What a home weather station adds to your irrigation
A soil moisture sensor tells you what the substrate is like right now. A weather station tells you why. The difference matters because irrigation is decided with two questions: “is there enough water?” and “how much is being lost?”. The first is answered by the soil sensor; the second by temperature, air humidity, wind and radiation.
Wind is the variable that surprises people most: a plant on a windy balcony uses more water than the same plant sheltered two metres from a wall, without the thermometer changing at all.
The four variables that really matter
| Variable | Why it matters | Where to measure it |
|---|---|---|
| Air temperature | Sets the rate of evapotranspiration | In the shade, at 1.5 m |
| Air humidity | When it drops, the plant transpires more | In the shade, next to the thermometer |
| Wind | Multiplies water loss from the leaves | Where there’s no turbulence from buildings |
| Rain | Supplies the water you won’t irrigate | Open, with no obstructions |
Radiation and atmospheric pressure are useful for evapotranspiration models, but they don’t justify a home sensor in a garden: reference stations exist for that.
Where to put each sensor, which is 80 % of the work
- Thermometer and air humidity, in the shade. A sensor in the sun reads the temperature of the sensor, not of the air. A car thermometer mounted on a balcony reads fifteen degrees more than the one in the living room.
- At the height of the plants. If you measure at two metres and then decide the irrigation for a low shrub, the decision doesn’t match the plant.
- Wind, out of the direct draught. Next to a building corner or behind a wall, the wind you measure isn’t the wind the plant suffers.
- Rain gauge, in the open, with the board’s casing protected.
- Everything together in the same spot if you can. Scattered around the garden, the sensors tell different stories.
If you only have one enclosure for several sensors, the shade of the enclosure and the position of the cables matter more than the model of each part.
Temperature and air humidity: the easy part
A temperature and relative humidity sensor on its own node is the usual starting point: it’s cheap, it runs on a battery and it integrates with the same ESPHome you use for the rest of the house. The BME280 wiring and calibration guide is a good reference for how an environmental sensor is documented and calibrated.
Mount it at leaf height, on a ventilated support, inside a casing that acts as a solar shield.
Rain: a rain gauge of its own
The rain gauge deserves its own article: it needs open space, height and a time margin to make a decision. What changes here is that in a weather station rain doesn’t just decide whether to irrigate: the total is also stored and compared with what the irrigation delivered. That comparison is what reveals leaks, clogging in the filter and lines that irrigate less than you think.
Wind: the sensor nobody fits and the one that saves the most
A professional anemometer is expensive; a wind sensor with a water output, the kind used in garden sprinklers, gives you all the information you need to make a decision on a balcony. The goal isn’t meteorological accuracy but detecting two things: that there’s wind, and how much.
With that variable you can make a simple, very profitable rule: if the wind speed exceeds a threshold, reduce the irrigation run time, because part of the water evaporates before it reaches the root.
From data to decisions: evapotranspiration
With temperature, humidity, wind and radiation you can calculate evapotranspiration, which is the water the plant consumes plus what evaporates from the soil. FAO-56 describes the reference method: a weather demand multiplied by a coefficient that depends on the crop and its stage.
Applied to a garden, the full method is unnecessary, but the idea is very useful: water demand isn’t constant, it can be calculated. If the crop coefficient for an ornamental shrub in your area is, say, 0.4 and today’s reference evapotranspiration is 4 mm, the plant has used around 1.6 mm. That’s exactly the criterion you adjust the controller with in summer.
Data from the surroundings: AEMET without building anything
You don’t have to measure everything: for the weather in your area, AEMET OpenData offers precipitation and temperature data from nearby stations. Integrating it into the dashboard turns irrigation into a decision with context: if it rained yesterday at your nearest station, today’s irrigation is skipped; if it hasn’t rained for three weeks, it’s extended.
The limit is that the reference station may be far away and its data isn’t the data for your street. It’s useful for the total volume of rain in your garden; to know whether the rain reached the root, you need the soil sensor.
When a home weather station doesn’t pay off
If you have three pots and you water them by hand, you don’t need any of this. A weather station starts to pay for itself when there are several zones with different plants, or when manual scheduling fails you often. Before that, weighing the pots gives you more information for less money.
Sources and next step
FAO-56 is the reference for calculating water demand from weather variables, and the University of Michigan’s smart irrigation guide explains how to combine soil sensors and weather data. To connect your nodes to the dashboard, the ESPHome integration in Home Assistant.
External sources
- FAO: Crop evapotranspiration, Paper 56 — reference method for estimating evapotranspiration from weather variables.
- MSU Extension: soil sensors for smart irrigation — combining soil sensors and weather data for irrigation.
- AEMET OpenData — temperature and precipitation data from reference stations.