While 2020 will be remembered for the coronavirus pandemic and how it has affected almost every aspect of our lives, arable growers will surely struggle to forget the weather seen this year.
From the deluge most of the country experienced over autumn and winter that dramatically reduced winter cropping, to the yield-reducing drought in April and May and the summer storms that interrupted harvest, it was a season of unwanted extremes.
But imagine if those patterns could be forecast up to two months ahead of time. While those days are still a little way off, progress is being made in improving long-term weather forecasts.
Recent research by a team led by the Met Office suggests that variations in winter weather across a decade are actually predictable based on variations in the North Atlantic atmospheric pressure patterns, or the North Atlantic Oscillation.
This potentially enables advanced warning of whether winters are more likely to be stormy, warm and wet or calm, cold and dry, although the study also found the models needed improving to correctly identify the predictive signal.
Long-term forecasting is also an area the Government-backed data marketplace Agrimetrics and seasonal weather forecasting firm Weather Logistics are investigating, to see whether data can be used to help farmers.
Seasonal weather forecasting is improving all the time, says Dr Chris Nankervis, founder and chief technical officer for Weather Logistics. His firm offers daily weather outlooks to farmers at a near field level resolution.
The improvements have been made through running multiple daily predictions of future weather events from several providers, based on current observations, to provide the probability of certain criteria happening, he explains.
The most accurate variable for long-term forecasting currently is extremes of temperature, for example, the number of days below a certain temperature.
“Rainfall is more difficult. The North Atlantic Oscillation is helpful for predicting winter temperatures, and also temperatures at other times of year, for example in May.
“But for rainfall it is more useful to look at the jet stream, and that fluctuates a lot more.”
Even so, in some circumstances it is possible to start predicting with a bit more certainty that a period will be wetter or drier through repeated daily prediction modelling, he says.
“In the middle of October last year some of the models were quite consistently suggesting November was likely to be wetter than average, and when the predictions range narrows that increases the confidence it is going to be correct.”
The interpretation of seasonal forecasts is all important, he stresses, but potential uses being explored already are helping a salad producer forecast harvest dates based on temperatures from drilling date to match supermarket demand, and with fruit crops, such as blackcurrants, where the number of cold days in winter affects yield in summer.
For the arable farmer, longer term forecasting could help make better drilling date decisions, suggests Alex Borthwick, farm agronomist for R.M. Cottingham, near Market Rasen, on the Lincolnshire Wolds.
“Temperature would be good, looking at the crops we’re growing, like potatoes and sugar beet. Sometimes you get that falsely mild spell at the start of March, but if you knew it was going to come mild and stay mild, you’d be more confident of drilling earlier.
“Of course, rainfall is the one we’d all love to know about; that would definitely help with drilling date. If you know it’s going to start raining and not stop, you’d be inclined to drill earlier, even with black-grass as at least you know the residuals will work better with moisture.
“It would also help with input decisions and help save money.”
Saving time and costs for growers is at the heart of the advanced weather stations offered by companies such as Metos UK, distributor of Pessl Instruments products, and French start-up Sencrop. These concentrate on providing growers with near real-time, interconnected weather data in an easily accessible platform, which can then also be used in conjunction with decision support systems.
Metos provides weather stations aimed at collecting data to help support either irrigation or disease modelling, starting at about £350 for an entry-level LoRain station, rising to more than £1,000 for a more comprehensive solution, while Sencrop offers three types of weather station: Raincrop (£380) – reporting rainfall, temperature, humidity and dew point; Windcrop (£320) which records wind speed, direction and gusts; and Leafcrop, a leaf wetness sensor.
Both companies claim to have several hundreds of stations on farms in the UK, with the data available via dashboards on either desktop or mobile applications.
Licences are included on Metos’ system, although an optional forecasting module is an extra £150/year, while entry level annual subscriptions for Sencrop start at £79/year for two stations, limited additional access to other networked stations, seven day forecasting from Dark Sky and one year’s historical data.
Most customers plump for the £149/year plan with five stations, extra networked stations, five years’ worth of data stored and access to decision support systems, says Harry Atkinson, Sencrop’s UK business development manager.
Having access to almost real-time data is a time saver for growers, he says.
“One of our potato growers uses the stations so they don’t have to drive out to each field to check rain gauges in the morning for two hours. With this they can just look at the app and plan the day accordingly.”
Alerts can also be set up – for example, if a certain amount of rain falls in a period – which again can be used to help decision-making.
But it is linking the weather stations to decision-support tools (see panel), which is going to be the biggest development for the UK in the next 12 months.
Pessl Instruments has been working on this since the early 1990s, says David Whatoff, managing director of Metos UK.
“The vision of Pessl Instruments has been using technology to improve yields and management. In the early 1990s, it developed weather stations monitoring temperature, rainfall, humidity and leaf wetness, which would tell local apple growers when to spray their crops for disease.”
Since then, the company has developed more than 80 disease models in 45 crops, at a cost of £75/year/crop. Unlike Metos (see panel) Sencrop is not developing support tools, but is partnering with others, with two to three likely to be launched in the next 12 months, says Mr Atkinson.
“They are a bit like the Blight-Watch service, but using direct data from the farm to say whether there is blight pressure in that field. I think there’s likely
to be a quick uptake of these services, but getting farmers to trust them will take longer probably. Once they do, it will become part of everyday life.”
Another area being looked at is the use of soil moisture probes, which could help with drilling decisions for crops such as oilseed rape, as well as irrigation.
“These are being developed as job specific sensors, with a much smaller hardware price point of a few hundred pounds,” says Mr Whatoff.
Cheaper data collection from low power wide area network (LPWAN) technology is driving opportunities for arable growers to use sensors and weather stations to actively monitor multiple fields rather than one per farm, he says.
“The future will be an amalgam of open data from a range of products, where you get data from a Metos weather station on one farm, a Davis or Sencrop one on a neighbouring farm, alongside some in-field sensors all linking with decision support tools to help drive change on-farm.”