make a FHIST simulation using modified flanduse_timeseries

[xiaoxiaokuishu]

Hi, all,
I plan to make a FHIST simulation using modified timeseries land cover data (time period 2000-2010) ,it means land cover in China needed to be changed annually, outside of china the land cover keep consistency as in flanduse_timeseries. Firstly, i change the value of PCT_CROP, PCT_NATVEG, PCT_LAKE, PCT_WETLAND, PCT_GLACIER, PCT_URBAN within china based on the surfdata_0.9x1.25_hist_16pfts_Irrig_CMIP6_simyr2000_c190214.nc (already make sure the sum of revised PCT_xx equals to 1) and successfuly the case. But the surfdata_0.9x1.25_hist_16pfts_Irrig_CMIP6_simyr2000_c190214.nc is only fixed for one year, my research is to use the time series land cover data in China. So I have to revise the "landuse.timeseries_0.9x1.25_hist_16pfts_Irrig_CMIP6_simyr1850-2015_c170824.nc"? I notice there is only PCT_CROP land cover in "landuse.timeseries_0.9x1.25_hist_16pfts_Irrig_CMIP6_simyr1850-2015_c170824.nc", other variable related with land cover is the percent for nature vegetation and crop, PCT_CFT, PCT_NAT_PFT. Is there any possible method to get the annually modified landuse.timeseries. If manually revised the PCT_NAT_PFT in "landuse.timeseries_0.9x1.25_hist_16pfts_Irrig_CMIP6_simyr1850-2015_c170824.nc", i need to get land cover map in china with the same classification scheme as in PCT_NAT_PFT? It seems very complex.....

 

[oleson]

Yes, you'd have to revise the landuse timeseries and translate your land cover map types to those types CLM uses.

 

[xiaoxiaokuishu]

Yes, you'd have to revise the landuse timeseries and translate your land cover map types to those types CLM uses.

I am wonder Is the mksurfdata_map is suitable to preprare data for my case? Beside, my annual land cover of China has the classification scheme:
Cropland,Forest,Shrub,Grassland,Water,Sonw/Ice,Barren,Impervious,Wetland, that is totally different from PCT_NAT_PFT.

 

[oleson]

You are still going to have to do the mapping yourself. The tool ingests the same CLM pfts and landcover types that are on the surface dataset.

 

[xiaoxiaokuishu]

You are still going to have to do the mapping yourself. The tool ingests the same CLM pfts and landcover types that are on the surface dataset.

Thanks, It is clear now.

 

[xiaoxiaokuishu]

You are still going to have to do the mapping yourself. The tool ingests the same CLM pfts and landcover types that are on the surface dataset.

A further question, I have modified surfdata_0.9x1.25_hist_16pfts_Irrig_CMIP6_simyr2000_c190214.nc. So the land cover in surfdata is not consistent with the landuse.timeseries of 2000, that means the PCT_CROP in surfdata is not equals to PCT_CROP in landuse.timeseries at year 2000, I am wonder which land cover the model really use during running if my simulation time is 1999-2005 that covers 2000.

 

[slevis]

In a transient simulation, the model will use land cover from the landuse timeseries rather than from the fsurdat file.