Effects of Land Development on Peak Runoff Rate and its Prediction for Brigalow Catchments in Central Queensland, Australia

Abstract

The estimation of runoff volume and peak runoff rate has been the focus of significant hydrological research worldwide. The results of these studies, usually in the form of empirical relationships or models, are intrinsically linked to the environment in which the study was conducted. This often limits the applicability and accuracy of the method of runoff estimation at alternative and ungauged locations. Within the brigalow belt of central Queensland, Australia, a scarcity of stream gauging stations to measure runoff volume and peak runoff rate has impeded research on the surface water hydrology of the region. Intermittent failure of these stations and consequently, multiple periods of missing data, have added further complexity and challenge to the understanding of catchment hydrology in the region. Commencing in 1965 and continuing today, the Brigalow Catchment Study in central Queensland has measured both runoff volume and peak runoff rate from three small catchments which initially contained native brigalow scrub. The natural hydrology of the three catchments was characterised during a 17-year calibration period from 1965 to 1981. In 1982, two of the three catchments were cleared, with one developed for cropping and one developed for improved pasture, while the third was retained as an uncleared control catchment. Study of the effect of land development on surface hydrology commenced in 1984. Twenty-one years of record was used to quantify the changes in peak runoff rate associated with land development. Results however, were confounded by missing data. To allow for robust analysis, estimates of missing data were generated via three different methods: (1) multiple variable regression analyses; (2) Soil Conservation Service curve number and graphical peak discharge methodologies; and (3) a simple variable infiltration rate model. The suitability of each technique for the estimation of peak runoff rate was assessed using both graphical and numerical evaluation.