Publication (Technical report): Characterisation of Flow in Regulated and Unregulated Streams in Eastern Australia

Publication Type:	Technical report / Consultancy
Publication Name:	Characterisation of Flow in Regulated and Unregulated Streams in Eastern Australia

Reference Information

Growns, J., Marsh, M. (2000) Characterisation of Flow in Regulated and Unregulated Streams in Eastern Australia. Technical Report 3/2000.

Optional Information

Attached document(s):

- Flows.pdf

Other information:

PDF report attached below

Characterisation of Flow in Regulated and Unregulated Streams in Eastern Australia

Jane Growns & Nick Marsh

EXECUTIVE SUMMARY

An understanding of the hydrologic characteristics of Australian rivers will help guide the responsible management of water resources, particularly the provision of environmental flows.

This report describes a new method of assessing the hydrology of rivers in a way that is relevant to the ecology of the rivers. It summarises the hydrological characteristics calculated for 107 regulated and unregulated streams and rivers in south-eastern Australia using 20 years of daily discharge data.
The aims of this project were:
1) to develop a comprehensive and ecologically relevant list of statistics that describe a river’s flow regime;
2) to characterise rivers by their flow regimes over a 20-year period; and
3) to answer questions such as

How has water resource development altered the flow regimes of Australian rivers?
Do distinct regions or climates produce characteristic flow regimes?
Do different forms of water resource development (irrigation, hydro-electricity, urban supply) impose characteristic changes on flow regimes?

The project used 333 hydrological variables in seven main categories to characterise the flow
measured at 107 stream gauging locations in south-east Australia. The seven major categories
analysed were:

1. long-term variables, such as mean daily flow, base flow index, maximum and minimum flow;
2. high-flow variables, i.e. number, duration and magnitude of events above a threshold flow;
3. low-flow variables, i.e. number, duration and magnitude of events below a threshold flow;
4. moving-average variables, i.e. 1-, 30- and 90-day moving averages;
5. cessation-of-flow variables, i.e. duration of periods with zero discharge;
6. variables concerned with the rise and fall of the hydrograph, i.e. durations of rising and falling limbs and comparison of differences in consecutive daily flow; and
7. monthly-flow variables, i.e. distribution of flow between months, and annual variability in monthly flow.

The variables for all 107 stream gauges were calculated from records consisting of 20 years of data, from 1/1/1973 to 31/12/1992. Of the 107 data sets, 42 were collected downstream of a regulating structure and 55 were collected from unregulated stations; also, data were simulated for five regulated and five unregulated stations. Stream gauges were classified as regulated if they were located downstream (regardless of distance downstream) of a structure that was likely to alter the stream hydrology. The status of those classified as unregulated was confirmed by checking with the relevant stream gauge management authorities.

The gauging stations were located in three climate types: arid (BS) and two types of warm temperate (Cfa and Cfb). The Cfa stations occurred in non-arid inland areas and coastal areas from Sydney north; Cfb stations occurred in mostly upland areas in south-eastern Victoria and NSW and in northern NSW and southern Queensland.

A cross correlation analysis of output data compared the mean and median values as measures of central tendency. Mean and median values appeared to be highly correlated (r > 0.8) for measures of duration (i.e. the duration of events above a certain threshold) and measures of number (i.e. the number of events above a certain threshold). Mean and median values were not highly correlated (r < 0.8) for measures of flow (i.e. the peak magnitude of events above a certain threshold). Highly correlated variables were removed, reducing the number of hydrological descriptors to 91 prior to multivariate analysis.

Multivariate analyses showed that the main gradient in the entire data set was that of intermittent-flow gauging stations versus permanent-flow gauging stations. Within the permanent group, stations from the BS and Cf climate groupings could be clearly distinguished from each other. Regulated and unregulated stations could not be compared, either within the BS and Cfb climate groupings or among the various types of water resources development, because of the small number of stations in many of the categories. However, within the Cfa group, the irrigation stations were significantly different from the unregulated stations, and many hydrological variables differed significantly between the two groups. The differences included much higher long-term maximum flows, 90th percentile flows and mean daily flows, and longer return intervals for the 2-year flood for regulated stations. Seasonality of flows was also different and several of the flow descriptors were less variable at regulated stations.

If the method developed in this study were applied to more stations within each of the categories, it would be possible to test differences within the other climate groupings and between various types of water resource development, and the variables that best categorise these groupings could be assessed. All of Australia’s river systems should be compared. If greater use were made of modelled data, regulated and unregulated hydrologies could be compared more directly.

It would be extremely useful if better information could be obtained about how and to what extent water is extracted from the rivers. Direct abstraction volumes should also be assessed; rivers where the volumes are found to be large should then be considered ‘regulated’ or at least part of a separate grouping.

The flow statistics derived in this study are only ecologically significant if they are correlated with patterns in stream biota, spatially or temporally. It is important to identify appropriate biological data sets and to test the correlations between flow variables and biota.

This work has important implications for management. As most water managers are aware, it will be a complex task to determine release strategies for ‘environmental flows’ for rehabilitating rivers. To develop a set of flow statistics that can characterise flow regimes, regionally or perhaps nationally, is a high priority. The results presented in this report suggest that each river would need to be considered in the context of (i) the natural flow regimes of nearby rivers, (ii) the climatic zone in which it occurs, and (iii) the main purpose for which it is regulated. However, the method described here needs further development to determine whether fewer flow statistics can achieve acceptable levels of flow characterisation.

Based on the results of this report, it should now be possible to talk with managers and engineers and discuss how water can be released from dams in ways that minimise both the hydrological differences between regulated and comparable unregulated rivers and the effects on consumptive use.

Search classifications

Administration