Search

AbstractThis paper uses the Budyko and Fu models to estimate mean annual runoff in 0.05° grid cells across southeast Australia, and assesses the estimates against observations in 323 catchments. The results show that the Budyko model can provide reasonable estimates of runoff and is suitable for data limited regions. In regions where there are some streamflow data available, the calibration of a single parameter in the Fu model gives better runoff estimates than the Budyko model and similar to regionalised rainfall-runoff models. In drier regions, the consideration of different vegetation types in the Fu model can improve the modelling results.

AbstractThis study describes a daily rainfall, potential evaporation and streamflow data set compiled for the important water resources region of southeast Australia, and the application of six commonly used lumped conceptual rainfall-runoff models to estimate daily runoff across the region. The daily climate data set and the daily modelled runoff are available from 1895 to 2008 at 0.05° grid resolution across the region. The modelling exercise indicates that the rainfall-runoff models can generally be calibrated to reproduce the daily observed streamflow (for 232 catchments in the high runoff generation areas), and the regionalisation results indicate that the use of optimised parameter values from a gauged catchment nearby can model runoff reasonably well in the ungauged areas. There are differences between the six models, but they are relatively small when used to describe aggregated results across large regions.

Abstract.: A GCM land surface scheme was used, in off-line mode, to simulate the runoff, latent and sensible heat fluxes for two distinct Australian catchments using observed atmospheric forcing. The tropical Jardine River catchment is 2500 km2 and has an annual rainfall of 1700 mm y–1 while the Canning River catchment is 540 km2, has a Mediterranean climate (annual rainfall of 800 mm y–1) and is ephemeral for half the year. It was found that the standard version of a land surface scheme developed for a GCM, and initialised as for incorporation into a GCM, simulated similar latent and sensible heat fluxes compared to a basin-scale hydrological model (MODHYDROLOG) which was calibrated for each catchment. However, the standard version of the land surface scheme grossly overestimated the observed peak runoff in the wet Jardine River catchment at the expense of runoff later in the season. Increasing the soil water storage permitted the land surface scheme to simulate observed runoff quite well, but led to a different simulation of latent and sensible heat compared to MODHYDROLOG. It is concluded that this 2-layer land surface scheme was unable to simulate both catchments realistically. The land surface scheme was then extended to a three-layer model. In terms of runoff, the resulting control simulations with soil depths chosen as for the GCM were better than the best simulations obtained with the two-layer model. The three-layer model simulated similar latent and sensible heat for both catchments compared to MODHYDROLOG. Unfortunately, for the ephemeral Canning River catchment, the land surface scheme was unable to time the observed runoff peak correctly. A tentative conclusion would be that this GCM land surface scheme may be able to simulate the present day state of some larger and wetter catchments but not catchments with peaky hydrographs and zero flows for part of the year. This conclusion requires examination with a range of GCM land surface schemes against a range of catchments.

Many physical‐based models of surface and groundwater hydrology are constructed without the possibility of multiple stable states for the same parameter set. For such a conceptualization, at the cessation of a transient hydrological disturbance of any magnitude the model will return to the same stable state and thus show an infinite resilience. To highlight and falsify this assumption, a numerical distributed ecohydrological model (coupled hillslope Boussinesq‐vertically lumped vadose zone) is presented, in which qualitatively different steady state water table elevations exist for the same parameter set. The multiple steady states are shown to emerge from a positive feedback arising from a reduction in leaf area index (LAI) and thus transpiration, as a saline water table approaches the surface. Limit cycle continuation is also undertaken to quantify the state‐space location of the threshold (repellor) between the steady states (attractors) and quantify the resilience. While the model is biophysically simple, it is sufficiently complex to challenge this potentially significant assumption within water resource planning.

This paper shows the feasibility of using steady state water balances of gauged catchments to calibrate a spatially explicit evaporation model and then applying this to estimate mean annual runoff for 120 gauged catchments in the Murray‐Darling Basin (MDB) of Australia from 2001 to 2005. We used remotely sensed leaf area indices from the Moderate Resolution Imaging Spectrometer (MODIS) mounted on the polar‐orbiting Terra satellite with the Penman‐Monteith equation, gridded meteorology, and a two‐parameter biophysical model for surface conductance (Gs) to estimate 8‐day average evaporation at 1‐km resolution. Parameters for the Gsmodel were optimized using steady state water balance estimates (precipitation minus runoff) in the gauged catchments in three precipitation zones of the MDB, and the calibrated evaporation model was then used to estimate evaporation (ERS) and runoff from gauged and ungauged catchments in the MDB. Mean annual calibrated estimates of ERScompared well with water balance estimates, indicated by a root‐mean‐square error (RMSE) of 78.6 mm/a and the Nash‐Sutcliffe coefficient of efficiency (CE) of 0.68. Reasonable agreement was obtained between the estimated mean annual runoff (RRS) (rainfall minus ERS), and the measured runoff (RMSE = 71.0 mm/a and CE= 0.75). Cross validation showed that estimated ERSand RRSwere almost as good as the calibrated ones. Furthermore, RRShas an accuracy similar to that of a seven‐parameter conceptual rainfall‐runoff model in the gauged catchments. The results show that the evaporation model can be easily applied to estimate steady state evaporation and runoff and that ERScan be used with rainfall‐runoff models to improve accuracy of estimated runoff in ungauged catchments.