Murray–Darling Basin
31.1 Evaporation from off-channel water storages

The volumetric value for the line item for the 2011–12 year was 1,611,332 ML. The line item represents the volume of water that passes into the atmosphere across a water/air interface from off-channel water storages within the Murray–Darling Basin (MDB) region. The following table presents breakdown information for the line item on a surface water resource plan area basis.

(1) Bureau of Meteorology (the Bureau): National Climate Centre daily climate grids (rainfall, temperature and solar radiation) and (2) Geoscience Australia: MDB human-made waterbody feature class and 9 arc-second digital elevation model (DEM).

The potential evaporation estimate produced by the Australian water resources assessment system landscape (AWRA-L) model version 2.0.0 (Van Dijk 2010) was used to calculate evaporation from off-channel water storages. The AWRA-L model uses a modified version of the Penman-Monteith method to produce the potential evaporation. A water balance tool based on a Fortran code was used to determine the amount of water available for evaporation from individual off-channel water storages.

Using daily gridded climate data for the MDB region (including precipitation, temperature and solar radiation data), open water evaporation data on monthly basis were calculated. Daily gridded climate data were available on a 0.05 degree (approximately 5 km) national grid.

The MDB was divided into 105 regions for the purpose of modelling the off-channel water store. The off-channel water store consisted of storages filled primarily by local catchment runoff. These were determined from waterbody mapping conducted by Geoscience Australia as those which:

• are not named storages (assuming that any storage with a name is unlikely to be a off-channel water storage)
• are above 600 m in elevation
• are below 600 m in elevation in areas that receive greater than 400 mm per annum in precipitation and are not within 50 m of a major or perennial stream.

The above rules attempt to divide storages into those that are likely to be filled primarily by local catchment runoff and those which are filled by abstraction from surface water, groundwater or floodplain harvesting. The catchment of each individual storage was determined via analysis of the 9 arc-second DEM.

The potential average evaporation depth across the MDB region was determined as the weighted mean of potential evaporation occurring from the relevant grid points within the region boundary. Points were weighted upon the area they represented within the MDB landscape to remove edge effects (where the area represented is not wholly within the MDB region) and the effect of changing area represented with changing latitude. The average potential evaporation was used as an input into the water balance tool. The tool determines the water stored in each off-channel water storage at each time step and determines the volumetric potential evaporation by multiplying potential evaporation by storage surface area. The tool assumes that actual evaporation will occur at the same rate as potential evaporation unless storage empties, at which time, evaporation will cease.

• AWRA-L potential evaporation estimates are subject to approximations associated with interpolating the observation point input data to a national grid as described in Jones et al. (2007).
• The spatial extent of water bodies subject to the assumptions and methods associated with the data provided by Geoscience Australia.

The uncertainty estimate was not quantified.

A change made to the calculation method resulted in the restatement of the 2010–11 year volume. The method used to quantify the line item was improved and resulted in a material change in volume.

The respective volumes associated with the change are detailed in the following table.

The volume of evaporation estimated for the comparison year for the 2012 Account (1,572,620 ML) is less than the volume reported for the 2011 Account (1,696,154 ML). 

The difference between the previously reported volume and the estimate produced for the comparison year can be attributed to the choice of the AWRA-L v2.0.0 model (instead of the previously used AWRA-L v1.0.0) to provide inputs to the water balance tool. The AWRA-L v2.0.0 model is more reliable than previous models in estimating runoff.

The difference of 123,535 ML represents a change 7% of the previously reported volume.