You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Land use: same as historical CMIP6 forcing (‘LUH2’), which as been extended to cover 850–2015 CE. From the looks of it, it’s LUH2 v2h, downloadable from http://luh.umd.edu/data.shtml
The text was updated successfully, but these errors were encountered:
Section 2.2.1 of the ESM paper states that the capability exists to modify land use through the model run. The details of how this is accomplished however are not specified
We can derive the weights for changing from LUH categories to CABLE categories by matching up the historical datasets, and then doing e.g. a scipy.optimize.minimize to work out the per-land tile weights. This produces a good match when I try for the melbourne grid cell, with the LUH data conservatively regridded to the ESM grid
Land use: same as historical CMIP6 forcing (‘LUH2’), which as been extended to cover 850–2015 CE. From the looks of it, it’s LUH2 v2h, downloadable from http://luh.umd.edu/data.shtml
The text was updated successfully, but these errors were encountered: