Open access publications in the SLU publication database

Abstract

Because of the interactions between machinery, biological material, weather and management, many problems related to hay-making can not be properly analyzed unless the entire hay production and utilization process is considered. The only way of covering all of this process is to use a calculation model which includes the whole process. It was found that the models available was not readily available for Swedish conditions, but would have to be modified. To some extent their resolutions were also lower than the one aimed at in this project. The hay production and utilization system was modelled as three models being parts of the integrated model of hay production and utilization, a growth model, a harvesting model and a hay-to-milk conversion model. The harvesting model consists of a model of field operations and management, a field drying model and a barn drying model. It is also complemented with models of field losses (shatter losses, respiration losses, leaching losses) and a field loss model. The model of field operations and management was designed as a discrete-event driven simulation program. It was developed using Simula, an object-oriented programming language, and DEMOS, a package for discrete modelling on Simula. The field operations model uses a statistical capacity calculation model rather than a detailed simulation model of the movements of field machinery. The management model concerns itself only with the general behaviour of the manager, leaving open the decision process for the application in question. The growth model used was adopted from Torssell et al (1982). In addition to simulating the dry matter growth, it also calculates the protein and energy content using a "quality" model described in Torssell et al (1983). The field drying model was adopted from Thompson (1981). This is a multi-layer model based primarily on meteorological laws. The barn drying model was developed earlier in this project by Jonasson (1983), further developed by Sundberg (1985). In adapting it to the integrated model, the resolution has been significantly improved. A hay-to-milk conversion model based on Linear Programming has been developed. Due to the fast development on the fodder ration calculation methods, this model is however not currently being used. Most of the models being parts of the integrated model was already validated. The field drying model was however calibrated and validated against Swedish experimental field drying data, resulting in a very good agreement between simulated and experimental values. The integrated model was also validated as a whole, using data from the hay harvesting at a real farm in 1981. The results of these were mixed. The growth model did underestimate the growth rate significantly, probably due to an unusually high nitrogen mineralization rate of the soil. The field operations and management model was instructed to follow the known behaviour of the real farmers as closely as possible, which it did quite well. The field drying model could not be evaluated in this validation due to lack of data. The barn drying model, finally, seemed to predict the drying process in the two barn dryers quite well, but the data available was insufficient to confirm such a conclusion. The integrated model is useful for analyzing a large class of problems requiring that the entire hay production and utilization process is considered. Its complexity however results in the model being useful primarily for research problems, whereas the extension service needs simpler tools.