Climate Risk Datasets: NARCliM 1.0
Climate risk datasets derived from Global Climate Models (GCMs) typically have a coarse spatial resolution (around 100-200km grid cells), whereas water models require data with much finer spatial resolution. The NSW and Australian Regional Climate Modelling (NARCliM) project addresses this by using dynamical downscaling on a subset of available Coupled Model Intercomparison Project (CMIP) Phase 3 GCMs (Evans et al., 2014). In the first phase of this project (NARCliM 1.0), four CMIP3 GCMs were remodelled onto finer resolution grids using three different regional climate models (RCMs). This process produced a suite of 12 modelled results for three distinct 20-year climate periods: 1990-2009, 2020-2039, and 2060-2079.
NARCliM 1.0 utilized the SRES A2 emissions scenario, which describes a highly heterogeneous world characterized by self-reliant regions with well-preserved local identities and a continuously increasing global population (projected to reach 15 billion by 2100). Economic development in this scenario is regionally focused, and per capita economic growth and technological change are more fragmented and slower compared to other emissions scenarios. Notably, potent greenhouse gases such as hydrofluorocarbons and land-use emissions increase rapidly in the latter half of the century, resulting in the highest overall emissions among the SRES scenarios. This scenario is projected to lead to an approximate 2.2-degree Celsius increase in average global temperature by 2079 (relative to the 1980-1999 baseline).
Average Monthly Scaling Factors from NARCliM 1.0
To inform the Regional Water modelling, future climate projections were combined with 10,000 years of long-term paleo-stochastic climate data. This approach aimed to stress-test our water systems under long-term climate variability and projected changes. For this reason, we selected the Global Climate Model (GCM) and Regional Climate Model (RCM) combination that projected the lowest rainfall compared to the present day.
We calculated monthly rainfall ratios by comparing projected rainfall data (2060-79) to current rainfall data (1990-2009) over 20-year periods at each climate station used in the water models. This calculation utilized bias-corrected NARCliM datasets, and the resulting ratios were then multiplied by the stochastic dataset to factor it. Similarly, potential evapotranspiration results from the same NARCliM scenario were used to factor the stochastically generated potential evapotranspiration results.
The resulting rainfall and evapotranspiration data, representing the dry future climate scenario used in the regional water strategy modelling, was derived from the CSIRO’s Mk 3.0 global climate model, downscaled by three RCMs under NARCliM 1.0. This allowed us to understand how proposed options in the draft strategies would perform under the most severe drought conditions in a future characterized by a slow global response to curbing carbon emissions. It is important to note that this is not the only future climate scenario being considered.
We apply this method of generating stochastic data from historic and paleoclimate datasets, and then factoring it using NARCliM 1.0 outputs, to NSW’s inland regions and north coast. For the south-coast region, our analysis is based on changes to East Coast Low seasonal frequency, informed by NARCliM 1.0. These changes were initially applied using available East Coast Low synoptic data. We then generated stochastic data from this altered dataset (DPIE, 2023).
Note: If you would like to ask a question, make any suggestions, or tell us how you are using this dataset, please visit the NSW Water Hub which has an online forum you can join.
