Soil moisture dynamics in response to rainfall and irrigation events were examined using data obtained from continuous point measurements carried out under rice and wheat crops for an agricultural plot located within the Ganga Basin, India. Soil moisture data were collected using SM100 and SMEC300 sensors at 18 subplots and at four different depths (0−80 cm) during the period from 5 August 2018 to 31 March 2019. Soil moisture was decomposed into temporal mean and temporal anomalies components, and its variability was characterized considering both absolute soil moisture and temporal anomalies. They exhibited similar patterns at all the depths under rice crop cover. However, it varies with depth under wheat crop cover due to periodic wetting and drying conditions and temporally variable atmospheric demand. Similarly, the spatial variance of absolute soil moisture was decomposed into time-invariant and time-variant components. The results revealed that the time-invariant component contribution was dominant at all the depths (72.49–101.46 %) and the contribution of each component varies with soil wetness and land cover. In addition, temporal stability analysis of soil moisture was carried out. It was observed that the spatial pattern at surface depth cannot be preserved for subsurface depths, and similar subplots were found to be temporally stable at the surface and bottom depths under different crop covers. The results are expected to help improve the understanding of the nature of soil water dynamics in agricultural fields.