Satellite imagery in rice crop mapping: Sowing the seeds for stronger food security

At the outer reaches of our atmosphere, roughly 700 kilometres from the Earth’s surface, satellites are capturing images – in this case, of crops and cropland over large areas. The images are beamed back to Earth where scientists analyse them in order to monitor agriculture, identify crops, and estimate the size of farmers’ yields. This technology is crucial in order to monitor large agricultural areas efficiently, and would be difficult or impossible to achieve through traditional ground surveys.

This accurate and rapid monitoring data is crucial to determine the minimum support price (MSP), which gives both farmers and consumers a fair and stable price for their crops in the market. When farmers have higher production, they can choose either to consume the surplus crops, or sell them at the market or to the government, using the already established MSP. This creates a stable environment for farmers and consumers, ultimately strengthening the food security of farming families.

In 2022–2023, Nepal achieved a remarkable 4.33 per cent increase in paddy production, churning out 5.72 million metric tons. The images captured by the satellites provided the crucial baseline information that was required to develop this production data. This technology is known as ‘remote sensing-supported crop area mapping’, and was conducted by the International Centre for Integrated Mountain Development (ICIMOD) in collaboration with Nepal’s Ministry of Agriculture and Livestock Department (MoALD).

The data was used to chart estimates of crop areas; since 2019, such estimates have been enabling the Government of Nepal (GoN) to make informed decisions based on reliable agricultural statistics to bolster the food security status of the country. The GoN also established agricultural information services to facilitate data and knowledge exchange between the relevant departments and the wider user community.

Much of this has to do with a collaborative exercise – between ICIMOD and MoALD – developing an operational service using remote sensing (RS) and machine learning (ML) for estimating crop area and yield. This collaboration culminated in the ministry’s first successful use of Earth observation (EO) data for crop field estimation, which took place during 2022–2023.

This collaborative technological pursuit came about following recent reports detailing the food and nutrition security status in Nepal. One report from the World Food Programme shows that approximately 4.26 million people (14.6 percent of the population) were food insecure in 2023. Similarly, the Global Hunger Index shows that 5.4 percent of the population was malnourished between 2020–2022. One acute instance is the case of Karnali Province in western Nepal, where 24.1 percent of households come in the bracket of ‘moderately and severely food insecure’. This status is exacerbated by climate change and poor harvesting practices.

Importantly, a high hunger line will continue to hinder the population in achieving their full human potential and limit the country in being able to accomplish the SDG goals related to education, health, and gender equality. To address this, the GoN felt the need to build the capacity of the overall food security monitoring system within the government. This includes the development of skilled staff, who would be trained and empowered to generate reliable and timely information, thereby aiding in the planning and policy decision-making processes.

The leadership at MoALD was keen to adopt new data science technologies like RS-based rice crop mapping to generate information more efficiently and to enhance data quality for evidence-based decision-making. Looking ahead, MoALD has expressed its resolve to build the capacity of its staff by utilising these tools and technologies, with potential support from ICIMOD.

Read here the related blog on how ICIMOD and MoALD have been cultivating an effective partnership through state-of-the-art technology.