Key points are not available for this paper at this time.
Physiological principles-based crop modelling and in situ sensor technology provide opportunities for smart nitrogen (N) management for sustainable agricultural production. We propose two N-management optimization methods, in which the mathematical 'bisection algorithm' is combined either with the crop modelling (CM method) or with an integrated remote sensing-crop modelling by data assimilation (RSCM method). Data collected from a field experiment of rice with six N treatments (each with four times of topdressing) were used to illustrate the methods, where the first two N topdressings (Ntop) were applied as in the experiment while the last two Ntop were optimized. The two methods were compared with three reference methods: farmer practice optimized by the yield response curve (FPopt), and the Sufficiency Index- or Response Index-based remote sensing (RS) methods. Crop growth and yields using N applications from these reference methods were also simulated by the same crop model. Compared with FPopt, the sum of the optimized last two Ntop of the CM method on average decreased by 37.9%, while that of the RSCM method decreased by 61.2%. The methods of CM, RSCM and RS decreased the simulated yield by 0.9%, 1.2%, and 4.4%, respectively, while they increased the profit by 2.8%, 4.4%, and −0.4%, respectively, compared with FPopt. The CM method relying on crop physiological principles tended to perform better than the methods of FPopt and RS in optimizing in-season N application, while the RSCM method further benefited from assimilating data from in situ remote sensing information into the CM framework, thereby potentially best suiting to guide smart fertilizer management.
Wang et al. (Thu,) studied this question.