An Investigation of Opium Production in Afghanistan Using Remote Sensing Technologies

Keywords: opium production, Afghanistan, remote sensing, artificial intelligence, human security
Abstract Type: Paper Abstract

Abstract

Afghanistan is the global leader in the cultivation and production of opium poppy which is used to produce illicit drugs such as morphine, codeine, heroin, along with synthetic derivatives fentanyl and methadone. Afghanistan alone accounted for more than 80% of the world’s opium supply in 2022 which was estimated at 7,800 tonnes. Opium poppy is a high-value, drought tolerant crop that generated $1.4 billion dollars to farmers in 2022 and is estimated to be 17 times more profitable than wheat. Understanding and monitoring global opium production is therefore of great importance to the international community with respect to economic development, illicit drug trade, human well-being, and political stability.

This research examines opium production through the lens of Human Security and Geospatial Intelligence (HSGI) by applying high-resolution satellite imagery to understand the spatial and temporal extent of opium cultivation in Afghanistan. While the United Nations Office on Drugs and Crime (UNODC) uses a combination of image classification and visual image-interpretation methods to create an annual survey of opium production in Afghanistan, mapping agricultural land accurately and rapidly is challenging, as current approaches use medium-resolution satellite imagery (such as Landsat-8 and Sentinel-2) and require resource-intensive and time-consuming manual image-interpretation. Recent advances in artificial intelligence and high-resolution satellite data have shown promising potential to improve this line of work. Specifically, research efforts aim to develop a deep learning-based image classification method using high-resolution satellite imagery to produce an automated workflow to monitor opium production in Afghanistan with better accuracy and reproducibility.