| dc.description.abstract | Accurate measurements of soil moisture are necessary for predicting weather patterns, mitigating floods and droughts, estimating precipitation and evapotranspiration, and calculating energy fluxes between the biosphere and the atmosphere. However, soil moisture variability is influenced by environmental conditions such as precipitation, soil properties, topography, temperature and vegetation cover. As part of the Hydrometeorology to Operations (H2O) project by the Norwegian Meteorological Institute, this study aims to investigate the temporal and spatial variability of soil moisture at Søråsfeltet in Ås, and to compare the effectiveness of satellite measurements to ground-based sensors.
Both ground-based and remote sensing methods were used to measure soil moisture, including the GroPoint Profile (SMIoT), SoilVUE10, COsmic-ray Soil Moisture Observing System (COSMOS), ThetaProbe ML2 (ADR), in addition to manual samples using the volumetric method, as well as data from the Sentinel-1 satellite. The data was collected from January to December 2022 at three locations in Ås with Søråsjordet as the main focus area.
The results showed significant temporal and vertical spatial variability of soil moisture. While ground-based sensors responded well to precipitation and provided reasonable soil moisture ranges, measurements from the Sentinel-1 satellite did not capture the same variability and its usage is not recommended. The ground truth data lies between the measurements of COSMOS and SoilVUE sensors, suggesting they provide a more accurate representation of surface soil moisture than the SMIoT sensors. However, the SoilVUE sensor experienced a malfunction or data transfer issue, resulting in incorrect data for soil moisture at depths of 10 and 50 cm.
The shallow soil moisture layers of the SMIoT and SoilVUE sensors exhibited more significant fluctuations than the deeper layers, consistent with the faster response of shallow soil layers to meteorological events. The overall trend suggests lower vertical and horizontal spatial variability when the soil is close to or at its saturation point.
The SoilVUE sensor consistently reported lower values than other in-situ sensors, but it showed an overestimation during heavy precipitation. A likely reason is poor contact with the soil due to the hysteresis effect of the soil's expansion/contraction characteristics, which resulted in air gaps after several wetting and drying cycles. This led to preferential flow during precipitation and poor soil contact during dry periods.
This study made several specific contributions to the understanding of soil moisture measurement in Ås: first, it compared various soil moisture sensors; second, it identified malfunctions in the SoilVUE sensor at Søråsfeltet; third, it contributed to the verification process for relocating a SMIoT sensor by discovering a drainage pipe that was affecting measurements in its original location; and fourth, it determined that satellite measurements are not appropriate for this region. | |
