Comparison of a Low-cost Prototype Optical Sensor with Three Commercial Systems in Predicting Water and Nutrient Contents of Turfgrass

Abstract

ABSTRACT Chemical soil and plant analyses are time-consuming, expensive, and labor-intensive. There are some optical systems used for this purpose; however, they are expensive and require expertise for their operation. The aim of this study was to develop a low-cost prototype optical sensor and compare it with three commercial systems (GreenSeeker NDVI (Normalized Difference Vegetation Index) meter, chromameter, Fourier transform-near infrared reflectance spectroscopy (FT-NIRS)) to determine water and nutrient concentrations including nitrogen (N), phosphorus (P), potassium (K), calcium (Ca), magnesium (Mg), iron (Fe), copper (Cu), manganese (Mn), and zinc (Zn) of turfgrass (Lolium perenne L.). Study was conducted on an experimental field to which four different levels of nitrogen fertilizer were applied. Prediction models were developed using PLSR (Partial Least Square Regression) and their performances were evaluated using the criteria of SEP (Standard Error of Prediction) and R2. With the prototype optical sensor, NDVI gave the best result among 10 different vegetation indices for the prediction of water (SEP = 1.43%) and N (SEP = 0.28%). The best results (lowest SEP) were obtained with the FT-NIRS. However, there are some disadvantages of this system along with the other two instruments (chromameter and NDVI meter) of being expensive and requiring expertise in their operation. Low-cost and easy-to-use prototype optical sensor gave similar results with the NDVI meter and chromameter to predict water and nutrient concentrations except K, Cu, and Zn. An optical sensor similar to the prototype sensor could be developed commercially with low cost and used to estimate the water and nutrient concentration of turfgrass.