Influence of season to season variability in weather on irrigation scheduling of wheat: A simulation study

Summary There is an increasing demand from farmers for irrigation scheduling advice. Where rainfall and evapotranspiration vary little from year to year, advice on a fixed irrigation schedule based on mean climatic data can be given. However where significant year to year variability in weather occurs a more flexible approach using actual weather data to predict the current level of soil water and mean climatic data to forecast the future rate of depletion and hence irrigation date may be needed. A technique for deciding the most appropriate scheduling approach was tested by using a simple model of crop growth combined with a soil water balance model to simulate year to year variability in scheduling advice. This technique was applied to irrigated wheat using a set of climatic data from 1968 to 1978 for Griffith in the Murrumbidgee Irrigation Area of New South Wales, Australia. A typical sowing date in early June was used and simulated irrigations were scheduled at an allowable soil water depletion (ASWD) of 62 mm for maximum yield and 93 mm for 80% of maximum. The analysis predicted that weather variability between years would cause the number of irrigations to vary from 2 to 7 for ASWD=62 mm and 1 to 4 for ASWD=93 mm. The interval between irrigations varied from 12 to 30 days, for ASWD=62 mm and from 16 to 28 days, for ASWD=93 mm. The first irrigation occurred between 76 and 131 days from sowing for ASWD=62 mm and from 100 to 140 days from sowing for ASWD=93 mm. The date of the last irrigation was similarly variable. This high degree of variability in the times and frequency of irrigations indicated that in south-eastern Australia accurate irrigation scheduling advice can only be given by using a flexible model using both actual and mean climatic data. A fixed schedule based on mean climatic data would lead to an inefficient use of water caused by the mistiming of irrigations.