Frost Prediction in Apple Orchards Based upon Time Series Models

Abstract

This Master Thesis is dealing with frost prediction in apple orchards based upon time series models. It is a part of the DASA (A Data Analysis Platform for a Sustainable Agriculture) research project of the Free University of Bozen·Bolzano, the Südtiroler Beratungsring and the Laimburg Research Center for Agriculture and Forestry. The DASA project has as a goal the development of tools for collection, quality control, monitoring and analysis of the agricultural data. The project commenced in January 2014 and will end in December 2014. The master thesis aims at the creation of frost prediction models, which require interdisciplinary knowledge in advanced statistics, agriculture, meteorology, and in diverse fields of computer science, especially time series data mining. The model should help in the design of an electronic monitoring system that permits intelligent forecasting of frost weather phenomena. Accurate frost forecasting should provide growers in South Tyrol with the opportunity to prepare for frost events in order to avoid frost damage. Based on the analysis of time series data the proposed linear regression and ARIMA models could be compared and evaluated. The best result provided the ARIMA model, achieving in case of forecast for the 95% confidence intervals lower bound the desired value of 1.0 for the recall. This means that all frost cases could be correctly identified. Despite the encouraging results, the rate of the false positives is high, which needs further investigations (e.g., testing VARIMA models, which are a multivariate extension of ARIMA models). The graphical illustration of the 95% confidence intervals lower bound of the ARIMA model forecast should be very helpful in frost prediction and could be integrated in the ”Beratungsring App”.