How big data is helping farmers save millions

How big data is helping farmer...
The predictive abilities of big data is being used to improve farming outcomes (Photo: Shutterstock)
The predictive abilities of big data is being used to improve farming outcomes (Photo: Shutterstock)
View 1 Image
The predictive abilities of big data is being used to improve farming outcomes (Photo: Shutterstock)
The predictive abilities of big data is being used to improve farming outcomes (Photo: Shutterstock)

Data scientists studying crop growth and weather patterns in Colombia have advised rice farmers not to plant crops, saving them millions of dollars. The International Center for Tropical Agriculture (CIAT) and the Colombian Rice Growers Federation have developed a computer model that can work out what crops work best under specific weather conditions in certain areas.

Farming has always been as much art as science: knowing what to plant and when is often intuitive for many farmers. However the vagaries of shifting weather patterns and climate change make this much more difficult and a crop destroyed by inclement weather or drought may cost small hold farmers and their families their livelihoods. In Colombia, where rice yields are already dropping and free trade agreements threaten local farmers, improving farming conditions by harnessing the predictive abilities of big data is showing signs of being terrifically useful.

According to CIAT: "Joint efforts on multi-environment trials with detailed physiological evaluation, studies on the adaptation of elite lines, historical data analysis, and crop modeling in Colombia provided important inputs for initiating the development of a system to better manage rice under highly variable weather scenarios."

Considering that there are predictions of massive crop yield fall offs in the coming decades thanks to climate change, such predictive computer modelling is going to become increasingly necessary. CIAT has found that climate accounts for 30 - 40 percent of crop production variability in some parts of Colombia.

Colombia’s rice yields have apparently been hit by climate change problems and harvests of the staple food have declined by a tonne per hectare in the past five years, a steep drop. Maize, potato, cassava and beans are also important crops and the modelling tool will eventually take these in also, under the two year agreement between CIAT and Colombia’s ministry of agriculture.

Last year, researchers from CIAT working with Colombian Rice Growers Federation, advised a group of Colombian farmers against planting crops as they predicted a drought would hit. They were right and saved farmers $3.8 million; those who did not take their advice were not as lucky and lost their crops. Their advice was well grounded: they had developed a complex computer model that had ten years’ worth of data from Colombian farms, from farm management and crop yield to the types of crops and the weather conditions that year. From this inferences could be drawn on what works where, and during what kinds of weather conditions. Whilst traditional farming knowledge is already based around such concepts – otherwise how could farmers grow anything? – this refinement seems to have led to greater accuracy if their success so far is any way to tell.

This project won the UN’s Big Data Climate Challenge in September and is already being looked at in other nations, such as Nigeria. The UN project wishes to, "bring forward data-driven evidence of the economic dimensions of climate change," using big data and analytics. The fields of study are diverse, from transportation and smart cities to agriculture and behavioral science.

(Photo: Shutterstock)

Encouraging to see the success of computer models for regional climate predictions. Of course they depend on data collection of longer time spans but as such data becomes more available, also from use of drones and satellites, so will bigger models.
Does it do any of the stuff that farmers REALLY need? Like keeping exchange rates and interest rates down?