A new tool for irrigation management

When it comes to irrigation scheduling, a farmer shouldn't have to rely on guesswork. It may rain tomorrow, or the next day, but often it's easier for a farmer to play safe and irrigate, rather than risk missing out on the water.

But what if that farmer could have instant access to soil moisture information and a reliable weather forecast tailored specifically to his farm? What if he could confidently know how much rain would fall, and when?

NIWA and five Canterbury farmers taking water from the Waimakariri Irrigation Scheme are working together with that aim in mind. The project is giving participating farmers new information that is helping them to make informed decisions about when, and how much, to irrigate.

The challenge

NIWA's goal is to find out what weather and soil-moisture information will be of most benefit to farmers, and the best way to get it to them so they can plan effectively.

Dry-land farming depends heavily on irrigation to maintain productivity. To make best use of it, farmers need to know the moisture content of their soil, when it's going to rain and how much will fall. They also need to know how much irrigation water is available to them at any given time, especially during dry periods.

To that end, NIWA researchers, and five members of the Waimakariri Irrigation Scheme, are working on the 'IM Tool', which generates customised, farm-specific weather forecasts to help farmers manage water-use better.

The benefits

The potential benefits are huge: not just to the farmer, who can reduce power, maintenance and operational costs and ensure he stays within his allocation limits, but also to the environment and sustainability of the national water budget. If water isn't needed, it's better stored, left in the river, or reallocated.

The IM Tool can help farmers make a range of farm-management decisions such as irrigation scheduling, stock movement and water-storage planning.

It can enable a farmer to optimise the scheduling of fertiliser and manure applications, to reduce the risk of them being washed into waterways by unexpected rain. But if drizzle is forecast, the farmer may consider applying urea, as drizzle can help reduce volatile losses by dissolving and carrying it into the soil.

How does the IM Tool work?

The IM Tool generates a daily report containing farm-specific weather forecasts, and automatically emails it to subscribing farmers.

The report is generated from 'NIWA forecast', NIWA's environmental forecasting and information service (powered by technology called EcoConnect), through its subscriber interface.

The on-farm forecasts are generated by a computer model that 'learns' how to predict local conditions. The model assimilates data from many sources, including NIWA's Climate Database, which archives the data recorded by nearly 200 weather/climate monitoring stations. These include stations in NIWA's National Climate Network, and a number of shared sites.

NIWA runs a model called the New Zealand Limited Area Model (NZLAM) that predicts the weather conditions for a particular cube of atmosphere above the Earth's surface, at any time.

NZLAM provides site-specific weather forecasts for more than 100 locations around the country. It then uses them to generate discrete forecasts for locations as close as 12km apart. NIWA is testing a new higher-resolution model, NZCONV, which will generate forecasts for points on a grid as close as 1.5km apart – enabling better handling of topographical features. The model has to 'run faster than the weather', so it's processed on NIWA's IBM supercomputer.

Local geography can produce local weather that may differ significantly from the wider forecasts generated by NZLAM, so we can improve forecast resolution still further. Here's how:

We install local weather stations as close as possible to the farm or region of interest.

Data from these stations provide the 'correction' information needed to 'localise' NZLAM's forecasts down to farm scale. NZLAM's raw forecast is compared with what the weather station actually recorded at the time, and the difference is used to correct the model. This process of fine-tuning effectively 'teaches' the model and so continuously improves its accuracy.

With this local correction, NZLAM can provide probabilistic weather forecasts down to less than half a kilometre, for periods out to 15 days.

On-farm IM Tool trials

We're currently trialling the IM Tool at four Canterbury farms: West Eyreton, Carleton, South Eyre and Burnt Hill.

We record rainfall, irrigation, soil moisture and soil temperature data at each farm.

Every day we email a customised, site-specific forecast to each farmer to show how local conditions are expected to change over the forecast period.

NIWA meets regularly with the farmers to subjectively review the results and gain a better understanding of the ways in which they are getting value from the IM Tool.

Who is involved?

As part of the Primary Innovation Project, the IM Tool project involves scientists from primary sector research organisations, growers, farmers and foresters. The project is funded by the Ministry of Business, Innovation and Employment (MBIE).


placeholder image
Principal Technician - Instrument Systems
placeholder image
Principal Scientist - Catchment Hydrology
Irrigator. [Dave Allen]
An example of a 2-day weather forecast for rainfall (green – none in this example), soil temperature (red) and relative humidity (blue)
An example of a 6-day weather forecast for rainfall (green) and soil temperature (red).
An example of a 15-day weather forecast for mean wind speed (blue) and rainfall (green) taken from a report.
An example of simultaneous conditions at Larundel and Tinomanas Farms taken from a report.