02 Feb 2010, jet Processing to create 1995-2105 dataset for RCP8.5 transient runs from decadal avg data provided by Jean-Francois Lamarque. For the RCP8.5 case I produced two versions of the aerosol and ozone datasets to see the impact of using the same 2005 data that was used in the 20th century runs vs using the 2005 data that JF generated for the RCP runs. The original data used for 2005 was actually a 2000 climatology. Dataset used for 20th Century runs is: inputdata/atm/cam/ozone/ozone_1.9x2.5_L26_1850-2005_c091112.nc see ozone_1.9x2.5_L26_1850-2005_c091112.readme for details of creating that file. Version 1: New data for 2005: ozone_rcp85_v1_1.9x2.5_L26_1995-2105_c100202.nc Begin with the years 1995 and 2000 from the 20th century dataset. Append the new datasets using mid-decade timestamps, i.e., 2005 through 2105. Version 2: Old data for 2005: ozone_rcp85_v2_1.9x2.5_L26_1995-2105_c100202.nc Begin with the years 1995, 2000, and 2006 from the 20th century dataset. Change the time coordinate of the 2006 data to be 2005. Then append the new datasets using mid-decade timestamps as above. Since 2006 (renamed 2005) was really a 2000 climatology this dataset will provide ramp up forcing for the RCP8.5 case that will mimic the tail end of the 20th century runs where the climatological forcing from 2000-2005 was held constant. The details follow: Decadal files produced by JF for RCP85 % msls /LAMAR/IPCC_AR5/Climatologies_20100115/RCP85 aero_1.9x2.5_L26_2000-2009.nc aero_1.9x2.5_L26_2010-2019.nc aero_1.9x2.5_L26_2020-2029.nc aero_1.9x2.5_L26_2030-2039.nc aero_1.9x2.5_L26_2040-2049.nc aero_1.9x2.5_L26_2050-2059.nc aero_1.9x2.5_L26_2060-2069.nc aero_1.9x2.5_L26_2070-2079.nc aero_1.9x2.5_L26_2080-2089.nc aero_1.9x2.5_L26_2090-2099.nc aero_1.9x2.5_L26_2100-2109.nc oxid_1.9x2.5_L26_2000-2009.nc oxid_1.9x2.5_L26_2010-2019.nc oxid_1.9x2.5_L26_2020-2029.nc oxid_1.9x2.5_L26_2030-2039.nc oxid_1.9x2.5_L26_2040-2049.nc oxid_1.9x2.5_L26_2050-2059.nc oxid_1.9x2.5_L26_2060-2069.nc oxid_1.9x2.5_L26_2070-2079.nc oxid_1.9x2.5_L26_2080-2089.nc oxid_1.9x2.5_L26_2090-2099.nc oxid_1.9x2.5_L26_2100-2109.nc ozone_2000-2009.nc ozone_2010-2019.nc ozone_2020-2029.nc ozone_2030-2039.nc ozone_2040-2049.nc ozone_2050-2059.nc ozone_2060-2069.nc ozone_2070-2079.nc ozone_2080-2089.nc ozone_2090-2099.nc 1. Convert decadal avg files from r8 to r4 data, and remove the variable T. Previously we removed PS as well, but it turns out that the interpolation scheme uses PS if it is present and that results in a more accurate vertical interpolation to the model grid. This step is done using the cam interpolation routine from tools/interpic_new/. /blhome/eaton/bin/interpic -e T -p 4 -t ozone_2000-2009.nc ozone_2000-2009.nc ozone_2000-2009.r4.nc /blhome/eaton/bin/interpic -e T -p 4 -t ozone_2010-2019.nc ozone_2010-2019.nc ozone_2010-2019.r4.nc /blhome/eaton/bin/interpic -e T -p 4 -t ozone_2020-2029.nc ozone_2020-2029.nc ozone_2020-2029.r4.nc /blhome/eaton/bin/interpic -e T -p 4 -t ozone_2030-2039.nc ozone_2030-2039.nc ozone_2030-2039.r4.nc /blhome/eaton/bin/interpic -e T -p 4 -t ozone_2040-2049.nc ozone_2040-2049.nc ozone_2040-2049.r4.nc /blhome/eaton/bin/interpic -e T -p 4 -t ozone_2050-2059.nc ozone_2050-2059.nc ozone_2050-2059.r4.nc /blhome/eaton/bin/interpic -e T -p 4 -t ozone_2060-2069.nc ozone_2060-2069.nc ozone_2060-2069.r4.nc /blhome/eaton/bin/interpic -e T -p 4 -t ozone_2070-2079.nc ozone_2070-2079.nc ozone_2070-2079.r4.nc /blhome/eaton/bin/interpic -e T -p 4 -t ozone_2080-2089.nc ozone_2080-2089.nc ozone_2080-2089.r4.nc /blhome/eaton/bin/interpic -e T -p 4 -t ozone_2090-2099.nc ozone_2090-2099.nc ozone_2090-2099.r4.nc /blhome/eaton/bin/interpic -e T -p 4 -t ozone_2100-2109.nc ozone_2100-2109.nc ozone_2100-2109.r4.nc 2. Strip off 1995,2000, and 2006 from the 20th century control ncks -dtime,180,191 /fs/cgd/csm/inputdata/atm/cam/ozone/ozone_1.9x2.5_L26_1850-2005_c091112.nc ozone_1.9x2.5_L26_1995_c091112.nc ncks -dtime,192,203 /fs/cgd/csm/inputdata/atm/cam/ozone/ozone_1.9x2.5_L26_1850-2005_c091112.nc ozone_1.9x2.5_L26_2000_c091112.nc ncks -dtime,204,215 /fs/cgd/csm/inputdata/atm/cam/ozone/ozone_1.9x2.5_L26_1850-2005_c091112.nc ozone_1.9x2.5_L26_2006_c091112.nc 3a. Concatenate 1995,1995,2000 for version 1 - These three years will be labeled 1994,1995,2000. 1995 is copied into the 1994 spot to allow starting from jan 1 1995. 3b. Concatenate 1995,1995,2000,and 2006 for version 2 - These four years will be labeled 1994,1995,2000,2005. 1995 is copied into the 1994 spot to allow starting from jan 1 1995 and still have a ncrcat ozone_1.9x2.5_L26_1995_c091112.nc ozone_1.9x2.5_L26_1995_c091112.nc ozone_1.9x2.5_L26_2000_c091112.nc ozone_1.9x2.5_L26_1994-2000_c091112.nc ncrcat ozone_1.9x2.5_L26_1995_c091112.nc ozone_1.9x2.5_L26_1995_c091112.nc ozone_1.9x2.5_L26_2000_c091112.nc ozone_1.9x2.5_L26_2006_c091112.nc ozone_1.9x2.5_L26_1994-2006_c091112.nc 4a. On to the three years of version 1 concatenate years 2005-2105 of the decadal averaged datasets processed in step 1. 4b. On to the four years of version 2 concatenate years 2015-2105 of the decadal averaged datasets processed in step 1. ncrcat ozone_1.9x2.5_L26_1994-2000_c091112.nc ozone_2000-2009.r4.nc ozone_2010-2019.r4.nc ozone_2020-2029.r4.nc ozone_2030-2039.r4.nc ozone_2040-2049.r4.nc ozone_2050-2059.r4.nc ozone_2060-2069.r4.nc ozone_2070-2079.r4.nc ozone_2080-2089.r4.nc ozone_2090-2099.r4.nc ozone_2100-2109.r4.nc ozone_rcp85_v1_1.9x2.5_L26_1995-2105_c100202.nc ncrcat ozone_1.9x2.5_L26_1994-2006_c091112.nc ozone_2010-2019.r4.nc ozone_2020-2029.r4.nc ozone_2030-2039.r4.nc ozone_2040-2049.r4.nc ozone_2050-2059.r4.nc ozone_2060-2069.r4.nc ozone_2070-2079.r4.nc ozone_2080-2089.r4.nc ozone_2090-2099.r4.nc ozone_2100-2109.r4.nc ozone_rcp85_v2_1.9x2.5_L26_1995-2105_c100202.nc 3. Fix the time/date variables. Here is the .cdl file that contains the times and dates. The time values are "days since 1994-01-01 00:00" ncdump datetime_1994_2105.nc netcdf datetime_1994_2105 { dimensions: time = UNLIMITED ; // (168 currently) variables: int date(time) ; date:long_name = "date" ; date:format = "YYYYMMDD" ; double time(time) ; time:long_name = "time" ; time:units = "days since 1994-01-01 00:00" ; time:calendar = "noleap" ; time:bounds = "time_bnds" ; // global attributes: :Conventions = "CF-1.0" ; :source = "CAM" ; :case = "ar5_cam_1850-2000_03" ; :title = " " ; :logname = "lamar" ; :host = "be0207en.ucar.ed" ; :Version = "$Name$" ; :revision_Id = "$Id$" ; :initial_file = " " ; :topography_file = " " ; :history = " " ; :nco_openmp_thread_number = 1 ; data: date = 19940115, 19940215, 19940315, 19940415, 19940515, 19940615, 19940715, 19940815, 19940915, 19941015, 19941115, 19941215, 19950115, 19950215, 19950315, 19950415, 19950515, 19950615, 19950715, 19950815, 19950915, 19951015, 19951115, 19951215, 20000115, 20000215, 20000315, 20000415, 20000515, 20000615, 20000715, 20000815, 20000915, 20001015, 20001115, 20001215, 20050115, 20050215, 20050315, 20050415, 20050515, 20050615, 20050715, 20050815, 20050915, 20051015, 20051115, 20051215, 20150115, 20150215, 20150315, 20150415, 20150515, 20150615, 20150715, 20150815, 20150915, 20151015, 20151115, 20151215, 20250115, 20250215, 20250315, 20250415, 20250515, 20250615, 20250715, 20250815, 20250915, 20251015, 20251115, 20251215, 20350115, 20350215, 20350315, 20350415, 20350515, 20350615, 20350715, 20350815, 20350915, 20351015, 20351115, 20351215, 20450115, 20450215, 20450315, 20450415, 20450515, 20450615, 20450715, 20450815, 20450915, 20451015, 20451115, 20451215, 20550115, 20550215, 20550315, 20550415, 20550515, 20550615, 20550715, 20550815, 20550915, 20551015, 20551115, 20551215, 20650115, 20650215, 20650315, 20650415, 20650515, 20650615, 20650715, 20650815, 20650915, 20651015, 20651115, 20651215, 20750115, 20750215, 20750315, 20750415, 20750515, 20750615, 20750715, 20750815, 20750915, 20751015, 20751115, 20751215, 20850115, 20850215, 20850315, 20850415, 20850515, 20850615, 20850715, 20850815, 20850915, 20851015, 20851115, 20851215, 20950115, 20950215, 20950315, 20950415, 20950515, 20950615, 20950715, 20950815, 20950915, 20951015, 20951115, 20951215, 21050115, 21050215, 21050315, 21050415, 21050515, 21050615, 21050715, 21050815, 21050915, 21051015, 21051115, 21051215 ; time = 14, 45, 73, 104, 134, 165, 195, 226, 257, 287, 318, 348, 379, 410, 438, 469, 499, 530, 560, 591, 622, 652, 683, 713, 2204, 2235, 2263, 2294, 2324, 2355, 2385, 2416, 2447, 2477, 2508, 2538, 4029, 4060, 4088, 4119, 4149, 4180, 4210, 4241, 4272, 4302, 4333, 4363, 7679, 7710, 7738, 7769, 7799, 7830, 7860, 7891, 7922, 7952, 7983, 8013, 11329, 11360, 11388, 11419, 11449, 11480, 11510, 11541, 11572, 11602, 11633, 11663, 14979, 15010, 15038, 15069, 15099, 15130, 15160, 15191, 15222, 15252, 15283, 15313, 18629, 18660, 18688, 18719, 18749, 18780, 18810, 18841, 18872, 18902, 18933, 18963, 22279, 22310, 22338, 22369, 22399, 22430, 22460, 22491, 22522, 22552, 22583, 22613, 25929, 25960, 25988, 26019, 26049, 26080, 26110, 26141, 26172, 26202, 26233, 26263, 29579, 29610, 29638, 29669, 29699, 29730, 29760, 29791, 29822, 29852, 29883, 29913, 33229, 33260, 33288, 33319, 33349, 33380, 33410, 33441, 33472, 33502, 33533, 33563, 36879, 36910, 36938, 36969, 36999, 37030, 37060, 37091, 37122, 37152, 37183, 37213, 40529, 40560, 40588, 40619, 40649, 40680, 40710, 40741, 40772, 40802, 40833, 40863 ; } 4. Use ncgen to convert the cdl file above to netcdf. Then extract the time,date variables from this file and use them to overwrite the time,date variables in the file containing the concatenated time samples. % ncgen -b -o netcdf datetime_1994_2105.nc netcdf datetime_1994_2105.cdl % ncks --append -v date,time datetime_1994_2105.nc ozone_rcp85_v1_1.9x2.5_L26_1995-2105_c100202.nc % ncks --append -v date,time datetime_1994_2105.nc ozone_rcp85_v2_1.9x2.5_L26_1995-2105_c100202.nc 5. I used cdo to get global averages of the fields and checked the new datasets against the overlapping years from original 20th century dataset as well as the original double precision decadal files from JF. All the files seem OK. ex. cdo fldsum -vertsum ozone_rcp85_v1_1.9x2.5_L26_1995-2105_c100202.nc ozone_rcp85_v1_1.9x2.5_L26_1995-2105_c100202.sum.nc 6. Move to final location: % mv ozone_rcp85_v1_1.9x2.5_L26_1995-2105_c100202.nc \ /fis/cgd/cseg/csm/inputdata/atm/cam/ozone % mv ozone_rcp85_v2_1.9x2.5_L26_1995-2105_c100202.nc \ /fis/cgd/cseg/csm/inputdata/atm/cam/ozone