A Canadian cranberry farm watches frost and soil tension from one solar station in the bed

The situation

A cranberry bed is a dense, low mat of vines growing on sand, and it sits in the lowest ground available. That is deliberate, because the beds are flooded for harvest and again for winter protection. It also means the bed is exactly where cold air ends up on a still night, since cold air is dense and drains downhill like water.

Cranberries are most vulnerable to frost around bloom and fruit set, and the standard defence is water: run the sprinklers, and as that water freezes on the plant it releases latent heat and holds the tissue at a survivable temperature. It works, and it is expensive to run and expensive to skip. Protecting on a night that did not need it costs a night of pumping. Not protecting on a night that did can cost the crop.

So the decision comes down to a number, and the number that matters is the air temperature just above the vines, in a specific bed, tonight. A regional forecast cannot supply that. Neither can a weather station a few kilometres away, or in many cases the bed next door.

What was installed

One Omni Genesis controller on a wooden stake set into the bed, with an SHT30 temperature and humidity sensor reading the air at canopy height and Irrometer tensiometers reading soil water tension below. A small solar panel on the same stake powers it, and reporting is cellular.

Everything about that arrangement is dictated by the bed. The station has to be short, because the crop is. It has to be staked, because there is nothing to mount it on. It has to be self-powered, because running mains out across a bed that gets flooded twice a year is not a serious proposition.

Two measurements, one trip

The temperature sensor is the reason the station exists, and accuracy is the reason the sensor choice matters. The whole decision lives within a couple of degrees of one threshold, so an instrument with a degree or two of error is not a minor imprecision, it is the difference between the right call and the wrong one. The SHT30 is rated to ±0.3 °C typical, which puts the measurement error comfortably inside the margin the grower is deciding on.

Humidity is on the same chip and answers the other half of the question. Temperature tells you where you are; the moisture in the air tells you where the night is going. A dry night with a low dew point keeps falling steeply. A humid one slows down sharply once the air nears saturation and condensation starts releasing heat. Two nights that look identical at nine in the evening can finish several degrees apart, which is why a frost decision made on temperature alone is being made on half the information.

Soil water tension is the third reading, and it is there because the station is already there. Cranberries are shallow-rooted on sand, so the profile holds little water and its status moves quickly, which makes tension worth following through the season in its own right. Adding it to an existing station costs a sensor and a cable instead of a second installation. That is the practical case for a controller that accepts any sensor rather than one built around a single measurement.

Why it matters

This is the same hardware that sits in vineyards and fish farms, staked into a bog in Canada and asked to do something quite different with it. Nothing about the controller is specific to the crop or the climate; the sensors define what the station is for, and here they make it a frost instrument that also happens to be an irrigation instrument.

It is also a reminder of what continuous measurement is actually for in horticulture. Nobody needs a temperature reading at two in the afternoon in June. They need one at three in the morning in the third week of bloom, in the low corner of the property, on the one night that decides the season. That reading has to be taken automatically, because the alternative is a person standing in a bog with a thermometer, and there is only one of them and several beds.