Abstract
The ECH2O GS3 soil moisture sensor (GS3, Decagon Devices, Inc.,) is suitable for measuring volumet-ric water content (θ), temperature, and bulk EC for agricul-tural materials. Two types of calibration equations are pro-vided for the GS3 water content measurements: a standard equation (θST) and an agricultural material equation (θAM). In this study, these two equations were evaluated based on the dry weight method θ (θM) for agricultural materials (Rock wool, Vermiculite, Peat moss, and Strawberry soil) and soils (Toyoura sand, Andisol and Kanto loam). The relative error of θST was 10 % – 14 % for agricultural ma-terials, 3 % for the sand, and 14 % for the Andisol and Kanto loam. The θST prediction was reasonably accurate for Toyoura sand, but practically not applicable for agri-cultural materials, Andisol, and Kanto loam. The relative error of θAM was 8 % – 10 %. The accuracy of θAM was slightly higher than for θST regardless of samples. The θST was corrected based on the linear regression line for the θM and θST relationship. The relative error for the cor-rected θST became 1 % – 5 % for all samples. Furthermore, the linear regression line was made according to three sat-urated, intermediate, and dried samples (3-point method). The relative error for the θST corrected with the 3-point method was 2 % – 6 %, which was much smaller than for the original θST. In conclusion, we recommend using the 3-point method to correct θST for the GS2 measurements for soils snf agricultural materials.
