Climate risk data and water modelling technical information and resource hub

Quality assurance and peer review

The resulting paleo-stochastic baseline has undergone rigorous scientific review to ensure its accuracy and usefulness. The NSW Office of the Chief Scientist and Engineer have assessed the methods, data sources and modelling approaches used to create it, and has formally endorsed it as being consistent with best-practice approaches to climate risk management.

Diagnostic checks and validation are performed for:

• daily rainfall distributions
• wet-day frequency
• spell lengths
• seasonal cycles
• spatial relationships between locations
• the joint behaviour of rainfall and evapotranspiration.

It also incorporates low-frequency variability linked to large-scale climate drivers such as the IPO, so that extended droughts and wet periods occur with realistic timing, duration and intensity.

Independent peer review findings and responses include:
 

Recent climate patterns are captured in instrumental observations from daily station and gridded datasets across NSW. These data are quality-checked, gaps are filled where needed and homogenised before being used to calibrate the model. They provide key information on seasonality, rainfall intensity and spatial patterns.

Climate variability beyond the instrumental record is captured in paleoclimate proxy information. The baseline uses reconstructed data from 5 tree rings, one continuous-growing coral and one historical archive. Tree ring and coral data are precisely dated to annual or seasonal scales. Details are outlined in Climate-informed stochastic hydrological modeling: Incorporating decadal-scale variability using paleo data.

These records help estimate how often and how long extended droughts and wet periods have occurred in the past. They also inform how these events relate to large-scale climate drivers that influence rainfall patterns over decades, such as the IPO.

Methods and evaluation

Multisite stochastic data generation

A multisite framework produces daily rainfall and potential evapotranspiration data across NSW river valleys. A stochastic generator is the set of statistical models used to create synthetic climate data. The generator is calibrated for wet day frequency, daily intensity distributions, dry  and wet spell statistics, inter day autocorrelation, cross correlation between sites and seasonal cycles. It also maintains the strong relationship between rainfall and evapotranspiration so that hydrological water balance behaviour is credible when used in models.

Embedding low-frequency variability

The baseline captures low-frequency (decade-scale) variability by using paleoclimate proxy data to model the longer-term wet and dry periods influenced by the IPO. Instrumental records were used to model the finer characteristics of those wet and dry periods.
Because the fine scale properties come from instrumental data, each catchment reflects its own observed sensitivity to IPO phases. In some valleys, the difference between IPO wet and IPO dry conditions is substantial (such as in the Macquarie), while in others the contrast is minimal (such as in the Murrumbidgee).

Temporal and spatial characteristics

The baseline dataset is provided at a daily time step over ~10,000 years. Spatial coherence is maintained within each valley and across the Murray–Darling Basin, reflecting the shared inland climate drivers. Spatial coherence is not preserved between the coastal valleys and the basin or between coastal regions due to differing climate drivers.

Correlation structures are aligned with observed gradients that reflect topography and coastal influences to the extent supported by the instrumental data.

More information on technical methods and evaluation