No abstract

Impedance probes are popular electromagnetic soil moisture monitoring devices used for a variety of applications but require site‐specific calibrations to provide accurate measurements. Several calibration techniques have been reported in the literature, although laboratory‐based procedures involving wet‐up (via upward or downward infiltration) and dry‐down are commonly performed for permanently installed sensors. Wet‐up calibrations can be completed substantially faster (<1 d) than dry‐down calibrations (1–2 wk), but it is uncertain which technique is preferable to provide the most accurate calibration. The objective of this study was to compare the results obtained from laboratory‐based infiltration wet‐up and dry‐down calibrations of the Stevens Hydra Probe soil moisture sensor. Soil samples for this study were obtained from agricultural sites in Saskatchewan, Canada, at depths of 5, 20, and 50 cm across a variety of textural compositions. Results demonstrate that utilizing either infiltration wet‐up (according to the procedure in this study) or dry‐down procedures provides accuracies of <0.061 m3 m−3 root mean square error (RMSE), which was superior to manufacturer calibration accuracy across all samples. However, superior calibration accuracies (i.e., the lowest RMSE) were achieved using the dry‐down procedure across all soil samples, resulting in a lower RMSE of 0.01 to 0.04 m3 m−3 (at 95% confidence). A significant correlation (r value = 0.61, p < 0.05) exists between the differences in infiltration wet‐up and dry‐down calibration RMSEs and clay content. This suggests that the difference between the two procedures tested in this study is more significant in finer textured soils. The findings of this study indicate that the dry‐down procedure produced the lowest RMSE and is therefore the preferred calibration procedure, particularly for finer textured soils.

Understanding the dynamics of spatial and temporal variability of soil moisture at the regional scale and daily interval, respectively, has important implications for remote sensing calibration and validation missions as well as environmental modelling applications. The spatial and temporal variability of soil moisture was investigated in an agriculturally dominated region using an in‐situ soil moisture network located in central Saskatchewan, Canada. The study site evaluated three depths (5, 20, 50 cm) through 139 days producing a high spatial and temporal resolution data set, which were analysed using statistical and geostatistical means. Processes affecting standard deviation at the 5‐cm depth were different from the 20‐cm and 50‐cm depths. Deeper soil measurements were well correlated through the field season. Further analysis demonstrated that lag time to maximum correlation between soil depths increased through the field season. Temporal autocorrelation was approximately twice as long at depth compared to surface soil moisture as measured by the e‐folding frequency. Spatial correlation was highest under wet conditions caused by uniform rainfall events with low coefficient of variation. Overall soil moisture spatial and temporal variability was explained well by rainfall events and antecedent soil moisture conditions throughout the Kenaston soil moisture network. It is expected that the results of this study will support future remote sensing calibration and validation missions, data assimilation, as well as hydrologic model parameterization for use in agricultural regions. Copyright © 2016 John Wiley & Sons, Ltd.

Passive microwave derived satellite soil moisture data was evaluated over in situ monitoring sites in Canada from two L-Band sensors. Soil moisture data from the Soil Moisture and Ocean Salinity (SMOS) and the Aquarius mission were used, which collect data at different spatial resolutions and using different retrieval models. Both sensors tend to underestimate soil moisture, with the underestimation from SMOS much more pronounced. Correlation coefficients show a reasonably good correspondence with in situ data, and this correlation tends to be better at sites where sub-grid soil moisture variability is represented in the in situ measured data. This highlights the importance of distributed in situ networks.