- Title
- Use of NARCliM rainfall data for simulating streamflow in the Williams River catchment
- Creator
- Lockart, Natalie; Willgoose, Garry; Kuczera, George; Kiem, Anthony; Chowdhury, A. F. M.; Parana Manage, Nadeeka
- Relation
- 36th Hydrology and Water Resources Symposium: The art and science of water (HWRS2015). Proceedings of the 36th Hydrology and Water Resources Symposium (Hobart, Tas. 7-10 December, 2015) p. 617-624
- Publisher
- Engineers Australia
- Resource Type
- conference paper
- Date
- 2015
- Description
- Reanalysis and General Circulation Model (GCM) rainfall projections can be used to predict future streamflow. This streamflow can then be used to assess the reliability of future water supply. The NSW/ACT Regional Climate Modelling (NARCliM) project has dynamically downscaled four GCMs (CCCMA3.1, CSIRO-Mk3.0, ECHAM5 and MIROC3.2) using three Regional Climate Models (RCMs) to produce twelve projections of high resolution spatially distributed climate data for three 20 year epochs (1990-2009, 2020-2039 and 2060-2079), as well as three 60 year reanalysis products for 1950-2009. This NARCliM data is expected to be used for a variety of climate change impact studies. In this study we assess the applicability of both the NARCliM bias-corrected and uncorrected rainfall data for water security assessment, using the Williams River Catchment as a case study. The NARCliM rainfall time series are input into a calibrated SimHyd model to generate streamflow. Compared to historical streamflows, the streamflows produced using all three NARCliM reanalysis have a similar seasonal cycle based on mean monthly streamflow volumes and autocorrelation analysis. The uncorrected streamflow volumes are higher than the historical volumes. The biascorrected streamflow volumes are much closer to the historical but underestimate the historical streamflow volume at the higher elevations. For the GCM/RCM datasets, the uncorrected streamflow volumes tend to overestimate the historical streamflow volumes while the bias-corrected streamflow volumes tend to underestimate the historical streamflow volumes. An autocorrelation analysis revealed that for the CSIRO-Mk3.0, ECHAM5 and MIROC3.2 models the persistence in the streamflows is much stronger than the historical persistence.
- Subject
- rainfall projections; streamflow volumes; catchment average; Chichester Dam
- Identifier
- http://hdl.handle.net/1959.13/1330620
- Identifier
- uon:26433
- Identifier
- ISBN:9781922107497
- Language
- eng
- Reviewed
- Hits: 2605
- Visitors: 2741
- Downloads: 0
Thumbnail | File | Description | Size | Format |
---|