The science behind on-farm weather stations for precision irrigation

Australian growers know that rain is not the same as stored water. A downpour can run off saturated soil, while a warm, windy week can lift crop water demand before the rain gauge records a single drop. In the Far North Queensland growing districts around Tully and the Atherton Tablelands, that difference matters as Fresh Del Monte manages crops with large, leafy canopies.

Australia’s climate can also shift sharply between seasons. Cool, humid conditions near Cairns behave very differently from hot, dry winds across the Darling Downs, and irrigation requirements change as a banana or pineapple canopy develops. Local measurements therefore provide a better basis for daily water decisions than relying only on a distant regional forecast.

An on-farm weather station combines sensors that record rainfall, air and soil temperature, humidity, wind, solar radiation and, depending on the installation, soil moisture or leaf wetness. Readings are logged at regular intervals and paired with the crop calendar, soil characteristics and forecast conditions. This turns the growing landscape into a sequence of measurable decisions.

Fresh Del Monte places this work within its wider focus on responsible use of natural resources and communities, described in its sustainability commitments. The value is practical: better evidence can reduce unnecessary pumping, support more consistent crop quality and help farming decisions remain viable through dry periods.

What the stations measure

The basic weather observations describe the atmosphere above the crop, but the most useful stations also measure conditions at the root zone. Soil-probe sensors estimate how readily water is available to roots, while soil temperature indicates how quickly roots can take up moisture. Rainfall totals show what has reached the surface; they do not reveal how much entered the root zone or was lost to runoff.

Placement affects the result. A rain gauge needs a clear, level exposure away from trees and irrigation spray, and air-temperature sensors require appropriate shielding. Soil probes should reflect the main rooting depth and the soil type used by the crop. A station placed in an unusually wet hollow or directly beside a pump can produce readings that look precise but represent neither the field nor the farm.

A second layer is crop and management context. Canopy cover changes solar interception and airflow, while fruit development, flowering and root growth alter water demand. Digital field records can link station data with irrigation dates, fertiliser applications and crop observations. The result is not a single number but a defensible picture of what is happening in the field.

Turning measurements into irrigation decisions

Agricultural weather data is commonly used to estimate reference evapotranspiration, the evaporation from a reference surface and transpiration from a representative crop. The widely used FAO-56 Penman–Monteith method combines solar radiation, temperature, humidity and wind. Because each input comes from the station, sensor quality and siting influence the final estimate.

Crop water use is then adjusted for the particular planting. A crop coefficient reflects differences between the reference surface and the crop’s canopy, leaf area and growth stage. The estimated crop evapotranspiration is compared with available root-zone water, incoming rain and the water already stored in the soil. When depletion reaches a managed threshold, the system can identify an irrigation date, duration and volume that suit the field rather than a blanket schedule.

Different approaches expose different gaps. A fixed calendar may suit a stable pattern, but it cannot respond to a heatwave or an unexpected storm. Weather alone indicates demand, yet it may miss water held below the surface. Soil-moisture readings reveal storage, but they need weather data to explain why it is changing.

Combining the methods turns an alarm into a plan. If a hot forecast raises expected demand but the root zone still holds adequate moisture, the response can be to check again rather than irrigate immediately. If stored water is already low and evapotranspiration remains high, the decision becomes one of timing and application rate before stress develops.

Why Australian conditions demand local data

The Far North Queensland example shows why local data matters. Heavy tropical rain can quickly recharge the topsoil, yet drainage and warm temperatures may move water below the roots within days. At the same time, cool nights, humidity and dense banana foliage can slow evaporation. An irrigation model that relies on long-term regional averages may miss these short cycles and recommend either too much water or too little.

Conditions across the Darling Downs and Swan Valley present a different set of variables. Strong heat and wind can produce high water demand, while cooler nights can slow crop growth. Growers in these districts often talk about whether forecast rain is “good rain”: useful enough to infiltrate and replace root-zone depletion, rather than merely wetting the surface. On some farms, salinity and water quality also influence how irrigation water should be applied.

Water availability adds another Australian reality. In the Murray–Darling Basin, entitlements, allocations and temporary water purchases can shape the cost of every megalitre used. Around the Darling Downs, growers must do the sums between crop value, pumping and water-market prices. Station information cannot guarantee cheap water or predict every allocation decision, but it can show when a field has reached a defensible irrigation threshold and when rain has already met part of that requirement.

Keeping the data trustworthy

A station is useful only when its record is fit for a decision. Sensors need routine calibration, cleaned rain gauges and checks for damaged cables or drifting values. Software should flag missing periods, implausible jumps and unusually large differences between nearby probes. Maintenance is particularly important around machinery, vegetation growth and irrigation infrastructure, all of which can change a sensor’s exposure over time.

Telemetry and agronomic review add another layer. Automatic alerts can identify a dry root zone or rising water demand quickly, but an agronomist still needs to consider crop stage, soil structure, drainage and recent field work. Data governance also matters: access should follow the farm’s operational requirements, and decisions should retain an audit trail showing which observations supported each irrigation event.

The same engineering principle appears in other industries. In cement production, secondary impact crushers depend on consistent material and process conditions; likewise, irrigation data is only as dependable as the instruments, maintenance and context behind it. Trustworthy records do not remove uncertainty, but they make uncertainty visible and manageable.

Using field intelligence to conserve water

Precision irrigation is not simply a promise to save a fixed percentage of water. Its value lies in matching applications more closely to crop demand, reducing losses that the field can avoid and improving resilience when rainfall is unreliable. For Fresh Del Monte, that can support more deliberate planning across regions with different climates, soils and water constraints while maintaining the quality and consistency expected of fresh produce.

The strongest programmes treat each irrigation event as part of a feedback loop. After water is applied, the team checks soil moisture, crop condition and weather data to see whether the application achieved the intended result. The next threshold or coefficient can then be adjusted. This approach also supports employee learning, since field observations and instrument records are compared rather than treated as competing sources of truth.

The practical takeaway is to irrigate from measured crop demand, verified root-zone availability and local weather, then document the result and adapt as the crop, soil and forecast change.