Evaluation of Performance of Organic, Biodynamic, and Precision Agricultural Carbon Uptake Modelling and Yield Prediction

Abstract

Photosynthesis modelling using environmental and crop health data can be key in accurately predicting yields, allowing farmers to find appropriate markets for their produce. This study uses on-site sensors from Arable Mark to collect data on solar radiation, precipitation, temperature, NDVI, and evapotranspiration in order to build localized models for photosynthetic capacity of tomato crops on four distinct New Jersey farms under different cultivation strategies and soil types. Each farm was found to have a peak level of photosynthetic activity that was significantly different from the others. Across all farms, this peak level of photosynthesis was directly corelated with yield later in the season. Multivariate nonlinear regressions were fitted to each farm’s photosynthetic activity over time, in order to produce a predictive model that could estimate peak photosynthesis and thus yield.