Automation of soil moisture monitoring and irrigation prediction for farmlan

Soil water potential sensors were buried at depths of 30, 50, 100, 150 and 200 cm in irrigated farmland, to establish criteria for irrigation scheduling. Fluctuations in soil moisture were found in the 0similar to100 cm soil horizon for wheat. The 30 and 50cm depths showed marked soil water changes while the 150 and 200 cm depths were relatively stable. Based on a large number of regression analyses using farm soil water content data, we concluded that on average, values at 20similar to40 cm and 40similar to60 cm were very similar to weighted means for 0similar to100 cm. Soil moisture for 0similar to100 cm, on farm was defined by soil water potential sensors buried at depths of 30 cm, and 50 cm. From indoor simulated experiments with undisturbed soil, a typical curve for soil water content was drawn and a mathematical model of theta(v) = a(-Phi(m))(b) was defined. A programme was written for irrigation forecasting software, based on observed parameters in irrigated areas. Automation of soil moisture sensing on farm, data collection by computer and irrigation prediction were realized. This system has been used and tested in irrigation areas and has shown a satisfactory precision: relative error was from -6.8% to 4.9%.