Skip to content

You are here: Home / What we do / Digital development / Data science and modelling

Farmers need access to high-quality evidence to support their decision making, particularly in the face of climate change and variability.

The opportunity

Data science and modelling are key tools for generating this evidence – identifying trends and patterns, filling information gaps, predicting likely future scenarios, and offering insights into the best course of action.

Our core strengths in this area help farmers tackle complex problems related to pests and diseases through the application of data-driven approaches and development of bespoke tools. This work can help farmers and policy makers build stronger agricultural systems and respond to a changing environment.

Providing solutions

CABI uses a wide range of data sources to build and validate models that offer insights for decision making. For example, information on the likelihood of the establishment, occurrence and potential damage caused by pests and diseases can help inform actions at both a farm and policy level.

At the farm level, models can be built to optimize the timing of interventions against pests and diseases, contributing to benefits such as rationalized use of pesticides or improving confidence in the effectiveness of slower acting products such as fungal biopesticides. At the policy level, these models can contribute to pest risk analyses, designed to prevent the entry and establishment of emerging or regulated pests.

Data-driven approaches to support pest and disease management are underpinned by access to high-quality datasets. The resolution and availability of datasets is improving continually and they provide the ‘building blocks’ necessary for accurate modelling. Environmental data can be collected remotely through earth observation sources and provide information on conditions such as temperature and humidity.

These datasets cover large areas – including inaccessible places – and are easily updated without multiple visits to the field. As well as environmental data, information relating to the biology and distribution of pests is key. These data can be obtained through globally accessible data sources, including CABI data resources, or primary data generated through research partnerships between CABI and local partners.

CABI is uniquely placed to collaborate with networks of partners, using data science and modelling to strengthen food systems and protect the environment. Through projects that combine data from many sources, we work with partners to model both challenges and solutions.

Our data science and modelling expertise

We develop research-led solutions to support sustainable development and the strengthening of food systems. Our strengths in data science and modelling include:

Alert

Alerting farmers about impending pest risks

We combine the best sources of weather and land-use data with knowledge about an organism’s biology and locally collected field data to create pest models. Advisory messages and recommended actions to take are shared with users.

Mapping

Mapping and monitoring the spread of invasive weeds

Our expertise helps us understand and prevent threats from invasive weeds. Parthenium hysterophorus, for example, is widely considered one of the most damaging weeds in the world. With partners, we were able to apply novel processing and analysis techniques to earth observation data to provide detailed mapping of this weed.

Predicting insects

Predicting the potential establishment of insects

Through our species distribution and environmental niche modelling work, we are able to better understand the potential distribution of invasive pest species and associated biological control agents. We use model outputs to advise on suitable survey areas for classical biological control agents.

Supporting uptake

Supporting the uptake of biological control options

By integrating advanced earth observation and environmental data, we can improve confidence in the use of biological control interventions.

Contact us

Using our expertise in digital development, we turn data and science-based knowledge into actionable, practical information that addresses real needs such as helping to prevent and tackle invasive crop pests. This helps to transform smallholder farmers’ livelihoods and helps agricultural and environmental professionals be more effective in their work.

To find out more about working with us on data science and modelling, contact us.

CABI In United Kingdom

Bryony Taylor

Head of Data Sciences and Modelling, Digital Development

T: +44 (0)1491 829014 E: b.taylor@cabi.org

Examples of our work

Maize farmers in Ghana tackling fall armyworm

PRISE

Pest outbreaks can be devastating for food security in Africa. Between 2017-2022, PRISE delivered pest alerts to farmers in Kenya, Ghana, Zambia and Malawi. Alerts identified the ‘time to action window’ which enabled farmers to take timely preventative action against key pests on maize, beans and tomato.

Parthenium Eradication.

Earth observation data

In Pakistan, CABI helped national decision makers use earth observation data to assess the extent of the parthenium weed problem in different parts of the country and create regional action plans. CABI coordinated field data collection campaigns to validate analyses produced using satellite data, built capacity and links with researchers in-country, and ran workshops to get results used in research, extension and policy-making circles.

 

Related projects

Satellite

Earth observation to improve critical datasets for pest risk modelling

Rising temperatures have led to pests, diseases and weeds establishing in areas of the world that were previously uninhabitable. Furthermore, growth in global trade and new trade pathways increase the risk of accidental movement of pests. Earth Observation (EO) and climatic data can help by improving predictions about where potential agricultural pests and diseases may be a threat. Information produced by models can help decision makers understand and prepare for future risks. Working with a consortium of researchers, this project will use EO data to improve the data layers used in models that predict where pests can establish, including irrigation, areas under protected agriculture and climatic canopy conditions, demonstrating the improvements made to species distribution estimations for key pests and biological control agents.

Establishing a digital plant health service in Malawi

Pests and diseases contribute to 40% of food loss leading to food insecurity. Synthetic pesticides are the predominant control method but these are associated with negative environmental and health concerns. The extensive use of chemicals has sparked a renewed interest in Integrated Pest Management (IPM) – an effective combination of control methods and the need for new innovative ways to manage pest and disease outbreaks. There are many digital systems that have been developed to identify, monitor, manage, control and predict outbreaks of a large number of pest and disease species. These systems provide useful information to aid decision-making and timing of integrated pest management strategies. By building on the successes of existing systems and data assets, this project aims to establish a digital agricultural plant health service for efficient pest and disease management in Malawi that will benefit over 100,000 farmers.

Crop field

Global Burden of Crop Loss

Efforts to reach Sustainable Development Goals in food security, nutrition and livelihoods are being hindered by crop loss. Up 40% of crop yields are lost to pests and disease but the data available to prove and show trends is limited. The Global Burden of Crop Loss project will collect, validate, analyse and disseminate data on the extent and causes of crop loss with the aim of gathering sufficient and reliable data that can act as evidence to enable prioritisation of research and policy in plant health, to improve our ability to predict the impact of emerging diseases.