Field evaluation of conventional and downhole TDR soil water sensors for irrigation scheduling in a clay loam soil

Authors: Marek, Gary; Evett, Steven - Steve; MAREK, THOMAS - Texas A&M Agrilife; PORTER, DANA - Texas A&M Agrilife; Schwartz, Robert

Submitted to: Applied Engineering in Agriculture
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 8/29/2023
Publication Date: 11/15/2023
Citation: Marek, G.W., Evett, S.R., Marek, T.H., Porter, D.O., Schwartz, R.C. 2023. Field evaluation of conventional and downhole TDR soil water sensors for irrigation scheduling in a clay loam soil. Applied Engineering in Agriculture. 39(5):495-507. https://doi.org/10.13031/aea.15574.
DOI: https://doi.org/10.13031/aea.15574

Interpretive Summary: Irrigation from the Ogallala Aquifer is used to supplement erratic and inadequate precipitation for crop production in the semi-arid U.S. Southern High Plains (SHP). However, decades of pumping with minimal recharge have reduced groundwater levels and well capacities in most regions in the SHP. Effective irrigation scheduling using time domain reflectometry (TDR) soil water sensors may extend these finite groundwater resources by limiting losses to evaporation and percolation. However, accurate measurements are paramount for effective scheduling and field studies are required to determine their usefulness. Researchers from USDA-ARS and Texas A&M AgriLife Research compared soil profile water contents from conventional and downhole TDR sensors with those from a neutron moisture meter in a clay loam field planted to maize near Bushland, TX. Results showed chronic underestimation by the downhole sensor for all irrigation treatments while an array of conventional TDR sensors estimated profile water well. Findings did not support use of the downhole TDR sensor for irrigation scheduling in a clay loam soil.

Technical Abstract: A field study was performed to evaluate the efficacy of two commercially available time domain reflectometry (TDR) soil water sensors for irrigation scheduling in a clay loam soil near Bushland, TX. Campbell Scientific SoilVUE10 and Acclima TDR-315 sensors were installed immediately adjacent to neutron moisture meter (NMM) access tubes in a research field planted to corn (Zea mays L) in 2020 and irrigated by a center pivot sprinkler system. Irrigation treatments included 50 percent, 75 percent, and 100 percent of evapotranspiration (ET) replacement with two access tubes installed in each plot, totaling six sensor evaluation sites. Semiweekly neutron measurements were used to monitor soil water status and schedule irrigation throughout the growing season. Soil profile water content values derived from SoilVUE10 and vertically distributed arrays of Acclima TDR-315 sensors installed at equivalent depths were compared with those from NMM measurements. Average profile values from the TDR-315 sensors trended well with those from the NMM having root mean square difference (RMSD) values of 10.6, 14.3, and 8.0 mm for the 50 percent, 75 percent, and 100 percent treatments, respectively. Corresponding differences in soil profile water on DOY 234 were 15.0, -6.6, and -0.1 mm resulting in overestimation for the 50 percent treatment, slight underestimation for the 75 percent treatment, and accurate estimation for the 100 percent treatment near the season’s end. In contrast, profile values from the SoilVUE10 sensors grossly underestimated those from the NMM for all three irrigation treatments with RMSE values of 94.4, 73.9, and 58.3 mmm for the 50 percent, 75 percent, and 100 percent treatments, respectively. Corresponding differences in soil profile water on DOY 234 were -124.0, -117.2, and -82.7 mm. Comparisons of volumetric water content (VWC) at each of the nine sensors depths revealed that values from the SoilVUE10 sensors were consistently less than TDR-315 values for all irrigation treatments. Underestimation at the near surface (5 and 10 cm depths) was attributed to loss of soil to electrode contact associated with clay shrinkage during periodic drying following irrigation. Although soil to electrode contact can be problematic at greater depths, the explanation for chronic underestimation of VWC was less obvious except to note that underestimation occurred immediately after installation, which indicated poor electrode-soil contact after installation despite use of manufacturer guidelines and tools. Other possible reasons include challenges for accurate TDR waveform analysis and measurement of permittivity for helical electrodes embedded in a plastic sensor body. Results from this study suggest vertically distributed arrays of TDR-315 sensors can provide profile water content values adequate for monitoring soil water status for irrigation scheduling in a clay loam soil. The chronic underestimation observed for the SoilVUE10 sensors does not support their use for irrigation and could lead to over irrigation. Additionally, the relatively short 1 m length is less than the rooting depth of many regional crops and thus not capable of determining percolation below the root zone. More accurate VWC and soil profile values may be achieved following site-specific field calibration.