Next Generation Technology for Monitoring Edge-of-Field (EoF) Water Quality in Agriculture

Emmanuel Deleon, Erik Wardle, Jay Ham, Dylan Casey, and Christina Welch


Irrigated agriculture has the potential to be a contributor to non-point source water quality issues. One of the most effective methods to quantify sediment and nutrient losses from fields is edge of field water quality monitoring. Industry standard sampling equipment can be cost prohibitive, thus limiting its use and reducing the availability of reliable data. The goal is to develop a low-cost water quality sampler with similar function and reliability at a fraction of the cost. The sampler is being tested at IIC using a side-by-side comparison with industry standard equipment. The low-cost sampler is built with a cellular microprocessors, peristaltic pumps, and water level sensors. It will be comparable to the industry standard in measuring flow weighted sampling, flow detection, logging and remote connectivity. For the industry standard, we used a Teledyne ISCO 6712 with a bubbler depth sensor, and a 1.5 H flume to measure flow at the edge of the field. Both samplers were automatically activated when an H Flume water height of 0.5 inches was reached. Samples (200 mL) were taken every hour during the irrigation event. The preliminary data from the water level sensor compared to the ISCO bubbler on two irrigation events show comparable results with an R2 = 0.99 and R2= 0.99 for irrigation 1 and 2, respectively.  More extensive field trails will be conducted next season to validate the low-cost sampler accuracy. This low-cost technology has the potential to provide similar results and reduce costs of monitoring equipment up to one tenth of the price of standard equipment. A reduced cost barrier makes this equipment widely available to farmers and other interested stakeholders, which results in more informed decision making.  Understanding that every field cannot be monitored, this technology is intended to provide additional data to enhance water quality model results.