Epsilon Archive for Student Projects

Abstract

Different means of yield forecasting have been investigated in many publications, although very few with a focus on pre-season forecasting. Climate change and the increasing risk of extreme weather events is a threat to food production, and early yield predictions could therefore be beneficial for farmer’s planning and management as a way towards more resilient food production. This thesis aims to explore if yield forecasting of malting barley in Southern Sweden can be done using climate markers of weather occurring during the 60 days before the beginning of the cropping season. Data from field trials, conducted between 1999 and 2018 and acquired from NTFS and SLU Fältforsk, were used as experimental data to calibrate and evaluate the CERES-Barley crop model. Three cultivars were used: Astoria, Irina and Propino. The model input additionally consisted of weather data from AgERA5 and soil profiles from ISRIC that were connected to the field trial sites. The climate markers and the simulated mean yields were tested with a linear mixed model analysis and a Pearson chi-square test in RStudio. The crop model was successfully calibrated, with Astoria being the best cultivar to mirror observed yields from different years and locations in Sweden. The results indicated, however, that none of the analysed climate markers during the 60 days prior to the beginning of the cropping season was able to explain malting barley yields, highlighting the difficulty in making early yield forecasts. One hypothesis for this lack of explanatory power from the climate markers on the yield of spring malting barley is that the effect of precipitation in the 60 days before the beginning of the cropping season is buffered by winter and early spring, when soils are usually at or close to field capacity.