WeatherFlow-2-MQTT for Home Assistant

BREAKING Announcement

As there is now a Home Assistant Core integration for WeatherFlow which uses the UDP API, I had to make a new Integration that uses the REST API, with a different name (WeatherFlow Forecast). The new integration is up-to-date with the latest specs for how to create a Weather Forecast, and also gives the option to only add the Forecast, and no additional sensors.

There is no Weather Entity in Home Assistant for MQTT, so after attributes are deprecated in Home Assistant 2024.3, there is no option to add the Forecast to Home Assistant. As a consequence of that, I have decided to remove the ability for this Add-On to add Forecast data to MQTT and Home Assistant. This Add-On will still be maintained, but just without the option of a Forecast - meaning it will be 100% local. If you want the forecast in combination with this Add-On, install the new integration mentioned above, just leave the Add sensors box unchecked.

Overview

There is not an exact date for when this will happen, but it will be before end of February 2024. This project monitors the UDP socket (50222) from a WeatherFlow Hub, and publishes the data to a MQTT Server. Data is formatted in a way that, it supports the MQTT Discovery format for Home Assistant, so a sensor will created for each entity that WeatherFlow sends out, if you have MQTT Discovery enabled.

Everything runs in a pre-built Docker Container, so installation is very simple. You only need Docker installed on a computer and a MQTT Server setup somewhere in your network. If you run either the Operating System or Supervised installation of Home Assistant, you will have easy access to both.

There is support for both the AIR & SKY devices and the TEMPEST device.

Please review Breaking Changes prior to updating your instance. Breaking changes will be listed in https://github.com/briis/hass-weatherflow2mqtt/blob/main/CHANGELOG.md

Table of Contents

1. Installation
2. Docker Setup
   1. Docker Volume
   2. Docker Environment Variables
3. Available Sensors
   1. Sensor Structure
   2. High and Low Values
   3. Long Term Statistics
   4. Sensors in Developement
4. Creating a Home Assistant Weather Entity
5. Creating a Developement Environment

Installation

Prerequisites

A MQTT Server must be setup and configured on your local network, before installing this component. If you run a Supervised Home Assistant installation, we recommend using the MQTT Add-On available for this, else you must set it up yourself. See more here for installing MQTT.

NOTE: WeatherFlow2MQTT will not discover your base station if it is on a separate VLAN/subnet from your Home Assistant (or Docker) instance.

Home Assistant Supervised

This is the easiest installation method. Just click on the blue ADD REPOSITORY button at the top of this README and you will be taken to your Home Assistant instance to add this repository to the Add-On Store. From there, scroll down to find "WeatherFlow to MQTT" and then install, configure (optional) and start the add-on.

NOTE: Make sure that you already have setup a Working MQTT broker, and easiest way to do that is by using the MQTT broker from the Official Add-On store.

Outside Home Assistant using Docker

• Ensure Docker is setup and running
• Ensure there is a MQTT Server available
• Open a Terminal on the Machine you want to run the Docker container on.
• Make a new Directory: mkdir weatherflow2mqtt (or some other name) and change to that directory.
• If you don't want all sensors setup, copy the config_example_TEMPEST.yaml or config_example_AIR_SKY.yaml file from this repo to that directory, and rename it to config.yaml. Then add or remove the sensors you want from the available sensors list. If you don't do this, all sensors from the Available Sensors will be added.
• Now start the Docker Container with the parameters described under docker-setup

If everything is setup correctly with MQTT and Home Assistant, you should now start seeing the sensors show up in HA. NOTE, it can take up to 1 min after startup, before all sensors are populated with data.

Docker Setup

The below command will pull the latest docker image and start WeatherFlow2MQTT for timezone Europe/Copenhagen and save data in the directory you are placed in when you launch the command. Ensure that you have replaced the Environment variables with your specific data. See description below.

If you are using docker-compose you can use the docker-compose.yml file and make the modifications for your environment.

docker run -d \
--name=weatherflow2mqtt --restart=unless-stopped \
-v $(pwd):/data \
-p 0.0.0.0:50222:50222/udp \
-e TZ=Europe/Copenhagen \
-e UNIT_SYSTEM=metric \
-e LANGUAGE=en \
-e RAPID_WIND_INTERVAL=0 \
-e DEBUG=False \
-e ELEVATION=0 \
-e LATITUDE=00.0000 \
-e LONGITUDE=000.0000 \
-e ZAMBRETTI_MIN_PRESSURE=960 \
-e ZAMBRETTI_MAX_PRESSURE=1060 \
-e WF_HOST=0.0.0.0 \
-e WF_PORT=50222 \
-e MQTT_HOST=127.0.0.1 \
-e MQTT_PORT=1883 \
-e MQTT_USERNAME= \
-e MQTT_PASSWORD= \
-e MQTT_DEBUG=False \
-e STATION_ID= \
-e STATION_TOKEN= \
-e FORECAST_INTERVAL=30 \
briis/weatherflow2mqtt:latest

The container is build for both Intel and ARM platforms, so it should work on most HW types. Please create an issue, if there is a platform for which it does not work.

Docker Volume

-v YOUR_STORAGE_AREA:/data Please replace _YOUR_STORAGEAREA with a directory where Docker will have read and write access. Once the program runs, a SQLite Database with the name weatherflow2mqtt.db will be created in this directory. This database is used to hold some calculated values, store temporary data used for calculations and to ensure that you don't start from 0 if you have to restart Home Assistant or the container. This is also where you can place the config.yaml file if you don't want all the sensors (see Installation).

Docker Environment Variables

A description of the Environment Variables available for this container. All of them have a default value, so you only need to add the onces where you want to change that.

• TZ: Set your local Timezone. It is important that you use the right timezone here, or else some of the calculations done by the container will not be correct. Default Timezone is Europe/Copenhagen (Required)
• UNIT_SYSTEM: Enter imperial or metric. This will determine the unit system used when displaying the values. Default is metric
• LANGUAGE: Use this to set the language for Wind Direction cardinals and other sensors with text strings as state value. These strings will then be displayed in HA in the selected language. See section Supported Languages
• RAPID_WIND_INTERVAL: The weather stations delivers wind speed and bearing every 2 seconds. If you don't want to update the HA sensors so often, you can set a number here (in seconds), for how often they are updated. Default is 0, which means data are updated when received from the station.
• ELEVATION: Set the hight above sea level for where the station is placed. This is used when calculating some of the sensor values. Station elevation plus Device height above ground. The value has to be in meters (meters = feet * 0.3048). Default is 0
• LATITUDE: Set the Latitude where the Station is located. Default is 00.0000.
• LONGITUDE: Set the Longitude where the Station is located. Default is 000.0000.
• ZAMBRETTI_MIN_PRESSURE: All Time Low Sea Level Pressure. Default is 960 (Mb for Metric) or Default is 28.35 (inHG for Imperial)
• ZAMBRETTI_MAX_PRESSURE: All Time High Sea Level Pressure. Default is 1060 (Mb for Metric) or Default is 31.30 (inHG for Imperial)
• WF_HOST: Unless you have a very special IP setup or the Weatherflow hub is on a different network, you should not change this. Default is 0.0.0.0
• WF_PORT: Weatherflow always broadcasts on port 50222/udp, so don't change this. Default is 50222
• MQTT_HOST: The IP address of your mqtt server. Even though you have the MQTT Server on the same machine as this Container, don't use 127.0.0.1 as this will resolve to an IP Address inside your container. Use the external IP Address. Default value is 127.0.0.1 (Required)
• MQTT_PORT: The Port for your mqtt server. Default value is 1883
• MQTT_USERNAME: The username used to connect to the mqtt server. Leave blank to use Anonymous connection. Default value is blank
• MQTT_PASSWORD: The password used to connect to the mqtt server. Leave blank to use Anonymous connection. Default value is blank
• MQTT_DEBUG: Set this to True, to get some more mqtt debugging messages in the Container log file. Default value is False
• DEBUG: Set this to True to enable more debug data in the Container Log. Default is False
• STATION_ID: Enter your Station ID for your WeatherFlow Station. Default value is blank. The correct STATION_ID is the number that you see when you access your Station from the Tempest Web APP. For example when you are on https://tempestwx.com/station/XXXXX/
• STATION_TOKEN: Enter your personal access Token to allow retrieval of data. If you don't have the token login with your account and create the token. NOTE You must own a WeatherFlow station to get this token. Default value is blank
• FORECAST_INTERVAL: The interval in minutes, between updates of the Forecast data. Default value is 30 minutes.
• briis/weatherflow2mqtt:<tag>: latest for the latest stable build, dev for the latest build (may not be stable due to development/testing build). Once dev build is verified latest build and dev will be identical. Latest features will be tested in dev build before released to latest.

Supported Languages

Currently these languages are supported for Wind Cardinals and other Text state strings:

• en: English
• da: Danish
• de: German
• fr: French
• nl: Dutch
• se: Swedish

If you would like to assist in translating to a new language, do the following:

• From the translations directory on this Github Project, download the file en.json
• Rename the file to YourLanguageCode.json - example for Spanish rename it to es.json
• Edit the file and translate the strings
• Make a pull request in Github and attach the file.

Available Sensors

Here is the list of sensors that the program generates. Calculated Sensor means, if No, then data comes directly from the Weather Station, if yes, it is a sensor that is derived from some of the other sensors. For a copy ready list see further below

Sensor Structure

See Available Sensors above for a description of each sensor.

sensors:
  - absolute_humidity
  - air_density
  - air_temperature
  - battery # voltage
  - battery_mode # support for Tempest devices only
  - beaufort
  - cloud_base
  - current_conditions
  - current_conditions_txt
  - delta_t
  - dewpoint
  - dewpoint_description
  - feelslike
  - fog_probability
  - freezing_level
  - illuminance
  - lightning_strike_count
  - lightning_strike_count_1hr
  - lightning_strike_count_3hr
  - lightning_strike_count_today
  - lightning_strike_distance
  - lightning_strike_energy
  - lightning_strike_time
  - precipitation_type
  - pressure_trend
  - rain_intensity
  - rain_rate
  - rain_start_time
  - rain_today
  - rain_yesterday
  - rain_duration_today
  - rain_duration_yesterday
  - relative_humidity
  - sealevel_pressure
  - snow_probability
  - solar_elevation
  - solar_insolation
  - solar_radiation
  - station_pressure
  - status
  - temperature_description
  - uv
  - uv_description
  - visibility
  - voltage
  - wbgt #Wet Bulb Globe Temperature
  - wetbulb
  - wind_bearing
  - wind_bearing_avg
  - wind_direction
  - wind_direction_avg
  - wind_gust
  - wind_lull
  - wind_speed
  - wind_speed_avg
  - weather
  - zambretti_number
  - zambretti_text

High and Low Values

For selected sensors high and low values are calculated and published to the attributes of the sensor. Currently daily, monthly and all-time values are calculated, but future values are planned. Only the sensors where it is relevant, will get a low value calculated. See the table further down, for the available sensors and what values to expect.

Here are the current attributes, that will be applied to the selected sensor:

The following sensors are displaying High and Low values:

Home Assistant Long Term Statistics

These sensors have long term statistics (LTS) enabled in Home Assistant (HA), no action by user is required to have LTS collected by HA this is done with MQTT autodiscovery. To display LTS in HA using the Statistics Graph Card (type: statistics-graph) the 'stat types' that can be used with this integration are Mean, Min, and Max. It is the user's discreation as to how to display the information, but cases such as rain today 'Max' would be a more logical selection without 'Mean' or 'Min'. During testing there has not been a use case for this integration that 'Sum' was of any valid use. Air, Sky, and Tempest sensors are specific to those units.

https://data.home-assistant.io/docs/statistics/

air_density
battery
beaufort
delta_t
dewpoint
feelslike
relative_humidity
illuminance
lightning_strike_count_today
lightning_strike_distance
lightning_strike_energy
rain_duration_today
rain_rate
rain_today
sealevel_pressure
solar_elevation
solar_insolation
solar_radiation
station_pressure
air_temperature
uv
visibility
voltage
wet_bulb_globe_temperature
wetbulb
wind_bearing
wind_bearing_avg
wind_gust
wind_lull
wind_speed
wind_speed_avg

Sensors in Developement

Local Current Conditions in Home Assistant format (Dev Branch)
Local Current Conditions in Human Readable Text (will need translations)
Probability of Snow (Dev Branch)
Probability of Fog (Dev Branch)

Creating a Home Assistant Weather Entity

If you have enabled the Forecast option, then there is a possibility to create a Weather Entity that can be used in all the different Lovelace Cards available for Weather. You can do this by using the Weather Template.

Edit configuration.yaml and insert the following (replacing hub_hb_00000001 and tempest_st_00000001 with the appropriate ids of your sensors):

weather:
  - platform: template
    name: My Local Weather
    condition_template: "{{ states('sensor.hub_hb_00000001_weather') }}"
    # condition_template: "{{ states('tempest_st_00000001_current_conditions') }}" # Local Only Option
    temperature_template: "{{ states('sensor.tempest_st_00000001_temperature') | float}}"
    humidity_template: "{{ states('sensor.tempest_st_00000001_humidity')| int }}"
    pressure_template: "{{ states('sensor.tempest_st_00000001_sea_level_pressure')| float }}"
    wind_speed_template: "{{ ( states('sensor.tempest_st_00000001_wind_speed_avg') | float * 18 / 5 ) | round(2) }}"
    wind_bearing_template: "{{ states('sensor.tempest_st_00000001_wind_bearing_avg')| int }}"
    visibility_template: "{{ states('sensor.tempest_st_00000001_visibility')| float }}"
    forecast_template: "{{ state_attr('sensor.hub_hb_00000001_weather', 'hourly_forecast') }}"

• The weather entity expects km/h when using metric units, so the above example converts m/s to km/h. If you are using imperial units, the line should just be {{ states('sensor.tempest_st_00000001_wind_speed_avg') }}, again replacing tempest_st_00000001 with the appropriate id of your sensor
• For the _forecasttemplate you can either use hourly_forecast or daily_forecast to get Hourly or Day based forecast.

Setup Dev environment

virtualenv -p `which python3` env
source env/bin/activate
python setup.py install

Then you just need to export the configuration

export TZ="America/Toronto"
export UNIT_SYSTEM="metric"
export LANGUAGE="en"
export RAPID_WIND_INTERVAL="0"
export DEBUG="True"
export EXTERNAL_DIRECTORY="."
export ELEVATION="30"
export WF_HOST="0.0.0.0"
export WF_PORT="50222"
export MQTT_HOST="..."
export MQTT_PORT="1883"
export MQTT_DEBUG="False"
export STATION_ID="..."
export STATION_TOKEN="..."
export FORECAST_INTERVAL="30"
export MQTT_USERNAME="..."
export MQTT_PASSWORD="..."

Then you can run the daemon with

weatherflow2mqtt