Dynamics of soil water content using field monitoring and AI: A case study of a vegetated soil in an urban environment in China

Maintenance of green infrastructure requires persistent monitoring of soil water content, electrical conductivity (EC), temperature and relative humidity (RH). There is need to create relation among these parameters to deduce soil water content as to limit cost of introducing sensors. The present investigation means to measure the progression of these four parameters in a green infrastructure site in an urban landscape and to additionally create correlations among them. An incorporated field testing and statistical modeling approach is embraced to accomplish the objective. Four sites including bare, grassed and treed soil are chosen for examination. Field monitoring was first directed to monitor the above mentioned four parameters for two months. This is trailed by statistical modeling using artificial intelligence (AI) approach (i.e., artificial neutral networks (ANN)). Correlations are created for evaluating soil water content as function of EC, RH and temperature for the four sites. Regardless of the sort of cover, EC is strongly correlated to soil water content, followed by RH and soil temperature. The correlation of EC is strongest in vegetated soil when compared with bare soil. The relationship of soil temperature with water content of soil do not have a decisive pattern. Ensuring drying of soils does not increase temperature fundamentally after precipitation. Uncertainty investigation additionally presumes that water content of soil follows normal distribution function in treed soils, while it follows skewed distribution in the cases of grassed and bare soils. (C) 2019 Published by Elsevier Inc.