Search

We aim to model the impact of variability in and changes to water availability in the Murray-Darling Basin on flows available to the environment and irrigation, and impact on the value of irrigated agricultural production. Our objective is to understand the opportunities for changed management of the basin, how they are constrained by climate change and other factors, and how they might affect the returns to irrigation and flows for the environment, so that we may provide information to help plan for the future. In this paper we describe the model: in other papers in this conference we describe analyses of water availability and use in the basin based on this model. The hydrology component of the model is based on a simple, monthly water balance stocks and flows model of the basin, subdivided into 58 catchments. In each catchment, the rainfall and potential evapotranspiration are used to partition the rain between actual evapotranspiration and runoff. Runoff accumulates in the rivers, and flows downstream; it is stored in dams, fills lakes and wetlands from which it evaporates, spills onto and is partly consumed on the floodplains, is diverted for irrigation, eventually (if enough water remains) flowing out of the mouth. This hydrology part of the model is calibrated against observations of flow at the downstream flow gauge of the 58 catchments (the records of which vary from a few years to the full 114 years of our typical simulation period from 1895-2009). It simulates reasonably well the full range of flows, and the development of dams and irrigation diversions. The economics part of the model is based on regressions with dependent variables: the observed areas, production, water use and gross value of production of irrigated agriculture. Each dependent variable is estimated as functions of water available, evaporation and rainfall, and crop prices, for ten major commodity groups. The regressions are based on data for 17 regions and four recent years during the drought: they cover a wide range of water uses, water availability, rainfall, evaporation and commodity price circumstances observed during the drought. We report separately in this conference on this statistical analysis (Connor et al, 2012a. In the integrated model, the hydrology model first determines the availability of water for irrigation in the 58 catchments and also calculates the flows, on a monthly cycle. Once per year, the water availability values are aggregated to the 17 economic regions, and the economic model determines the irrigation outcome in terms of areas under each commodity group in each region and the gross value of production. The integrated model has some unique features in comparison to existing MDB economics models: the coupling of economics with detailed hydrology; the ability to simulate active management of environmental flows and the resulting consumptive water use economic impacts; and, the ability to simulate the dynamics of the water balance and economic impact over 114 year historical and simulated future climate sequences.

The recent drought saw the lowest inflows on record in the Murray-Darling Basin in 2006. Water use by irrigation in 2007-8 and 2008-9 was about one third that of pre-drought levels. Understanding how irrigation adapted to less water is key both to provide information to help plan for the future and data for model calibration. Complicating this objective was the concurrent international food price crisis of 2008 and an increase in water trading in the basin. This presents a challenge in unpacking the effects of commodity price changes, input substitution, productivity gains and water trading from the effect of reduced water availability. In this paper we seek Our objective is to unpack the various effects. We used publicly available price and production data to calculate a price index and adjust gross value data in actual dollars using this index; some price and volume data are not available, so the adjustment is approximate for total irrigation. We calculated productivity both in terms of output volumes per unit of water and in terms of adjusted gross value per unit of water. In 2007-8 and 2008-9, overall water use in irrigation was down to about 31 % and 33 % respectively of the 2000-01 value. In 2008-9, the gross value of irrigated agricultural production was down to 86 % in unadjusted terms, but 80 % in commodity price adjusted terms. The gross value (adjusted) per unit of water more than doubled (at 241 %). The aggregated data obscure the variation of water use in different industries. Water use by cereals, grapes and fruit and nuts changed little. Rice, cotton, meat (pasture) and dairy all reduced their water use substantially (to as little as 1 % of the 2000-1 value in the case of rice in 2007-8). Dairy, cereals, rice and meat all experienced significant price rises in 2007-8, whereas prices of cotton and grapes fell. In cereals, the higher prices and the modest change to water use resulted in a large increase (to 187 %) in gross value; some of this was due to higher prices, the remainder to productivity gains. Cereal production increased in the north of the MDB and fell in the south; we speculate that cotton growers swapped to winter cereals, whereas rice growers did not. For rice, which is a high water use crop, the price rise was insufficient to outweigh the decline in water availability, and production fell. All industries show increases in productivity by value or by volume per unit of water. The greatest increase was in dairy: dairy farmers adapted to water scarcity and high water prices by substituting bought-in feed for irrigated pasture. Water trading increased after 2006; importantly, inter-regional trade increased greatly, from a few tens of gigalitres (GL) per year before 2006 to more than 500 GL/yr in 2008-9. The source of traded water was from the Murrumbidgee and to a lesser extent Murray NSW; buying regions were South Australia and to a lesser extent Murray Victoria. Much of this trade was away from rice and dairy and towards horticulture. Trade thus offset some of the impact of reductions in water availability, buffering South Australia and the Victorian Lower Murray from very low allocations.