Betting the farm

New Zealand’s rain fuels world-leading agriculture, but the challenge for farmers is when and how much? Mark Blackham went to Fieldays this year to find out how technology is helping the farmer understand weather.

Experience meant everyone ‘knew’ that it was going to rain this year at farming’s biggest gathering of products and services, Fieldays at Mystery Creek near Hamilton. But the question of exactly when it would rain, and how much, was answered by Chris Brandolino of NIWA, the official forecaster for the event.

Using a new service called FarmMet, Brandolino was able to forecast the weather for Mystery Creek – at 112 hectares, about the size of a small dairy farm – with a high degree of accuracy.

His regular forecasts were broadcast around the site every few hours. And there was plenty of rain – 112.6mm in total, including Tuesday, the non-public day. Thursday, was the wettest, with 45mm of rain.

Where once a weather vane and wisdom of the almanac helped farmers prepare for rain, advances in weather science are fuelling a revolution in farm use of weather forecasting.

The advances come from a combination of new weather monitoring technology including data from weather stations to satellites, steadily improving forecast models, and computers powerful enough to calculate local forecasts.

It has been shown in the US that today’s two-day forecast is as accurate as the one-day forecast was in 1988. The seven-day forecast now is as accurate as the five-day forecast was then.

The accuracy of modern forecasting is worth real money to farmers. If you know the moisture content of soil and the amount of rain due tomorrow, if any, you can decide when and how much to irrigate, apply fertiliser, and even judge stock levels on paddocks. This maximises grass and crop growth, and minimises expenditure on fertiliser, electricity, water, and hardware maintenance.

The whole supply chain of the agricultural sector benefits from improved weather forecasting. The range of decisions affected by weather and climate is considerable; from scheduling of planting/harvest operations, choices of crops and herds, crop sequencing and rotations or stocking rates, land use, insurance and finance arrangements, and processing and storage.

Economic studies into the value of weather forecasting have found they can add 1 – 2% to the value of agricultural output.  The studies are also indicating a new dimension to more accurate forecasts: anticipation of economic responses. Farmers aware of impending dry or wet periods can adjust production ahead of the supply and demand changes that are likely to arise.

In addition, forecasts enable farmers to reduce the costs of weather disasters by taking action ahead of them.

Advances in the science of weather forecasting are now attracting big money. Global giant Monsanto last year paid $930 million for The Climate Corporation, which specialises in local weather prediction and analytics. Monsanto estimates that there is $20 billion of “untapped yield opportunity,” if farmers use “data science.”

NIWA’s similar service, FarmMet, was launched at Fieldays this year. It is based on the best international science, tailored by New Zealanders for New Zealand to provide forecasts localised to the customer’s farm.

Finessing the forecast

NIWAFarmMet’s accurate, near farm-scale weather forecasts are the outcome of complex number crunching using extraordinary computational power, multiple local and global data sources, and models that continually check against locally recorded data and adjust their outputs accordingly.

“NIWAFarmMet draws on a number of different weather ‘models’ – highly sophisticated computer programmes – to produce local two-day, six-day and two-week forecasts for farmers,” explains NIWA’s Chief Scientist, Climate and Atmosphere, Dr Murray Poulter.

“Our own high-resolution model for New Zealand, run on our supercomputer in Wellington, provides the two-day forecasts. They’re the most detailed and accurate in the NIWAFarmMet package. Medium-resolution six-day forecasts are delivered via a global model. Two-week forecasts are the combined and ‘downscaled’ outputs from an ensemble of models, and provide a low-resolution ‘heads up’ of weather expected during the fortnight ahead.

“Each model solves complex mathematical equations that describe how the atmosphere changes, incorporating changing pressures and flows, and the weather that will develop as a result. They use data from a wide range of local and global sources including ground- and ocean-based monitoring stations, satellites, aircraft and weather balloons. They run on incredibly powerful supercomputers.”

Alongside the big global data, a local monitoring station plays a critical part in ensuring farmers are given the most accurate picture possible of the weather expected at their place, Dr Poulter says.

“Our two-day model works by producing discrete forecasts for points on a 12km grid covering the entire country. But because New Zealand’s terrain is so varied, within any 12km square there may be mountains, basins, ridges, valleys, lakes or the sea – all of which can have a marked influence on local weather and climate.

“So we need to modify the grid forecast by anchoring it to a nearby point that best represents the climatological conditions on the farmer’s property. That’s where the local monitoring station comes in.”

Subscribers to NIWAFarmMet enter their address as part of the online registration process. The website then automatically selects the NIWA monitoring station closest to their property. The farmer can change that station if they believe a nearby alternative will better represent the climate on their property.

“The selected station provides data on the current weather at that location – effectively giving the models the correct ‘starting point’ for their forecasts,” Dr Poulter explains.

“It also contributes to a process of continual forecast improvement. The models aren’t perfect; they may have biases – perhaps a tendency to under-estimate temperature or windspeed at a particular location. So the data measured locally are used to remove those biases and correct the forecast”.

“The forecast system compares the last six weeks’ of measured data with their own forecasts generated for that period. Any discrepancy is adjusted out before the next forecasts are produced. It’s an ongoing process. The models are continually looking back and comparing, then finessing, their next forecasts.”

The local station also provides data for the range of climate products – including accumulated rainfall, growing degree days, and days with frost – that form part of the NIWAFarmMet package. These products enable farmers to compare current conditions with past seasons and long-term averages.

NIWA continues to look for other ways to improve the accuracy of its forecasts – not only for farmers but other end users with specific forecasting needs.

“Improving resolution – reducing the size of those grid squares – is a key way we can deliver better forecasts,” says Dr Poulter. “The smaller the squares, the more of those terrain-related microclimates we can factor into our base forecasts. It’s a high priority for us. Watch this space.

“We’re also expanding our monitoring network, prioritising those areas where we currently have fewest stations. As our network grows, we’ll be much better placed to provide end users with a data source that truly represents the climate at their location.

“The models will do the rest!”

For more information:
http://farmmet.niwa.co.nz

Fieldays technology

This year’s Fieldays had a particularly appropriate theme for the science sector.

“Managing resources for a competitive advantage” summed up the practical role science has in assisting farmers achieve greater productivity and profit.

The challenge of this theme was embraced by many exhibitors displaying innovative solutions using technological advances to improve efficiency on the farm and embrace a new generation of tech-savvy farmers.

Waikato Milking Systems introduced a product called the Bail Marshal. It had been designed to enable all technology devices on a milking system to work together seamlessly and continually communicate with each other.

Also on show was an electronic drench gun receiving data via wifi to calculate exact dosages needed. Then there was a capsule device placed inside a cow’s rumen which measures temperature and pH levels and can deliver the data to a farmer’s smartphone.

The growth of data-driven technology accessed by farmers from mobile phones was particularly evident at Fieldays, offering a glimpse into the future where everything a farmer needs to know is available on one device at the touch of a button.

Automation – from feeding to milking – was also a prominent theme, driven partly, agricultural leaders say, by New Zealanders’ love of innovation.

It is innovation that has helped the dairy industry reach annual exports of more than $13 billion, helped by a $450 million spend on research and development by the agricultural sector spread across business, government and higher education.

Farming today is about engineering, electronics, information technology and a range of science disciplines. All of which, Fieldays Chief Executive Jon Calder says, has led to the kind of innovation that has made New Zealand an agriculture superpower.

Fieldays Factiods

• 300 + One-to-one demonstrations of NIWAFarmMet given to farmers.
• 450 + Brochures on NIWAFarmMet given away to farmers.
• 25 Frenzied minutes, on day one, when simultaneous visits by Prime Minister John Key and Science and Innovation Minister Steven Joyce made the NIWA stand the centre of action and attention.
• 5000 Packets of NIWA lollies given away to sweet-toothed visitors.
• 1200 NIWA rain ponchos given away to soaked visitors during the first two days.
• 24 Live rural weather presentations delivered by NIWA forecaster Chris Brandolino.
• 40 Throat lozenges consumed by Chris Brandolino to sustain his voice.
• 12 Width, in metres, of screens positioned around the Fieldays complex on which Chris’s Mystery Creek forecasts were broadcast up to six times daily.
• 384 Collective hours spent by NIWA staff on their feet, manning the stand.
• 179 Entries into the draw to win an iPad Air and 12-month subscription to NIWAFarmMet. Congratulations to lucky winner Mr Paul Smith of Hawke's Bay.
• 21 Separate media entities that ran stories on NIWA’s activities at Fieldays.