Use of remote sensing to predict the optimal harvest date of corn

Accurate prediction of optimal harvest date (OHD) is important to maximize the yield of corn as it approaches maturity. Yield loss occurs if harvest occurs in advance of, or is delayed past, OHD; both scenarios are undesirable. If corn is harvested prematurely, high residual moisture makes it prone to mildew during storage, which produces aflatoxin. The grain quality then deteriorates, leading to a serious decrease in yield. If corn is harvested too late, the crop quality declines severely. Previous studies have demonstrated the utility of three corn kernel indicators (kernel moisture, milk line, black layer) in predicting corn harvest date; however, these indicators do not satisfy the demand of modern agriculture. Therefore, in this study, remote sensing techniques, which have proven useful in the application of Precision Agriculture, were used for timely estimation of OHD for large-area crops. We proposed a new method for predicting corn OHD in fields using remote sensing by assessing corn kernel moisture (CKM) and identifying an effective biochemical indicator to relate OHD and remote sensing data. We estimated CKM using canopy chlorophyll content (CCC), a significant biochemical parameter that can be estimated from remote sensing images using PROSAIL, a popular used radiative transfer model. Since CKM decreases at a regular rate after entering the maturity period, OHD can be predicted based on the estimated CKM using a baseline assumption that OHD occurs when CKM drops to 30%. The prediction method was evaluated by analyzing measured CKM, and the temporal variation in one hundred-grain weight and yield. The results of this study enabled us to successfully and accurately estimate CKM using a non-linear model (R2 = 0.92) and predict CKM (R2 = 0.64) and corn OHD. This study provides new avenues for predicting crop OHD using remote sensing multispectral imagery and suggests that remote sensing techniques are effective for accurately predicting corn OHD across large areas.