Preservation of green infrastructure (GI) needs continuous monitoring of soil moisture. Moisture content in soil is generally interpreted on the basis electrical conductivity (EC), soil temperature and relative humidity (RH). However, validity of previous approaches to interpret moisture content in urban landscape was rarely investigated. There is a need to relate the moisture content with other parameters (EC, temperature and RH) to economize the sensor installation. This study aims to quantify the dynamics of the above-mentioned parameters in an urban green space, and to further develop correlations between moisture content and other parameters (EC, temperature and RH). An integrated field monitoring and statistical modelling approach were adopted to achieve the objective. Four distinct sites comprising treed (younger and mature tree), grassed and bare soil were selected for investigation. Field monitoring was conducted for two months to measure four parameters. This was followed by statistical modelling by artificial neural networks (ANN). Correlations were developed for estimating soil moisture as a function of other parameters for the selected sites. Irrespective of the type of site, EC was found to be the most significant parameter affecting soil moisture, followed by RH and soil temperature. This correlation with EC is found to be stronger in vegetated soil as compared to that without vegetation. The correlations of soil temperature with water content do not have a conclusive trend. A considerable increase in temperature was not found due to the subsequent drying of soil after rainfall. A normal distribution function was found from the uncertainty analysis of soil moisture in the case of treed soil, whereas soil moisture was observed to follow a skewed distribution in the bare and grassed soils.