One of the main analytical variable to indicate the evolution and the phases of the composting process is temperature, whose constant monitoring is fundamental for decision making. However, studies usually perform collection of temperature data with a daily frequency due to the operational difficulty in obtaining this information from manually collected samples. Thus, the aim of this study was to determine the ideal frequency of temperature data collection in composting layers. Eight composting layers containing tree prunings + domestic organic residues were installed and four temperature sensors were installed in each layer. The temperature data were collected and recorded from minute to minute by means of a datalogger developed with an Arduino board during 70 days of composting. Thus, the collected temperatures were used as a pilot sample, and therefore the ideal temperature collection rate was estimated for different estimation error limits. No significant difference was found between the different collection times according to the Kruskal-Wallis test at a significance level of 5%. Therefore, the ideal collection frequency can be determined from the error limit of temperature estimation that is acceptable to the researcher.
Keywords: Arduino; Systematic sampling; automation of data collection; composting; non-parametric test.