High-resolution (5km) Australian future projections from CABLE

Just wanted to advertise high resolution CABLE future runs that I’ve done for Australia in case these are of any interest to others.

These are 5km spatial resolution runs across Australia covering the period 1960-2099. There are 16 runs in total for RCP4.5 forced with downscaled, bias-corrected met forcings from the Bureau’s hydrological projections project (4 GCMs from CMIP5 x 4 bias-correction methods). I have monthly outputs of the core hydrological variables, sensible heat, GPP, NPP and respiration.

For those interested in CO2 effects, I also have runs with constant CO2 (set to 1960 level).

The slight caveat is that the main purpose of these runs was to compare to AWRA projections from the Bureau’s project so I’ve matched the PFTs to AWRA (meaning only evergreen broadleaf and C3 grass were used) and run with AWRA historical mean LAI rather than observed. But we’ve used a high-res Australian gridinfo otherwise.

These runs used up a fair bit of resources on Gadi so if they are of use to anyone else, that would be fantastic!

Here is a little teaser of the drought changes we are looking at in these runs:


Is this a data collection at NCI @aukkola?

very cool! Are the forcing data on gadi as well?

not yet but I should do that!

The daily data are available under a Bureau project and I have codes to turn them into sub-daily CABLE forcing files using a weather generator. The forcings are too large to store long-term so I’ve had to create them on the fly


Hi @aukkola ,

I am not very familiar with land surface model. But as a coastal ocean modeller, I am interested in the runoff information they provide. I am more familiar with AWRA than CABLE, and used the Australian Water Outlook outputs before. I was wondering if CABLE produces runoff gridded product as outputs (as AWRA does)? and if these outputs were accessible on NCI?


1 Like

Hi Clo,
Yes CABLE produces gridded runoff and I have saved these outputs. However, it doesn’t have a river routing scheme so the runoff is produced at each pixel but not routed along a river network into the ocean. I think (but could be wrong) AWRA also lacks this mechanism so in that regard the models are similar. I am storing the outputs on NCI but have not published them officially. Happy to share them in the meantime though, if you are interested you can email me at a.ukkola@unsw.edu.au

1 Like