Low-flow hydrological classification of Australia

  • Mackay S
  • Marsh N
  • Sheldon F
 et al. 
  • 15


    Mendeley users who have this article in their library.
  • N/A


    Citations of this article.


This report presents a national classification of low-flow hydrology in Australia. It outlines key low-flow metrics that may be used as suitable calibration metrics for testing hydrological models and provides a basis for extrapolating the results of site- or region-based hydro-ecological studies. A preliminary low-flow classification based on 35 low-flow metrics was calculated for 830 stream gauge records. These 35 metrics are a subset of the 120 flow metrics used by Kennard et al. (2010a) for the classification of Australian flow regimes. A six-group classification was developed and the robustness of the group membership tested. The six-group low-flow classification shows a strong latitudinal gradient relating to climatic patterns and a strong secondary gradient related to position in catchment – which may reflect catchment size and its effect on measures of low flow (i.e. smaller streams cease to flow more often than larger streams). We used a novel technique of cluster stability to assess the ‘validity’ of the low-flow classes. These results can be used by water managers to determine where hydro-ecological studies should be focused, since ‘weak’ flow classes may not yield robust flow-ecology relationships. The low-flow classification process was revised in the light of a series of site-specific hydro-ecological studies that tested how the flow metrics correlated with the biological traits of aquatic taxa (see Marsh et al. 2012). A second and simpler classification using only four low-flow metrics (the flow exceeded 90 per cent of the time, the baseflow index, the number of zero-flow days and the average of the annual minimum flow divided by catchment area) yielded the same general patterns as for the 35 low-flow metrics classification. In acknowledgement of the limitations of using categorical class membership, we also produced a series of maps demonstrating the spatial distribution of five key low-flow metrics across Australia. These metrics were the flow exceeded 90 percent of the time, the baseflow index, the coefficient of variation of the baseflow index, the number of zero-flow days and the coefficient of variation in the number of zero-flow days. In addition to the classification, we compared the point-based class membership for the 830 sites with the long-term Normalised Vegetation Difference Index (NDVI). The NDVI is a satellite-derived vegetation vigour index which has been shown to be a good predictor of water deficit. Using the long-term NDVI, sites could be allocated to the correct low-flow class (as identified by the simplified classification) with an error of approximately 50 per cent of the time (and within the correct or adjacent class greater than 90 per cent of the time). This site allocation is clearly not precise but provides a useful illustration of the spatial distribution of flow classes within a region. We therefore used the national long-term NDVI scores to spatially extrapolate the low-flow classification to generate a map for helping managers transpose the results of site- or region-specific studies and identify areas of potential ecological risk through alteration to low-flow hydrology.

Author-supplied keywords

  • Hydrological classification
  • Hydrology
  • low Flows Series
  • low flow

Get free article suggestions today

Mendeley saves you time finding and organizing research

Sign up here
Already have an account ?Sign in

Find this document


  • S Mackay

  • N Marsh

  • F Sheldon

  • M Kennard

Cite this document

Choose a citation style from the tabs below

Save time finding and organizing research with Mendeley

Sign up for free