Development of a real-time decision support system for generating and relaying irrigation schedule forecasts for lowland rice

Abstract

The absence of real-time decision support systems (DSSs) that generate, forecast, and relay weather and irrigation schedules to farmers has accelerated unproductive use of water in lowland rice production systems. The objective of this study was to design a real-time mobile phone application decision support system that generates and relays weather and irrigation schedule forecasts to rice farmers and extension workers. In addition, the study also aimed at testing the DSS functionality in a rice production system. The iRice DSS was designed on the Model - View - Controller architectural pattern to render simplified outputs for decision making using Java, HTML and MySQL programming languages. It uses the FAO Penman-Monteith equation, Alternate Wetting and Drying water management practice for lowland rice irrigation, and the water balance equation, to monitor soil moisture depletion at different crop growth stages. iRice DSS was tested on lowland rice crop grown in an irrigation scheme at National Crops Resources Research Institute, Namulonge in Wakiso District in Central Uganda. The accuracy of the DSS weather forecast was tested using the R2 of the historical forecast and observed wind speed, relative humidity, temperature, and precipitation, while the irrigation schedule forecast was tested using the RMSE, model efficiency, Willmott (1981) Agreement Index and R2 of the forecast and observed soil moisture content. The developed iRice DSS consists of iRice Web System as the engine and the iRice mobile phone application as the relay. The DSS monthly average forecast in relation to the observed revealed R2 = 0.972 of Relative Humidity, R2 = 0.652 of Temperature, R2 = 0.614 wind speed, and R2 = 0.784 of precipitation. The iRice prediction of irrigation schedules in relation to the observed was RMSE = 0.055 mm, coefficient of determination (R2) = 0.999. The model efficiency = 0.997 and Willmott (1981) Agreement Index = 0.999. The results imply that iRice DSS has capability to enhance rice yields and improve irrigation water use efficiency through accurate forecasting and relaying of weather and irrigation schedules to rice farmers in Uganda.