Processing of soil temperature data for computational simulations of an agroecosystem

Abstract

Ecosystem services, such as natural pest control, are included in agro-ecosystem management strategies and their use can be optimized based on knowledge from computational modelling of pests, predators, and landscape. In this article we focus on the processing of soil temperature data, necessary for the operation of the predator-pest models Bactrocera oleae (olive fly) and Haplodrassus rufipes (soil spider) in the olive grove. The processing methodology allowed us to (1) complete the gaps in the series of soil temperature data, collected on an hourly basis, (2) convert it into a series with daily periodicity, and (3) create an additional variable, the average daily temperature of the soil during the twilight. This last step prevents the loss of information regarding that specific period of the day, reducing the error associated with the deaquation of the temporal resolution of the average daily climatic variable to express a phenomenon that occurs only during a certain number of hours of the day.