AbstractsBiology & Animal Science

Challenges in real-time precision farming: a case study of modelling biomass accumulation

by Mikko Hakojärvi




Institution: University of Helsinki
Department: Department of Agricultural Sciences, Agrotechnology
Year: 2015
Keywords: agroteknologia
Record ID: 1143883
Full text PDF: http://hdl.handle.net/10138/153257


Abstract

Precision farming is a set of advanced technologies and cultivation practices that aim to enhance crop growing conditions in a site-specific manner. Current machinery offers many technical tools for such actions, but information about when and where it is necessary to use the developed machinery effectively is lacking. The use of crop models has been proposed to combine the effects of growing season conditions and field spatial properties. However, in a timely and spatially variable environment, which a field often is, much information about site-specific growing conditions should be available in order for an advanced crop model to reproduce the site-specific growth in a detailed manner. Unfortunately the information from fields has often been very limited, and insufficient for such purposes. Furthermore the set of precision farming tools and the number of growth factors that can be managed is limited. For describing maximal biomass accumulation, a simple crop model was introduced and evaluated in this thesis. The model is mechanistic, and it uses a minimal number of parameters that all are based on physics, chemistry or physiology. The model can be used for calculating the radiation or radiation and water limited biomass accumulation of a C3-crop. A field experiment equipped with continuous measurements was used for model establishment and after model establishment the model was evaluated with a field experiment with various radiation, nitrogen fertilization and precipitation conditions. In both the studies the crop model was found to produce the maximal biomass accumulation when parameter values measured in the experiment were used. The model was applied in a study evaluating the effects of selected site- and depth-specific soil properties on yield variation on three different clay soil fields located in southern Finland. In order to evaluate the effects of selected soil properties under various weather conditions a Monte Carlo method was used with the biomass accumulation model and generated precipitations. The yield variation was evaluated according to temporal mean biomass yield and temporal standard deviation of the biomass yield. For studying the use of the model in a spatial environment, the introduced biomass accumulation model was applied in a simulator built for simulating a fully automated crop farm. In addition, the use of continuous soil moisture measurements for measuring the crop water use and further for biomass accumulation was tested. According to the results, the crop model was capable of simulating the highest biomass accumulation of the crops used in the experiments. This was the case for all radiation-limited simulations and for most of the water-limited simulations. In a few cases the values of the observed soil properties were found to cause too low biomass accumulation in simulations but in such cases the problem was also present in the comparisons of observed soil properties and observed soil water content during the growing season. For future research with the model, the next phase…