Environmental change point analysis and its application to agriculture

Field environmental information such as air temperature, humidity, and solar radiation would be important not only for evaluation of crop growth progress, status and quality or the occurrence of pests and diseases, but also for optimization of agricultural production processes such as plating, irrigation, fertilization, pest control, and harvesting. However, measured data stored in an agricultural information database, would be vast data with the increase in the number of monitoring devices and sensors installed in such devices. Thus, in order to extract feature values from such vast data, it is necessary to develop some numerical technologies and methods. In this study, the change point analysis based on Singular Spectrum Transformation (SST) was applied for extracting the difference of cultivation environmental conditions in each the tomato greenhouse. The change point scores were detected from the change of CO2 concentration data clearly. They agreed well with the farm work history recorded by a corporative farmer.