In this study, soil water balance model was implemented to estimation moisture in two sub-humid areas located in north-western part of Jordan namely Irbid and Ras Muneef. In addition, in situ observations of soil moisture were collected from 16 randomly distributed sampling locations and used for monitoring the spatiotemporal variability of soil moisture in the study area. Sampling was performed during the growing season in the study area with a total of seven sampling occasions from 11 March to 12 April 2017. The results showed that the estimated soil moisture in Ras Muneef was slightly higher than Irbid. This might be referred to variations of, for instance, Ras Muneef receiving higher precipitation and lower temperature values comparing to Irbid. Also, we noticed that the spatiotemporal variability of the observed soil moisture is directly linked with the precipitation in our study area. The coefficients of the spatial and temporal variabilities were in the range 3 to 40% and 8 to 15%, respectively. Due to the high Spearman's rank correlation values which range from 0.42 to 0.73, soil moisture is temporarily stable. Strong and positive relationship was found between the estimated and observed soil moisture with determination coefficients (r2) and root mean square error (RMSE) around 0.7 and 0.41, respectively, whereas negative relationship between evapotranspiration and observed soil moisture was shown with R3and RMSE are 0.34 and 0.24 respectively. This indicates that actual soil moisture can be predicted from soil water balance model.
Keywords: Evapotranspiration; Spatial variability; Temporal stability; Water balance.