DETERMINATION OF INFILTRATION CHARACTERISTICS BY VOLUME BALANCE FOR BORDER CHECK IRRIGATION

The infiltration characteristics of a cracking clay soil were predicte d by a regression approach to the volume balance technique utilising a utomatic data gathering techniques. The study was conducted at Deniliq uin, N.S.W. on a clay loam soil during the irrigation of millet and tr iticale and the pre-irrigation for a wheat crop. The analysis techniqu e developed enables the fitting of any form of infiltration function b y least squares regression. The parameters of three common infiltratio n equations could be predicted accurately from measured field data, an d the best fitting equation identified, however, the parameters of the fitted equations were extremely sensitive to errors in measuring fiel d data (inflow and surface storage). Accurately rated hydraulic water control structures are the preferred method of collecting inflow data, although syphons are suitable if entry and exit conditions can be con trolled. In this case, large variations were measured in the rate of d ischarge of syphons operating under identical head conditions. The ave rage depth of water flowing over an irrigation border was related to t he water depth measured at the head of the border, the rate of inflow to the border and the slope of the border. This model of surface stora ge can be improved by the inclusion of further parameters (e.g. biomas s density) not measured in this experiment. Measurement of depth of fl ow at various points within an irrigation border should be made to det ermine with confidence the average depth of flow and surface storage o f irrigation water. The volume balance technique assumes that water in filtrates only into the soil which is covered by the advancing irrigat ion water. This assumption holds for soils in the ''fully irrigated'' state where soil moisture levels are high enough to prevent extensive cracks, it is invalid in dry soils where extensive shrinkage cracks de velop. Under these conditions the form of the infiltration characteris tic is wrongly predicted leading to an overestimate in the amount of i nfiltration. Rates of irrigation advance approached a linear form and the rate of advance was more strongly influenced by inflow rates than by border slope or the stage of crop growth. Further experimental work with accurately measured rates of inflow is needed to define better t he relationships between inflow rate and rate of advance.