Monitoring Sustainability Practices in Dry-Land Crops with Farm Digital Twins

Abstract

Conservation agriculture together with organic practices could help reduce greenhouse-gas (GHG) emissions and global warming potential (GWP) to counter the adverse effects of climate change and sequester carbon in the soil improving its fertility. In light of this, there is a growing need to model and monitor farming environments to precisely address local and global concerns. We propose in this paper a farm digital twin framework that was used to monitor dry-land crops: maize, soybean, groundnut and rajma (kidney beans) to forecast and validate yields using a process-model approach while studying the overall impact on various other fronts. Deep learning models were used for early diagnosis of disease and pest conditions with imaging that drove timely operations to minimise impact on crop growth. Results indicated that application of organic manure was a key driving factor in high sequestration of carbon while no-tillage helped reduce the GWP of the farms.