- Title
- Evaluation of a behavioural approach and a regression approach for the modelling of household-scale outdoor water use
- Creator
- Micevski, T.; Thyer, M.; Coombes, P. J.
- Relation
- 32nd Hydrology and Water Resources Symposium (H2009). H2009: Proceedings of H2009, the 32nd Hydrology and Water Resources Symposium (Newcastle, N.S.W. 30 November - 3 December, 2009) p. 1070-1080
- Relation
- http://www.engineersaustralia.org.au
- Publisher
- Engineers Australia / Causal Productions
- Resource Type
- conference paper
- Date
- 2009
- Description
- Integrated urban-water management (IUWM) aims to reduce demand on traditional water-supply systems by emphasising household and cluster-scale water-management solutions (e.g. rainwater tanks, grey water re-use etc.) Effective and efficient design using the IUWM paradigm relies on greater understanding of household water-use dynamics at the allotment-scale. Outdoor water use is the most spatially and temporally variable of all household water end-uses and generally governs water infrastructure design. Thus, accurately predicting household outdoor water use is key to providing accurate estimates of the effectiveness of IUWM on mains-water savings and reducing peakdemands. This paper evaluates two approaches for modelling daily household outdoor water use. The behavioural model was compared to a linear regression model - which is traditionally used for household water use modelling . The second is the behavioural-modelling approach of Coombes et al. (2000) which models the occurrence and amount of outdoor watering as a probabilistic behavioural response to daily weather conditions, such as rainfall and temperature. Both models were calibrated to monthly outdoor water use data from 130 individual households in the Lower Hunter Valley Region of New South Wales, Australia. Examination of the statistics of this data set illustrated that outdoor water use at the individual household scale exhibits a large variability with CVs in the range 0.5 to 1.5 at the monthly time scale. The performance of both models was evaluated in terms of reproducing the observed monthly statistics. Both models did well at capturing the observed mean, but both significantly underestimated the variability in the observed data. For the behavioural model the likely reason for this is that it was originally developed and calibrated using zonally averaged household data (consisting of approx 13 houses) which has a reduced variability compared to individual household data. Despite the marginal difference in terms of model performance the behavioural model is preferred because it operates at the daily time step and for IUWM modelling outdoor water use estimates are preferred at the daily time step. In contrast the regression model operates at the monthly step. Furthermore, the behavioural approach is conceptually more appealing as it offers considerable scope for further improvement, whereas the regression model has limited scope for improvement. Future research will aim to improve the behavioural approach and capture the variability in outdoor water use at the individual household scale.
- Subject
- integrated urban water management; water supply systems; household water use; Hunter Valley (N.S.W.)
- Identifier
- http://hdl.handle.net/1959.13/919627
- Identifier
- uon:8929
- Identifier
- ISBN:97808258259461
- Language
- eng
- Reviewed
- Hits: 1821
- Visitors: 983
- Downloads: 0
| Thumbnail | File | Description | Size | Format |
|---|