Soil moisture heterogeneity and sensor deployment in uniformly managed field with unitextural soil

• Published in: Agronomy journal Vol. 114; no. 3; pp. 1800 - 1816
• Main Authors: Irmak, Suat, Kukal, Meetpal S., Sharma, Kiran
• Format: Journal Article
• Language: English
• Published: 01.05.2022
• Subjects: agronomy; confidence interval; rhizosphere; soil water; volumetric water content
• ISSN: 0002-1962; 1435-0645
• DOI: 10.1002/agj2.21064

This research investigated the spatial and temporal distribution of volumetric soil water content (VWC) and total soil water across 64 sampling locations (8 × 8 grid) in a production‐scale field, reported as a unitextural soil unit by SSURGO. Consequently, the required number and placement of soil moisture monitoring locations were calculated to represent soil water status with varying degrees of variability. Each soil layer in the root zone was subject to 5–9% of mean spatial VWC variability and maximum variability as high as 21% during the 2014 and 2015 growing seasons. Total soil water showed spatial behavior, which was as high as 13%, with mean spatial variability of 5%. In 2014, the field minimum and maximum temporal CVs in VWC were 9 and 27%, respectively, which increased to 14 and 33% in 2015, respectively. Representative soil moisture and its variability were accounted for using 1–45 monitoring locations, depending on observed variability, acceptable measurement error, and confidence interval (CI). When the commonly used marginal error of 2% was considered, the numbers of sensor deployment locations were 8 (90% CI) and 11 (95% CI), which reduced to 5 (90% CI) and 4 (95% CI) sensors with 3% marginal error. We demonstrate that although SSURGO data can be valuable to make preliminary assessments, relying on it to spatially characterize soil units and their properties for effective within‐soil unit management, especially for variable rate water and nutrient applications, can create significant challenges. Core Ideas Soil water status was grid‐sampled for 64 grids and mapped across a uniformly managed field. VWC showed 5–9% of seasonal mean spatial variability (as high as 21%). TSW exhibited as high as 13% spatial behavior as well with a mean spatial variability of 5%. Quantitative sensor deployment approach was presented for representing spatial variability. VWC variability was accounted for using 1–45 sensors, depending on spatial variability.