Model of ammonia emission from manure applied on the field

A model is described, which is able to determine the maximum accumulated emission of ammonia from manure applied on the field, and the time course of this ammonia emission. The model is based upon a neural network, which uses the manure-specific parameters: dry matter content, pH, and ammonium concentration, and the external parameters: maximum and minimum air temperature, precipitation, wind velocity, and daily radiation sum, as well as a variable describing vegetation as input. The output comprises two parameters of a hyperbolic function, which describe consistently the time course of ammonia emission. A reliable prediction of the maximum ammonia emission is important in estimating the value of manure as fertiliser. Calculating the amount of nitrogen remaining would prevent over-fertilisation through additional application of mineral fertiliser or manure. The consistent description of the time course of the ammonia emission is necessary for evaluating the effectiveness of the available technical options to reduce ammonia emission. It can be demonstrated that the different reduction techniques affect the emission rate shortly after application, which usually leads to a reduction in final accumulated emission.