Evaluating the Uniformity of Sensor-Based Recommendations on Irrigation Timing

Investigators: Himmy Lo, Jacob Rix, Drew Gholson, and Lyle Pringle

Date: 2022

Project Summary

Introduction

The timing of irrigation is important for growing profitable crops while stewarding water resources. When timing is too early, irrigation is applied more frequently, and the soil stays too wet. Such conditions reduce the ability of the ground to soak up rain and the availability of oxygen and nutrients around plant roots. Not only are irrigation water and expenses wasted, but also crop yield is decreased especially if heavy rain occurs shortly after irrigation. On the other hand, when timing is too late, the soil becomes too dry to supply the water needs of the plants. While such delays can conserve irrigation water, drought stress can limit crop yield. Given all these tradeoffs, appropriate irrigation scheduling may be easier with tools that can suggest irrigation timings when the benefits would most exceed the risks.

One such tool is soil moisture sensors. These devices provide convenient, objective, and science-based assessment of water availability in the plant root zone. Typically, irrigation would be recommended whenever the sensor reading reaches a predetermined trigger value. However, sensors of the same make and model rarely report the same reading even when installed at the same depth within the same field. So how different are the sensor readings and their corresponding recommendations on irrigation timing? Although replicate installations of soil moisture sensors may be uncommon on commercial farms, the answer to this question describes the expected uncertainty in irrigation timing recommendations based on only one sensor installation.

Materials and Methods

Two soil moisture sensor models — Irrometer Watermark 200SS and Sentek Drill & Drop — were compared over two soybean growing seasons in a rectangular, precision-leveled field of the Sharkey soil series near Stoneville, MS. At twelve locations across this field, a Watermark installation and a Drill & Drop installation were placed side by side each season. Each of the twelve locations was 360 feet downstream from the polypipe out of a total furrow length of 570 feet. All locations were within 1120 feet of each other and were managed identically. During three drying cycles per season, the date when the sensor reading reached the equivalent of 70 centibars was identified for each installation of each sensor model. The standard deviation in these dates was calculated to summarize the uniformity of sensor-based recommendations on irrigation timing.

Preliminary Findings and Next Steps

The Sentek Drill & Drop was found to be less uniform than the Irrometer Watermark 200SS during four of the six drying cycles studied (Table 1). Nonetheless, the readings from replicate installations were quite variable for both sensor models (Figure 1). The largest differences from the field average were caused by rain occurring between the date when a particular installation reached the equivalent of 70 centibars and the date when the field average reached the equivalent of 70 centibars. Given these observations, NCAAR researchers have been developing and testing strategies to minimize the uncertainty of sensor-based recommendations on irrigation timing. Successful strategies, once proven, will be shared with stakeholders to improve irrigation scheduling across the Lower Mississippi River Basin.

Project Photos

• Topic:
• Irrigation Scheduling
• Irrigation