Data available from the Climate Data Portal

The Climate Data Portal is the access point for all NARCliM data including the 2014 projections.

Users of the Climate Data Portal are able to construct and submit data requests to extract Regional Climate Model (RCM) data for the simulations, locations, time periods and climate variables that are of interest to them. Additional Information is provided for the data sets.

The information provided below is applicable to the 2014 NARCliM data.

Please contact us if you need further advice on how to use the Climate Data Portal or the data available through the portal.

Domain and resolution

All of the data available from the Climate Data Portal is available over two domains and at two resolutions. The NARCliM domain covers southeast Australia at a horizontal resolution of approximately 10 kilometres. The Coordinated Regional Climate Downscaling Experiment (CORDEX) domain covers Australia and surrounds at a resolution of approximately 50 kilometres (See About NARCliM).

Data types

Two types of data are available from the Climate Data Portal:

Site-specific: Data can be extracted for any point location within the selected domain. This data is in comma-separated values (CSV) format. This can be read by standard text editors and spreadsheet software.

Gridded: Data can be extracted for a grid, or spatially continuous polygon, within the selected domain. This data is in geographic information system (GIS)-readable American Standard Code for Information Interchange (ASCII) format and is on a regular latitude-longitude grid.

Simulations

The NARCliM dataset includes a set of simulations in which three configurations of the weather research and forecasting (WRF) regional climate model (RCM) (essentially three different RCMs) are forced with output from four different global climate models (GCMs). Data from these 12 simulations, listed below, are available through the Climate Data Portal.

  • WRF RCM configuration 1 forced with MIROC3.2(medres) GCM
  • WRF RCM configuration 2 forced with MIROC3.2(medres) GCM
  • WRF RCM configuration 3 forced with MIROC3.2(medres) GCM
  • WRF RCM configuration 1 forced with ECHAM5/MPI-OM GCM
  • WRF RCM configuration 2 forced with ECHAM5/MPI-OM GCM
  • WRF RCM configuration 3 forced with ECHAM5/MPI-OM GCM
  • WRF RCM configuration 1 forced with CGCM3.1(T74) GCM
  • WRF RCM configuration 2 forced with CGCM3.1(T74) GCM
  • WRF RCM configuration 3 forced with CGCM3.1(T74) GCM
  • WRF RCM configuration 1 forced with CSIRO-MK3.0 GCM
  • WRF RCM configuration 2 forced with CSIRO-MK3.0 GCM
  • WRF RCM configuration 3 forced with CSIRO-MK3.0 GCM

A second set of NARCliM data simulations include the same three RCMs with a single NCEP/NCAR Reanalysis Project (NNRP) reanalysis, resulting in three simulations:

  • NNRP Reanalysis – R1
  • NNRP Reanalysis – R2
  • NNRP Reanalysis – R3

For additional information, please see the NARCliM Technical Report 4

Time periods

Simulations are available for the following time periods:

  • 1990-2009
  • 2020-2039
  • 2060-2079

The three reanalysis-forced simulations were run for the period 1950-2009.

Temporal resolution

All data are output as time series and are available as daily and monthly values. Monthly values are averages of daily values, with the exception of precipitation, which is an accumulation over the month.

Climate variables

The following surface variables are available from the Climate Data Portal:

  • precipitation accumulation (including bias-corrected data)
  • mean surface air temperature
  • mean surface pressure
  • mean surface specific humidity
  • mean near-surface wind speed
  • mean surface evaporation flux
  • mean soil moisture content
  • mean sea surface temperature
  • mean downward shortwave radiation
  • mean downward longwave radiation
  • mean upward longwave radiation.

Bias-corrected data are corrected using the method described by NARCliM Technical Note 3.

Important additional information

No climate modelling exercise produces error-free data, and the NARCliM modelling is no exception. The following information describes known issues with the NARCliM data. If you have any questions about the NARCliM data, please email: narclim@environment.nsw.gov.au

NARCliM CCCMA3.1 simulations

There is no daily data for any variable for 29 February on any year for the NARCliM CCCMA3.1 simulations. This is because the CCCMA3.1 Global Climate Model that is being downscaled works on a calendar that has no leap years.

Evaporation data

At present, evaporation data is available from the Climate Data Portal. A dataset for evapotranspiration is being generated from other climate variables and will be made available through the data portal in the future.

Mean surface air temperature (tasmean)
The daily tasmean datasets contain spatial discontinuities over the ocean. This is likely due to discontinuities in data from the Global Climate Models that are being downscaled. These discontinuities coincide with discontinuities in the daily hussmean (mean surface specific humidity) datasets. Monthly data and data over the land are not affected.

Precipitation accumulation (pracc_fl)
A known issue with the Regional Climate Model's convection scheme resulted in unrealistically large values for isolated gridpoints for some time steps in the NARCliM simulations. The NARCliM data at approximately 50-kilometre resolution is most impacted. The pracc_fl files contain data in which these unrealistic values of daily precipitation above 1200 millimetres have been filtered out. Over the entire domain, a very small number of data points are affected.

Precipitation accumulation - bias corrected (pracc_bc)
There are some spatial discontinuities in this dataset in some remote inland areas of the Northern Territory, Queensland and South Australia (see Figure 1). This is due to the scarcity of observational climate data in these areas. In these areas, the Australian Water Availability Project (AWAP) gridded dataset used for bias correction relies heavily on interpolation of data between a small number of meteorological observation stations. Therefore, an alternative bias-correction approach that used regional-average data, rather than data for individual grid points, was used in these areas. For further information, please see NARCliM Technical Note 3.

 

Figure 1: Example map of approximately 10 kilometre resolution NARCliM pracc_bc data showing unrealistic spatial discontinuities in Northern Territory, Queensland and South Australia.

Mean surface specific humidity (hussmean)
The daily hussmean datasets contain spatial discontinuities over the ocean (see Figure 2). This is likely due to discontinuities in data from the Global Climate Models that are being downscaled. These discontinuities coincide with discontinuities in the daily tasmean (mean surface air temperature) datasets. Monthly data and data over the land are not affected.

 

Figure 2: Example map of approximately 10 kilometre resolution NARCliM huss data showing an unrealistic spatial discontinuity in the ocean to the east of New South Wales.

Mean sea surface temperature (sstmean)
The daily sstmean datasets for the NARCliM MIROC3.2  simulations contain spatial discontinuities in December 1991. This is due to discontinuities in data from the MIROC3.2 Global Climate Model that is being downscaled. Monthly data are not affected.

Data available from the AdaptNSW website

You can download maps of future changes (2020-2039 and 2060-2079) from the NARCliM projections for temperature, hot days and cold nights, rainfall and Forest Fire Danger Index. These maps are in ASCII file format. They are for the NARCliM domain and are at approximately 10 kilometre resolution. Each download also has a ‘read me’ document.

The data is available for the multi-model mean of the 12-model ensemble for the NARCliM 10 kilometre domain. This, and all other NARCliM data is available from the Climate Data Portal.