LTMIP EXPERIMENTS

Please read this file before downloading any data. Data can be accessed as 
individual, uncompressed files or an entire experiment directory can be 
downloaded as a compressed tar file. 

EXPERIMENT NAMES

Since the UVic2.8 model does not have a weathering component with climate 
feedbacks, experiments designated with "_csw" were not done. The experiment 
with all feedbacks is still designated as "_cswv" even though there is no 
weathering feedback. This was done so that the experiment name would be the 
same as for models with weathering feedbacks.

TIME

All experiments start from the same year 1800 equilibrium initial state. All 
model years start from year 1800.

The time in all output files is for the end of any averaging period. Since all 
output here is annual average all time should have 6 months subtracted from it 
to get the centre of the averaging period. Annual average output is written at 
midnight on January 1 of the next year so if you are just using the year to 
plot variables against, you should subtract 1 year to get the year that the 
data was averaged over. For example if an annual average is written at the end 
of 1994, the time stamp (ss:mm:hh:dd:mm:yyyy) will be 00:00:00:01:01:1995. In 
this case the "year" variable will be 1995 and the "time" variable will be 
1995. To get the correct time in decimal years you should subtract half the 
averaging period or 0.5, since 1994.5 represents the middle of the averaging 
period. An alternative is to just subtract 1 year since the year that is 
represented by the data is 1994 not 1995. I apologize for this confusion. It is 
an unfortunate legacy feature.

CALCULATING CARBON

Time Series Variables (ts_intgrls.nc files):
  LAND
    CV = total vegetation carbon (kg)
    CS = total soil carbon (kg)
    RESP_S = total flux from soil respiration (soil to air) (kg s-1)
    NPP = total flux from net primary production (air to vegetation) (kg s-1)
  ATMOSPHERE
    co2ccn = co2 concentration (ppmv)
  OCEAN
    dicbar = average dissolved inorganic carbon concentration (moles m-3 carbon)
    pbar = average phytoplankton concentration (moles m-3 nitrogen)
    zbar = average zooplankton concentration (moles m-3 nitrogen)
    dbar = average detritus concentration (moles m-3 nitrogen)
    diazbar = average diazotroph  concentration (moles m-3 nitrogen)
    dicflx = average air-sea flux (moles m-2 s-1 carbon)
  SEDIMENT
    ttrcal = average dissolution flux (sediment to ocean) (moles m-2 s-1 carbon)
    rain_cal = average rain flux (ocean to sediment) (moles m-2 s-1 carbon)
    weathflx = total weathering flux (sediment to ocean) (kg s-1)
  
  CONSTANTS (for calculating carbon inventories or fluxes):
    Ocean Surface (or Bottom) Area = 3.56684e14 m2
    Ocean Volume = 1.35769e18 m3
    Atmospheric conversion of ppmv to Pg = 2.12995
    Nitrogen to Carbon Redfield ratio = 7.
    seconds in a year = 86400.*365. = 3.1536e7
  
  INVENTORY (Pg):
    Atmospheric = co2ccn*2.12995
    Land = (CV+CS)*1.e-12
    Ocean = (dicbar + (pbar+zbar+dbar+diazbar)*7.)*12.e-15*1.35769e18
    Sediment = only the change in inventory can be calculated, either by
    (+Rock)    tracking the change in total Atmosphere + Land + Ocean 
               inventories or integrating the net ocean to sediment (burial)
               and land to ocean (weathering) fluxes.

  NET FLUXES (Pg year-1):
    air to land = (NPP-RESP_S)*1.e-12*3.1536e7
    air to sea = dicflx*12.e-15*3.56684e14*3.1536e7
    ocean to sediment (burial) = (rain_cal-ttrcal)*12.e-15*3.56684e14*3.1536e7
    land to ocean (weathering) = weathflx*1.e-12*3.1536e7

ADDITIONAL DATA

For most experiments complete three dimensional output is saved every 100 
years. Only selected output is on the web site to reduce the volume. Additional 
data is available on request.

ADDITIONAL EXPERIMENTS

Other doubling CO2 experiments, not specifically designated on the LTMIP web 
pages, are: 2xPI_c, 2xPI_cs, 2xPI_cswv. The naming convention follows that for 
the CO2 neutralization experiments. A run with historical CO2 but no feedbacks, 
or changes in other forcing, has also been done (Hist). This is comparable 
to the OCMIP historical CO2 experiments.

HISTORIC ANTHROPOGENIC INVENTORY

Several other variations of the HIST_cswv experiment have also been done. 
Another HIST_cswv integration was started from the year 1300 in order to access 
the models sensitivity to the initial equilibrium state. This experiment shows 
that the model uptake of CO2 from 1800 to today is quite sensitive to the 
initial state with diagnosed uptake differing by more than 10%. Since the HIST 
experiment starts from a year 1800 equilibrium it underestimates anthropogenic 
uptake from 1800 to present day by more than 10 Pg. This should not be an issue 
for emissions experiments but it should be considered when comparing the UVic 
model's anthropogenic inventory to other models that equilibrated to lower 
levels of CO2 (as in 1765). Output from additional HIST experiments is 
available on request.