A systematic review the of literature on remote sensing tomato crops productivity

The limitations of the sensing capabilities of earth observation sensors have allowed for the advancement of robust high-resolution technologies that are flexible and easy to operate at a low cost, especially in the context of mapping and monitoring the health of crops such as tomatoes in the global south. Although a lot of research efforts have been exerted towards assessing the literature on remote sensing of tomato crops, there are very limited studies that have quantitatively and systematically assessed the findings of those studies to identify the most optimal, sensors, spectral features, modelling algorithms as well as the spatial distribution of those studies. In this regard, this work assessed the progress, opportunities, challenges, and gaps of remote sensing techniques used in characterizing the productivity of tomato crops. Seventy-four articles were retrieved and systematically reviewed from Google scholar, Science Direct, Scopus and Web of Science databases. Results showed that about 44 % of the studies retrieved were conducted in Europe, with the most contributions coming from Italy, while a few studies were from Africa. The contribution of biomass, LAI, chlorophyll, and canopy yield was explored as the most prominent attributes and proxies for estimating the productivity of tomato crops. The most widely used sensors and algorithms which exhibit optimal accuracies in tomato productivity are Hyperspectral sensors (ASD), Unmanned Aerial Vehicles (UAVs), Sentinel 2 Multispectral instruments (MSI), multi-variate techniques, and Machine Learning algorithms. The community of practitioners remains challenged by the high acquisition costs or remotely sensed data and weather constraints due to the restricted spatial properties of sensors in mapping and monitoring crop health to optimise agricultural productivity.