Conceptual or lumped parameter rainfall-runoff models are more suitabl
e for assessing the effects of climate variables on stream discharge t
han either physically based or empirical models. A lumped parameter ra
infall-runoff model is applied for this purpose to daily precipitation
and discharge data from an upper subcatchment of the Queanbeyan River
in south-eastern Australia, just above the junction with Tinderry Cre
ek. This is a difficult catchment to model and provides a stem test of
the capability of conceptual models, because of the high quick flow p
eaks and small but slow baseflow recession. Models were estimated for
five different calibration periods, each of approximately 2-y duration
. Simulations or verifications were then performed with each of the mo
dels to test their performance in predicting discharge on independent
periods of the climatic record. The results show the importance of cha
racterising antecedent conditions in the model, and that the model is
able to be used for assessment of possible changes in water regime wit
h fluctuations in temperature and precipitation levels in the region.
However, improvements are possible.