SOIL WATER SENSORS FOR IRRIGATION SCHEDULING: CAN THEY DELIVER A MANAGEMENT ALLOWED DEPLETION?

Abstract

Soil water sensors are widely marketed in the farming sector as aids for irrigation scheduling. Sensors report either volumetric water content (theta(v), m3 m-3) or soil water potential, with theta(v) sensors being more common. To obtain yield and quality goals, irrigations are scheduled to keep soil water in the root zone above a management allowed deficiency (MAD) level, which is specified as a percentage of the plant available water and which may change over the cropping season and with soil type and water quality. For example, representative values for maize grown in a silty clay loam and irrigated with good quality water may be (during pollination): MAD = 40%; field capacity = 0.33 m3 m-3; wilting point = 0.18 m3 m-3; and water content range of the MAD = 0.33 - 0.27 = 0.06 m3 m-3. In a sandy soil, the water content range of the MAD would be much smaller, e.g., 0.03 to 0.04 m3 m-3. An important question then is: are available soil water sensors accurate enough in the field to be reliable for irrigation scheduling using MAD? We evaluated the accuracy of several down-hole access tube type electromagnetic (EM) sensors (and the neutron probe) and several EM sensors that can be buried or inserted into the soil. All sensors required soil-specific calibration, with the possible exception of conventional time domain reflectometry (TDR with waveform reduction). The EM sensors based on capacitance measurements were found to be least accurate, most affected by soil bulk electrical conductivity and temperature, and generally ineffective for irrigation scheduling by MAD. The neutron moisture meter and gravimetric sampling were accurate enough to use with MAD. Some insertion type EM sensors showed promise for MAD-based irrigation scheduling. The capacitance sensors suffer from a fundamental problem in that their EM fields do not uniformly permeate the soil, instead preferentially following paths of greater bulk electrical conductivity. Because of differing arrangements of conductive pathways at each sampling location, the capacitance sensors exhibited a variability that did not reflect the actual field variability in water content.