1. INTRODUCTION

Two-thirds of the world’s food needs are produced by smallholder farmers in developing countries. Despite their immense contribution, these farmers are among the world’s poorest people. Unpredictable climatic conditions (e.g., witnessing floods and droughts), land tenure insecurity, pests and diseases, unproductive soils and lack of extension services/information on good farming practices are some of the many challenges they face. Smallholder farmers require relevant information of their farms and the embedding economy to make wise decisions and employ good farm practice. For example, decisions on the crop to cultivate, the timing and amount of fertilizer to apply and whether supplemental irrigation is needed are essential decision making stages in the food production process. Provision of such information through a knowledge-based system can increase agricultural productivity in smallholder systems and improve the livelihoods that depend on this branch of the economy.

Remote Sensing technology has been a useful source of information to improve agricultural management in high-income countries. The collection of vast amounts of data through satellites, aircraft and, recently, unmanned aerial vehicles (UAVs) in these regions permit the delivery of pertinent farm-level information such as crop conditions, soil moisture availability, crop area estimation and yield forecasts to farmers and others interested in such information. There is potential to improve agricultural management and to better yields and sustained and improved food production.

Despite its usefulness in other parts of the world, the potential of RS technology has barely been harnessed in smallholder agricultural systems. This is partly attributable to the typical landscapes in smallholder regions (i.e., small plots interspersed with natural vegetation), the cropping system practiced (e.g., intercropping whereby multiple crops are cultivated on the same land) and high variability in farm practices (e.g., variable planting and harvesting dates, fertilizer application, weeding practices). Additionally, adequate satellite images are often not available for many smallholder farming areas due to persistent cloud cover. The low number of operating satellites (e.g., Landsat) have been a further limitation in harnessing the potential of RS technology in these areas. But recent advances in space technology (e.g., the rise of UAVs), and the launch of a large number of satellites (Sentinels, Landsat-8, RapidEye, SPOT-6/7, WorldView-2/3), present an opportunity to reconsider the role of RS technology in improving agricultural management in smallholder farming systems for increased and sustained food production.

This portal summarizes present knowledge on the potential uses of RS technology in smallholder agricultural systems. This is a wide domain, and the knowledge generated and lessons learnt do not address systematically everything that can be learnt in it. However, the portal specifically aims at:

  • Providing readers with a better understanding of why remote sensing technology has not been used widely in smallholder systems;
  • Sharing remote sensing-based work streams that produce improved information for a number of regional use cases where the demand for such information has been identified for specific stakeholders;
  • Elaborate on recent technological advancements in the field of remote sensing (e.g., the use of high-frequency time series of satellite images and unmanned aerial vehicles) that can improve agricultural management in smallholder systems for increased productivity;
  • Providing insight into the equipment and survey instruments that are required for crop performance monitoring, their usage and limitations;
  • Provide readers with a better understanding of the essential treatments that are required for remote sensing data prior to information retrieval.

We focus here on crops as the primary target of study; we need to point out, though, that generally we could discuss vegetation as the object of study, to include natural vegetation and the related subjects of landscape monitoring, drought risks and desertification. We will not always make this extensive dintinction between the cases in this knowledge portal. Although it is believed that knowledge and experience here described is beneficial to all readers, the portal specifically addresses two types of target audience:

  • Professionals, students and young scientists in research and educational institutions that focus on agriculture and remote sensing domains and are interested in trying out and extending this knowledge and experience. 
  • Operational, donor and policy organizations, representing a continuum of organizations from those that provide only funding to those that define and execute policy. In-between these organizations are government agencies and other organizations with various levels of funding capacity such as international bodies but also donor governments and their aid infrastructure, which may include embassies.

It is assumed that the audience of this portal has basic understanding of principles in remote sensing, although an attempt has been made to explain concepts as much as possible. Where necessary, relevant links to external materials that provide detailed explanations are given.