29 Aug Priyantha Jayakody
Priyantha Jayakody, Australia.
Priyantha Jayakody is a hydrologist in the Surface Water Resource Management group at DPI Water (NSW Department of Primary Industries). He received B.S. and M.S. degrees from the University of Peradeniya Sri Lanka, and a Ph.D. in Agricultural Engineering from the Mississippi State University. His doctoral research was focused on water quantity and quality modelling under different climate scenarios using the Soil and Water Assessment Tool (SWAT). Priyantha has been active in the area of water resource management for over 15 years and has been a frequent contributor to the development of new tools and methods required for water resource assessments. He has worked for the International Water Management Institute and the Bureau of Meteorology before joining DPI Water. Priyantha has a sound knowledge of water allocation processes in the regulated rivers of NSW, and is currently evaluating the variability of transmission losses in those systems.
Presentation Title:River transmission loss forecast using multiple variables.
River transmission losses in some New South Wales regulated rivers are highly variable. In large river systems, unaccounted water can accumulate to substantial volumes. In this context, the size of storage reserves required for future water delivery is often difficult to predict. The Lower Namoi regulated river water source is one such system. During the hot and dry summer of 2013/14 in particular, reserves were rapidly diminished due to very high observed transmission losses. The aim of this analysis is to improve seasonal forecasts of river transmission loss using multiple hydrological variables. Variables were selected based on their availability and the strength of their correlation with losses. Unaccounted water – an analogue for transmission losses – was calculated between river gauges, with extreme high flow events removed from the analysis to avoid interference from overbank flows and breakouts. A regression equation was developed to match river transmission losses at a monthly to seasonal scale over the period 2005 to 2016, to match the longer term temporal variability of losses observed in the Namoi river system. The following significant variables were used in the regression equation: rainfall, tributary inflow contributions, soil moisture content and nearby groundwater levels. Three-month forecasts were developed for each variable to enable future loss rates to be predicted. Seasonal forecasts from the Bureau of Meteorology were incorporated wherever possible. The statistical accuracy of the developed regression equation was evaluated using coefficients of determination. Initial results indicate that high, medium and low loss rates can be forecast with greater than 70 per cent accuracy. Results of this analysis can enable water managers to better manage the risk of shortfall due to transmission losses.