How to Build a Modular Weather Station: Sensors, Power and Mounting

How to Build a Modular Weather Station: Sensors, Power and Mounting

Most weather stations are sold as a fixed package. You get five parameters whether you need five or two, and if you later want soil moisture or solar radiation, you buy a second station.

A modular build works the other way round. You start with a mast and add the instruments the site actually needs. It costs more up front and less money over three years.

This guide covers what to measure, how to power the station where there's no outlet, and where each instrument has to sit to give readings worth trusting.

What a weather station measures

Six parameters cover most requirements. Which ones you need depends entirely on what decision the data is feeding.

Parameter

What it's for

Instrument

Air temperature and humidity

Baseline for everything else

Air temp and humidity probe

Rainfall

Irrigation scheduling, runoff, records

Tipping bucket rain gauge

Wind speed

Spray windows, structural load, evaporation

Anemometer

Wind direction

Drift risk, frost movement

Wind vane

Barometric pressure

Short-range forecasting, storm tracking

Barometric pressure sensor

Solar radiation

Crop modelling, greenhouse control, solar yield

Pyranometer

If you're irrigating, start with rainfall, temperature and humidity. If you're spraying, add wind speed and direction. If you're modelling crop growth or running a greenhouse, add solar radiation.

Three builds that make sense

Basic site record

Air temperature and humidity, plus rainfall. Two instruments, and enough for a farm to keep a defensible record of conditions. This is the build that answers "how much rain did we actually get on the north block" without argument.

Agricultural station

Add wind speed, wind direction, and solar radiation. Five parameters get you into evapotranspiration estimates, spray decision windows, and frost risk. This is where a weather station starts paying for itself in decisions rather than records.

Research or full met station

Add barometric pressure, and often soil sensors at the base of the mast.

Once you're logging pressure alongside the rest, you have a complete surface observation set. Soil moisture and temperature at 10cm and 30cm turn it into something that explains crop behaviour rather than just describing weather.

The weather station frame is galvanised steel and takes any combination of these, so the build can grow. Start at basic and add instruments as the questions get more specific.

Powering a station with no outlet

Weather stations sit in fields, on roofs and at the far end of sites. Mains power is usually the constraint that decides everything else.

Solar with battery backup is the standard answer. A solar panel rated IP67 handles the daily load, and the battery carries the station through nights and overcast runs.

Battery only works for short deployments and trial plots. A battery booster pack takes 4x18650 cells or 8xAA and delivers a stable 12V, which is what the RS485 instruments expect.

Mains is worth running if the station is within reach of a building. It's the cheapest and most reliable option by a distance, and cable is cheaper than call-outs.

Two things to size properly before ordering. Work out how many instruments are drawing from the supply, and check what your logging interval does to consumption. A station reporting every minute uses far more than one reporting every fifteen, and most sites don't need the minute.

Mounting: where each instrument has to sit

This is where most self-built stations go wrong. The instruments are fine. The siting isn't, and bad siting produces confident wrong numbers.

An anemometer and wind vane go at the top of the mast, clear of everything. The rule of thumb is at least twice the height of the nearest obstruction away from it. A wind sensor mounted beside a building measures the building, not the wind.

Rain gauge needs open sky above it and a level mount. Tilt it, and the tipping bucket calibration is wrong from day one. Keep it clear of overhanging trees and roof edges, which both funnel and shed water unpredictably.

Air temperature and humidity probe must sit inside a radiation shield, not in open air. Direct sun on a temperature probe can read several degrees high, and that error moves with the season, which makes it harder to spot than a constant offset.

Pyranometer must be levelled, mounted so nothing shades it at any point in the day, and kept clean. Check it against a spirit level at install and again after any storm.

Barometric pressure sensor is the easy one. It needs shelter from rain and nothing else. Pressure doesn't care where it sits.

Wiring the station together

Instruments that output RS485 Modbus connect to the logger's RS485 bus, which on a GS1 sits behind the 3.5mm audio connector.

One port serves multiple instruments through a splitter, with one rule that catches people: each port must carry a different instrument type. A rain gauge, an anemometer and a pyranometer on one splitter is fine. Two identical soil probes is not, because the logger can't tell them apart on the bus.

Cable runs are limited. Extension cables chain up to four deep for roughly 13 metres of total reach, which is enough for a mast and a soil pit but not for a sensor at the far side of a field. Plan the layout before you order, because a station that needs one more metre of cable than exists is a station you rebuild.

Five mistakes worth avoiding

  1. Mounting the anemometer too low. Any reading taken in a building's wind shadow is a reading about the building.
  2. Skipping the radiation shield. It's the difference between temperature data and sunshine data.
  3. Levelling nothing. Rain gauges and pyranometers both fail quietly when tilted.
  4. Logging too fast. Minute intervals drain batteries and rarely change a decision.
  5. Building for today's questions only. Leave mast space and bus capacity for the sensor you'll want next season.

Build the station your site actually needs

A fixed weather station gives you someone else's idea of what matters. A modular one gives you the parameters your decisions depend on, and room to add the rest later.

Start with the weather station frame and pick instruments from external sensors, or take a shortcut with the bundles if you'd rather begin from a configuration that's already matched.

Not sure which parameters your site needs, or how to power a mast at the far end of a field? Tell us what you're growing, storing, or managing at UbiBot USA and we'll spec the build with you. It's a shorter conversation than a rebuild.

FAQs

How many sensors can one weather station mast hold? 

Physically, as many as the frame takes. Electrically, six per splitter, provided each is a different instrument type. Most agricultural builds land between three and six.

Do I need solar power for a weather station? 

Only if mains isn't within reach, where a cable run is practical, mains is cheaper and more reliable. Solar with a battery is the standard answer for remote sites.

How high should the anemometer be mounted? 

As high as the mast allows, and at least twice the height of the nearest obstruction away from it. Comparability with published data improves the higher and clearer it sits.

Can I add sensors to the station later? 

Yes, that's the point of a modular build. Check the logger's remaining bus capacity and the mast's mounting space before ordering, and leave headroom for one more instrument than you need today.

Does the rain gauge need levelling? 

Yes. A tipping bucket mechanism relies on a level mount to tip at the correct water volume. Tilt introduces a permanent error that no calibration corrects.

What's the difference between a silicon and thermopile pyranometer? 

Silicon sensors measure a narrower part of the spectrum and cost far less. They're well suited to agriculture, greenhouse control and general monitoring. Thermopile instruments are what you need when a formal instrument class is specified.

Back to blog