----------------------------------------------------------------------------------------- Case name: b30.108 Machine: Bluesky Run scripts: /fis/cgd/ccr/paleo/ccsm_runs/b30.108 Run directory: /ptmp/bruceb/b30.108 Data location: NCAR Mass Store at /CCSM/paleo/b30.108 Source code: CCSM3.0 (June 2004 release) with dynamic vegetation hydrology mods Resolution: T31 gx3 Initialization: Year 301 of the Holocene Dynamic Vegetation run- case b30.107.dv5, a T31 gx3 started from year 300 of the CCSM3 control PI run; PI = Pre-Industrial. This Holocene spin-up used trace gas values from the beginning of the greenhouse gas forcing dataset (see below) and solar constant from PI: < co2vmr = 2.63204e-4 < ch4vmr = 5.84438e-7 < n2ovmr = 2.60206e-7 < f11vmr = 0.0 < f12vmr = 0.0 < scon = 1.36500e+6 Orbital used that from previous 6kyr BP Holocene runs: < orb_year = -4050 Dynamic vegetation model results show an acceptible northward movement of grasses in the African Sahel, but not as much movement as desired in the Northern Hemisphere high-latitude boreal forests, compared to the PI dynamic vegetation run. It is not known if there are drifts in the deep ocean in the PI run used to initialize b30.107.dv5, and another 300 years of equilibration occurred during the dyanamic vegetation spin-up. In general the top-of-atmosphere radiation balance of the T31 gx3 CCSM3 is less than 0.1W/m2, implying very small ocean drfits. Run Frequency: 4 model years each submission Run Time: Approximately 5 hours 15 minutes on Bluesky with 2*32-way nodes = 64 processors total. This implies that at best the run will proceed at 16 years per day (~4 submissions per day), or about 500 model years per calendar month. Thus, at best one could hope to achieve 2500 year simulation in five months. Run Cost: 81 gaus/submission for csl_rg32 (regular) class. This implies 20,250 gaus/1000 years of run, or a total of 50,625 for a 2500 year run. Time: Previous Holocene runs used orbital year -4050, which is in calendar years 6kyr BP relative to 1950. Therefore, the transient was started for calendar year -4050 (4050 BC) which must be model year 1 for forcing dataset and interpolation reasons. Hence, calendar year of the run is: model_calendar_year = -4050 + modelYear - 1 Time coordination of the forcing datasets was necessary to ensure that the solar and volcanic historical forcings are accurate. Forcings: Ozone for PI All aerosols except for volcanic for PI Orbital- Standard CCSM3 formulae implemented as follows- Due to code limitations, the orbital year could not be updated annually, but rather with each submission. A new coupler namelist parameter "orb_refyear" was defined, and the orbital year computed on each run submission as: orb_year = orb_refYear + modelYear - 1 where the reference year is: orb_refyear = -4050 This means a orbital parameter resolution of 4 years. From the following table: year eccentricity obliquity(deg) longitude perhelion(deg) -4050 .01868182 24.10538 0.8696128 -4046 .01868085 24.10509 0.9361134 we can see that the changes in the orbital parameters are small enough to justify using four year update frequency. Solar constant- Supplied by Raimund, based on inferences from solar modulation parameter SM in MEV, inferred from C_14 and Be_10 ice core records. The solar irradiance (constant) first was fit to estimated satellite irradiance data for the last 25 years: scon = 1364.4 + 0.0030967*SM - 1.1723*10^(-6)*SM*SM (W/m2) where SM=solar modulation in MEV. Trends in the solar modulation data were adjusted so that solar modulation minima (~200Mev) were similar across entire record from -4050 to 1997. For the entire 6k record, we have: mean scon = 1365.729 scon min and max = 1364.756 1366.397 range of scon = 1.641 relative to 1364.4 = 0.001203 so that the amplitude of solar constant variations is on the smaller side of the full range proposed in recent years by various studies. The solar constant data vary annually, and are applied in CCSM3 as an adjustment at all wavelengths (i.e. no preference is given to uv, visible wavelengths over near-ir). The mean solar constant value is larger than the PI and Holocene spin-ups (1365.729 verse 1365.000), implying an initial warming of the run compared to the Holocene spin-up. However, the implied solar forcing ( 0.729/4 * .70 = 0.128 W/m2 is mostly offset by the mean volcanic forcing as described below). The solar irradiance values are punctuated by several Maunder-type minima during the first millenium or so (-4000 to -2500), followed by generally higher than the mean values with few Maunder minima from -2500 to -1500. There are higher frequency but lower amplitude variations of order a few centuries throughout the record. The solar irradiance forcing data can be found in netCDF format on: /fis/cgd/ccr/paleo/ccsm_runs/b30.108/data in the file: CCSM_solar_Transient-Holocene_from_-4050_to_1996_c051111.nc Greenhouse gases- Supplied by Fortunat, based on spline fits to ice core data. CO2, CH4, N2O are varied through the run on an annual basis. The data range from year -4050 to 2003. These trace gases vary slowly and smoothly during the first three millenia (-4050 to -1050) with 260-275 ppmv range for CO2, 590-610 ppbv for CH4, and 260-265 ppbv for N2O. For CCSM3, the freons F11 and F12 must be included, and their modern (i.e. non-zero) values were taken from past millenial simulations (850AD to 1999AD) done by Caspar. The greenhouse gas forcing data can be found in netCDF format on: /fis/cgd/ccr/paleo/ccsm_runs/b30.108/data in the file: CCSM_ghg_Transient-Holocene_from_-4050_to_2003_c051111.nc Volcanic- Supplied by Caspar, with statistical estimates of the distribution of the location, timing and magnitude of eruptions from -4050 to 849 based on statistics from 850 to 2000, while the 850 to 2000 forcings are from historical reconstructions. No tropical eruption is allowed to have a monthly volcanic mass larger than that estimated for Tambora 1815, while no extratropical (poleward of 30 degrees latitude, both hemispheres) eruption is allowed to have a monthly volcanic mass exceeding the peak of Laki, Iceland in 1783. Depending on the seasonality of the eruption, it is possible for global annual radiative forcing to slightly exceed (in magnitude) -6W/m2. Thus, forcing range imposed on the entire record is that estimated historically for the last two centuries. Three types of volcanoes: Tropical spread to both hemispheres depending on season; mass moves towards winter hemisphere with time scale order several months and e-folding to pre-eruption time scale order; 1 year tau_vis = 1.0 max (~ Tambora) Frequency 5 events/100 years NH 30 degrees latitude to pole tav_vis = 0.5 max Frequency 4 events/100 years SH 30 degrees latitude to pole tav_vis = 0.5 max Frequency 1 events/100 years Peak mass in vertical is from 50 to 80 hPa, range 30 to 120 hPa. Volcanic data is zonal and specified monthly during model run. The volanic aerosol is assumed to be sulphate, with a visual extinction coefficient of 3700 m2/kg. Only shortwave heating is allowed; in keeping with all other aerosols, no longwave heating is included. Auxillary History Fields: In addition to the standard monthly mean output for all components, the following fields were added to CAM output: fincl1 = 'FLN200','FLN200C','FSN200','FSN200C','MVOLC:A' The fluxes are net longwave and shortwave at 200 hPa and "MVOLC" is the mass of volcanic aerosol. Also, a daily mean set of fields is being written out from CAM: fincl2 = 'T850:A','U850:A','V850:A','T300:A','U300:A','V300:A', 'PSL:A','TREFHT:A','PRECL:A','PRECC:A' for purposes of storm track analysis and paleo proxy analysis. The former requires 850 and 300 hPa T,U,V. The latter requires daily mean 2m air temperature and precipitation. The total extrae data volume these fields add to the run is estimated at ~ 10%, while the impact on the running time is minimal (< 1% extra). Processing: We plan to put up a web site with some time series and other graphical analysis to monitor the run. More information on this will be forthcoming. Periodic progress reports will be emailed every few weeks or so. -----------------------------------------------------------------------------------------