Statistical analysis of environmental measurements for design of energy-efficient monitoring systems

Environmental monitoring systems often operate in remote locations and thus must be designed for energy-efficiency and reliability. As the main tasks of such systems are sensing, logging and delivering environmental measurements, the frequency with which are these operations executed significantly affects the overall energy consumption of the monitoring devices. This work presents the results of statistical analysis of environmental measurements (air temperature, air humidity, soil moisture and photosynthetically active radiation), and evaluates how the frequency of their collection affects the accuracy of collected samples. In particular, two independent approaches are discussed. The first approach is based on the concept of stationarity for evaluating time series models, while the second seeks to determine the probability density function through the combination of descriptive statistics with ANOVA parametric analysis. The results of these analyses show that different environmental variables should be sampled with different frequencies. Implementation of this principle will decrease energy requirements of the environmental monitoring devices, and allow their energy-efficient design and long-term uninterrupted operation under demanding field conditions.