
cheaper sensors have limited accuracy, and some stations briefly send wrong values after a restart or loss of connection, which need to be filtered out.

That is why designing weather automations starts with the question of what should happen when data is missing, and only then moves on to thresholds and scenes.

BEFORE YOU GET IN TOUCH

What to include when you write about a weather station

Before you write about choosing or integrating a station, it helps to gather a few details. They make it clear straight away where to start:

whether you already have a station, and if so which model and how it sends data today;
what hardware Home Assistant runs on and whether it already has Zigbee or Bluetooth;
where the station could be mounted and how far that is from the server or the nearest Zigbee devices;
what the station should control: awning, blinds, heating, ventilation, irrigation or energy;
which devices are already in Home Assistant and how you control them today.

HELP

Want the weather to actually control your home?

I can help you choose the right station for your home, integrate it locally with Home Assistant and put its data to work in automations for awnings, blinds, heating, ventilation, irrigation and energy, with safe behaviour when data is missing.

→ Get in touch

SOURCES

Sources and freshness

Checked on 30 September 2026 against current materials: