How insights from external data yield more efficient crop production

Platforms such as the CGIAR (Consultative Group on International Agricultural Research) and the CIAT (International Center for Tropical Agriculture) are developing models to create strategic foresight scenarios which consider the impact of climate, water resources, and technology on agriculture. The models apply these factors to future scenarios and make detailed predictions on potential outcomes, which farmers can use to make decisions, significantly reducing risk and increasing their chances of successful yield.

The international Center for Tropical Agriculture (CIAT) is making progress in exploring a new approach to agronomy (the science of soil management and crop production) that leverages external data which can help farmers make better decisions and improve their agricultural outcomes.

This approach is called data–driven agronomy. It takes into account field observations, external data from new sources like satellites and weather forecasting, and contextual information to generate tailored recommendations for farmers, guiding them on what to plant, and when and how best to manage their farming practices. CIAT has partnered with a number of farmers to test this strategy, primarily across Latin America.

CIAT in Columbia

Conventionally, crop breeding and agricultural research is conducted in labs and at research stations. CIAT, however, uses data gleaned from farmers in the field to delve deeper into productivity constraints fueled by inefficient management practices, and external factors such as climate and socioeconomic conditions. By uncovering insights from this external data, the CIAT team is able to more easily  identify trends and offer timely, relevant and locally specific recommendations.

CIAT carries out data-driven agronomy at its headquarters in Columbia. In order to do this, the team has iterated on data mining techniques and worked towards building strong relationships with various public and private sector stakeholders in Columbia. Thanks to the country’s commitment to open data policies, the data-driven work benefits from a favorable digital landscape.

The Colombian Government provides open climate data through the National Institute of Hydrology, Meteorology and Environment Studies (IDEAM). Data is also collected from the national crop growers’ associations in the country, which are legal entities that represent local farmers in both the private and public sector and hold historical data about farmers, farming practices, crop planting trends, weather stations and climate data sets.

According to a report published by CIAT, data-driven agronomy is defined as a “set of complementary approaches that enhance traditional agronomy through the use of increased observational information, data mining, and contextualised information.” These approaches allow researchers to analyse different data sets and make tailored recommendations that would help farmers know what to plant, if they should plant, and also when to plant a specific crop.

The CIAT uses the data-driven agronomy approach with crops such as Andean blackberry, sugarcane, limes, plantain, avocado, lulo, beans, mango, maize, rice and banana. A project on the Andean blackberry carried out by the CIAT, for example, demonstrated the benefits of big data methodologies for site-specific agriculture. It saved years of experimental cropping systems research, hefty financial resources to identify ideal management practices for the largely understudied crop.

Applications and results of data-driven farming recommendations

In 2014, during a dry spell in Montería (Córdoba), the CIAT recommended farmers delay planting their rice crop. This advice was followed by 170 farmers. Evaluations estimated that farmers who adhered to the recommendation avoided crop losses valued at over USD 3.6 million and produced a better crop yield in comparison to the neighbouring communities who went ahead with the rice farming as per usual.

A similar initiative by GRET through the Agro-Ecology Learning Alliance in South East Asia (AliSEA) program, along with  CIAT, RT Analytics, and An Giang University-Research Center for Rural Development, involves a program designed to help rice farmers produce better yield through external data analysis.

In Cho Moi, a rice farmer can upload on a mobile application any data relevant to their farming and management practices in a standard template. This data includes farm location, amount of water required/used for production, amount of chemicals used etc. Data inputs from each farm can be entered as a quick response – QR – product code, which can then be also then be presented to consumers at point-of-sale by providing information on the sustainability of the rice product from a specific farm.

“The data which they enter into the app are then used to help them optimize their farm operations, as well as provide buyers with full traceability,” Le Dang Trung, Chief Scientist at RT Analytics, said.

RT Analytics is also incorporating artificial intelligence to the mobile application, in order to provide better support for the farmers. This would involve things like the ability to help farmers identify which disease their crop has been infected with by uploading a picture of the infected crop to the app.

As farmers become more tech savvy and continue uploading a data and images onto the app, like farming schedule, water use, weather observations, pesticide use, crop or yield observations – analysts will be able to combine these data inputs with satellite weather data and soil data in order to generate timely and effective advice. The more data farmers input, and the more data the analysts have to work with, the better recommendations they’ll be able to produce for effective farming practices.