Location: Water Management and Systems Research
Project Number: 3012-13660-010-030-S
Project Type: Non-Assistance Cooperative Agreement
Start Date: Aug 1, 2024
End Date: May 31, 2026
Objective:
The Cooperator and ARS will collaborate to develop and test methods for monitoring soil moisture when a Cosmic Ray Neutron Sensor (CRNS) rover is driven on gravel roads. Specifically, the project aims to develop a method for mapping soil moisture patterns over large spatial extents relevant to distributed watershed modeling in semi-arid Colorado.
Approach:
A new soil moisture observation technology uses nearly constant incoming flux of cosmic rays that produce high energy, fast neutrons moderated by interactions with hydrogen in the environment. A cosmic ray neutron sensor (CRNS) detects and counts the number of epithermal neutrons which is inversely proportional to the mass of hydrogen (water) in a support volume (an areal footprint of ~200 m radius ~30 to 60 cm deep). Once calibrated, the CRNS infers the soil moisture. Stationary CRNS’s have been deployed by ARS, and the Cooperator (CSU) owns and operates a mobile CRNS rover to be used in this project. CRNS tubes can be mounted on a vehicle to map soil moisture over large spatial extents. However, the rover must be driven on gravel roads. Accurate estimation of the rangeland or farm field soil moisture requires removing the effects of the road and the ditches. A recent study developed a correction that adjusts the neutron counts to remove the effect of the road in humid areas, but applicability for semiarid rangelands and roads flanked with drainage ditches remains unclear.
Corrections will be tested using the rover on gravel roads and comparing the rover’s estimate of rangeland soil moisture to in-situ measurements collected within the rover’s footprint. The testing will occur primarily at the Central Plains Experimental Range (CPER) in northern Weld County, Colorado. This facility is operated by ARS and spans 6280 ha. Road moisture will be estimated by collecting samples of the road material, and by treating the road moisture as a rover calibration parameter. The preliminary estimate of the rangeland soil moisture will be estimated using the uncorrected estimate from the rover and the average soil moisture from Soil Moisture Active Passive (SMAP) satellite data (9-km grid). The ditch moisture will be characterized by collecting paired in-situ soil moisture measurements in ditches and adjacent lands.
We anticipate performing calibration at three locations. One location will be away from roads near an existing stationary CRN sensor at CPER. The second calibration site will be near a road at CPER. At both CPER calibration sites, 25 measurements will be taken with dielectric sensors at varying distances from the rover within the footprint. The third calibration site will be located near two existing stationary CRN sensors at summit and toe slope positions within Drake Farm, a smaller ARS rangeland study site not impacted by roads.
On 8-10 dates, the rover will be driven on gravel roads that connect CPER and Drake Farm. Three 1-km road segments will be selected that have varying widths and adjacent landscape moisture conditions. Within each segment, ten locations will be identified, and in-situ measurements of soil moisture will be collected laterally out from those road locations to characterize the rangeland moisture. The corrected rover measurement will be compared to the average of the in-situ measurements to assess the rover’s accuracy. The observed errors will also be used to determine whether multiple passes of the rover are needed to reduce the uncertainty. Maps of soil moisture at <1 km resolution will be produced.
