ClimGen data for the QUEST-GSI project

Summary of status on 31-Mar-2009

(green=complete, pink=incomplete, red=see warning in info log)

Observations Click for info log or data

Prescribed-change time-slice scenarios Click for info log or data

Prescribed-change transient scenarios Click for info log or data

Emissions scenarios Click for info log or data

Data files

The data files (and in some case, graphics files containing figures) are grouped according to the structure below. It is essential to read the log, latest entry first, for each group so that you are aware of errors, updates, caveats etc.

When you click on the link to each group of files, you will find that you have two options (see the file starting 000_readme in each group): either download all files individually or download the files that have been grouped and compressed. The latter files begin with 00 and end with .tar.gz. These will be quicker and easier to download (fewer files and smaller overall), but you will need to uncompress and then ungroup them -- see the 000_readme file for further information.

All files are text (i.e., ASCII) files. The layout of the contents is fairly self-explanatory. Descriptions of the file format and the file naming system are available on the main ClimGen webpage.

Data files made to order (e.g. different format, combined files, specific regions) are available: click here for files made for CEFAS : click here for files made for Southampton.

Log -- please check this for updates, corrections, etc.

• Emissions-based scenarios from priority GCMs
  
  Click here to access the files for emissions-based transient scenarios from priority GCMs
  
  
  • 2009-03-31:
    • VAP scenarios for the final two priority GCMs are now available. For ukmo_hadgem1, the VAP changes are based on data interpolated to the surface from humidity changes in the atmosphere above the surface (because no surface humidity data were available in the CMIP3 database for this GCM). For csiro_mk3.0, the VAP changes are based on the multi-model mean changes from all GCMs that had humidity data in the CMIP3 database (because there were no surface or atmospheric humidity data in the CMIP3 database for this GCM).
  • 2008-10-30:
    • Future scenarios for SST (sea surface temperature) from all 7 priority GCMs for A1B (and from ukmo_hadcm3 also for A2, B1 and B2) have been added. Extra files for seasonal-mean time series for some quasi-standard regional averages are also available. The definition of these regions is shown in this map. All are compressed as before.
    • Note that the SST scenarios are based on the same pattern-scaling approach that was used for most other variables, such as TMP over land. Users should be aware that (i) the assumption of a linear relationship between local change and global-mean temperature change that is at the heart of the pattern-scaling approach is violated more frequently in GCM output over the oceans than over the land and thus these scenario data will differ more markedly from the original GCM simulation in some regions (e.g., in some GCMs, there is cooling, or weak warming, over parts of the North Atlantic Ocean in simulations that exhibit a significant weakening of the Atlantic thermohaline circulation, but this is often not proportional to the global-mean temperature change and may be linked with the rate of global-mean temperature change or some more complicated behaviour that cannot be captured by a pattern-scaling approach); and (ii) as with most other variables, the year-to-year variability is taken from the observational record, and the mean climate changes are added on to obtain the scenario; as a result, there are no changes in future ocean variability (e.g. in ENSO behaviour) in these scenarios, even though some GCMs simulate such changes.
  • 2008-10-23:
    • Future scenarios for WET (wet-day frequency) from all 7 priority GCMs for A1B (and from ukmo_hadcm3 also for A2, B1 and B2) have been added. See note below (2008-07-05 under time-slice scenarios) for an explanation of what this variable represents.
  • 2008-10-17:
    • Future scenarios for PRE (precipitation) from all 7 priority GCMs for A1B (and from ukmo_hadcm3 also for A2, B1 and B2) have been added.
  • 2008-09-14:
    • Future scenarios for VAP (vapour pressure) from 5 priority GCMs for A1B (and from ukmo_hadcm3 also for A2, B1 and B2) have been added. No surface humidity data are available for the other 2 priority GCMs (csiro_mk3.0 and ukmo_hadgem1) in the CMIP3 data base.
  • 2008-09-12:
    • Future scenarios for TMP (mean temperature), DTR (diurnal temperature range) and CLD (cloud cover) from all 7 priority GCMs for A1B (and from ukmo_hadcm3 also for A2, B1 and B2) have been added.
    • Note that for 4 GCMs, the change patterns for DTR are from the multi-model mean of all CMIP3 GCMs, though the time series of global temperature change used to scale these patterns is taken from MAGICC emulation of each GCM for each scenario.
  • 2008-09-08:
    • Data files are provided for 2006-2100 (because the observed record runs to 2005), except for CLD which covers 2003-2100 (because for CLD the observed record runs only to 2002). The same sequence of variability is repeated for 2010-2039, 2040-2069 and 2070-2099, and this sequence is derived from variability observed during 1961-1990. So these four 30-year periods can be intercompared.
    • Data files are provided with monthly data for all land grid boxes, plus monthly and seasonal-mean time series of country/national-averaged data. All files are compressed (see above) due to their large size.
    • When they are uncompressed, you will see that there appear to be more than one SRES scenario in each filename. This is because the filenames also include the name of the GCM simulations from which the climate change patterns were diagnosed (which are usually SRES scenarios themselves). The actual scenario for which the data have been generated is given AFTER _dt. So if the filename has ____dtB1__MG1__ in it, then this represents the B1 scenario derived from this model.
  
  Click here to access the files for prescribed-change transient scenarios from priority GCMs



• Prescribed-change transient scenarios from priority GCMs
  
  Click here to access the files for prescribed-change transient scenarios from priority GCMs
  
  
  • 2008-11-16:
    • Scenarios for SST from all 7 priority GCMs for a transient scenario with a prescribed global temperature change of +2.0 deg C by 2050 have been added to the archive. For ukmo_hadcm3, SST data for additional transient prescribed-warming scenarios of 1, 1.5, 2.5, 3, 3.5 and 4 degC have also been added.
  • 2008-11-15:
    • Future scenarios for DTR (diurnal temperature range) from 3 priority GCMs for +2 degC transient prescribed-warming scenarios (and from ukmo_hadcm3 also for 1, 1.5, 2.5, 3, 3.5, 4 degC transient prescribed-warming scenarios). For the remaining four GCMs (cccma_cgcm3.1, ipsl_cm4, mpi_echam5 and ukmo_hadgem1) no DTR or minimum/maximum temperarture data have been stored in the CMIP3 archive and therefore we cannot generate scenarios of DTR for these models. Rather than assuming no change in DTR for these models, we will instead use the multi-model mean DTR pattern computed from all those models that did provide DTR data to the CMIP3 archive. Multi-model mean scenarios are provided for the +2 degC transient prescribed-warming scenario.
    • Future scenarios for VAP (vapour pressure) from 5 priority GCMs for +2 degC transient prescribed-warming scenarios (and from ukmo_hadcm3 also for 1, 1.5, 2.5, 3, 3.5, 4 degC transient prescribed-warming scenarios). No surface humidity data are available for the other 2 priority GCMs (csiro_mk3.0 and ukmo_hadgem1) in the CMIP3 data base.
    • Scenarios for PRE, WET and CLD from all 7 priority GCMs for a transient scenario with a prescribed global temperature change of +2.0 deg C by 2050 have been added to the archive. For ukmo_hadcm3, PRE, WET and CLD data for additional transient prescribed-warming scenarios of 1, 1.5, 2.5, 3, 3.5 and 4 degC have also been added.
  • 2008-11-13:
    • Scenarios for TMP from all 7 priority GCMs for a transient scenario with a prescribed global temperature change of +2.0 deg C by 2050 have been added to the archive. For ukmo_hadcm3, TMP data for additional transient prescribed-warming scenarios of 1, 1.5, 2.5, 3, 3.5 and 4 degC have also been added.
  • 2008-11-13:
    • Data files are provided for 2006-2100 (because the observed record runs to 2005), except for CLD which covers 2003-2100 (because for CLD the observed record runs only to 2002). The same sequence of variability is repeated for 2010-2039, 2040-2069 and 2070-2099, and this sequence is derived from variability observed during 1961-1990. So these four 30-year periods can be intercompared.
    • Data files are provided with monthly data for all land grid boxes, plus monthly and seasonal-mean time series of country/national-averaged data. All files are compressed (see above) due to their large size.
    • When they are uncompressed, you will see that "_dtsA1B_nn__" appears in each filename. This indicates that the global-mean temperature time series that has been used to drive the pattern scaling for generating each scenario is based on the MAGICC multi-model mean simulations for the A1B scenario, and subsequently scaled so that the global temperature rise by the year 2050 is n.n degrees Celsius above the 1961-1990 mean. So, "_dtsA1B_20__" represents a transient global temperature rise of +2.0 degC above 1961-1990 by 2050.
  
  Click here to access the files for prescribed-change transient scenarios from priority GCMs



• Prescribed-change time-sclice scenarios from priority GCMs
  
  Click here to access the files for prescribed-change time-slice scenarios from priority GCMs
  
  
  • 2008-10-30:
    • Future scenarios for SST (sea surface temperature) for 1 prescribed amount of global warming (2.0 degrees Celsius above the 1961-1990 mean) have been added to the archive for the 7 priority GCMs. Extra files for seasonal-mean time series for some quasi-standard regional averages are also available. The definition of these regions is shown in this map. All are compressed as before.
    • Note that the SST scenarios are based on the same pattern-scaling approach that was used for most other variables, such as TMP over land. Users should be aware that (i) the assumption of a linear relationship between local change and global-mean temperature change that is at the heart of the pattern-scaling approach is violated more frequently in GCM output over the oceans than over the land and thus these scenario data will differ more markedly from the original GCM simulation in some regions (e.g., in some GCMs, there is cooling, or weak warming, over parts of the North Atlantic Ocean in simulations that exhibit a significant weakening of the Atlantic thermohaline circulation, but this is often not proportional to the global-mean temperature change and may be linked with the rate of global-mean temperature change or some more complicated behaviour that cannot be captured by a pattern-scaling approach); and (ii) as with most other variables, the year-to-year variability is taken from the observational record, and the mean climate changes are added on to obtain the scenario; as a result, there are no changes in future ocean variability (e.g. in ENSO behaviour) in these scenarios, even though some GCMs simulate such changes.
  • 2008-10-23:
    • When creating the WET (wet-day frequency) data for the other 6 GCMs, I found a relatively minor error in the "detrended observations" for the prescribed-change time-slice scenarios.
    • Today I have uploaded correct "detrended observations" WET data. If you downloaded the old "detrended observations" WET version (before 23rd October) then please delete them and download the new data files. I hope this hasn't caused you too much inconvenience.
    • Please note that this does not affect the WET data for the full observations (just the detrended observations), nor the WET data for any of the scenarios, all of which appear to be correct.
  • 2008-10-23:
    • Future scenarios for WET (wet-day frequency) for 1 prescribed amount of global warming (2.0 degrees Celsius above the 1961-1990 mean) have been added to the archive for the other 6 priority GCMs (data had already been added for ukmo_hadcm3). Extra files for country/national averages are there too. All are compressed as before.
  • 2008-10-17:
    • Future scenarios for PRE (precipitation) for 1 prescribed amount of global warming (2.0 degrees Celsius above the 1961-1990 mean) have been added to the archive for the other 6 priority GCMs (data had already been added for ukmo_hadcm3). Extra files for country/national averages are there too. All are compressed as before.
  • 2008-09-24:
    • Sorry, but I found that all the CLD data for the prescribed-change time-slice scenarios were wrong and so I deleted them on the 14th September. This applies to the "detrended observations", the +2 degC scenarios for the 7 priority GCMs, and the +0.5, 1, 1.5, 2.5, 3, 4, 5, 6 degC scenarios for ukmo_hadcm3.
    • Today I have uploaded correct CLD data. If you downloaded the old CLD versions (before 14th September) then please delete them and download the new data files. I hope this hasn't caused you too much inconvenience.
    • Please note that this does not affect the CLD data for the full observations (just the detrended observations), nor the CLD data for the emissions scenarios, all of which appear to be correct.
  • 2008-09-14:
    • Future scenarios for VAP (vapour pressure) from 5 priority GCMs for prescribed warming of 2 degrees Celsius above the 1961-1990 mean (and from ukmo_hadcm3 also for 0.5, 1, 1.5, 2.5, 3, 4, 5, 6) have been added. No surface humidity data are available for the other 2 priority GCMs (csiro_mk3.0 and ukmo_hadgem1) in the CMIP3 data base.
  • 2008-09-07:
    • Detrended observations for VAP (vapour pressure) have been add to the archive: vap_obscru_ts_3_00_detrended.... This is the baseline file with no climate change pattern added.
  • 2008-09-05:
    • Future scenarios for DTR (diurnal temperature range) for 1 prescribed amount of global warming (2.0 degrees Celsius above the 1961-1990 mean) have been added to the archive for two of the other priority GCMs (csiro_mk3.0 and ncar_ccsm3.0; data had already been added for ukmo_hadcm3). Extra files for country/national averages are there too. All are compressed as before.
    • For the remaining four GCMs (cccma_cgcm3.1, ipsl_cm4, mpi_echam5 and ukmo_hadgem1) no DTR or minimum/maximum temperarture data have been stored in the CMIP3 archive and therefore we cannot generate scenarios of DTR for these models. Rather than assuming no change in DTR for these models, we will instead use the multi-model mean DTR pattern computed from all those models that did provide DTR data to the CMIP3 archive. Multi-model mean scenarios for 1 prescribed amount of global warming (2.0 degrees Celsius above the 1961-1990 mean) have been added to the multimodelmean subdirectory for DTR, and should be used for all GCMs that do not have their own DTR scenarios.
    • The same rules as described below should be used to generate minimum and maximum temperature scenarios from the DTR and TMP data.
  • 2008-09-05:
    • Future scenarios for CLD (cloud cover) for 1 prescribed amount of global warming (2.0 degrees Celsius above the 1961-1990 mean) have been added to the archive for the other 6 priority GCMs (data had already been added for ukmo_hadcm3). Extra files for country/national averages are there too. All are compressed as before.
  • 2008-09-01:
    • Future scenarios for CLD (cloud cover) from the HadCM3 GCM for 9 prescribed amounts of global warming (0.5, 1.0, 1.5, 2.0, 2.5, 3.0, 4.0, 5.0 and 6.0 degrees Celsius above the 1961-1990 mean) have been added to the archive.
    • Due to the large file size (~170 MB per file, so a total of 9x170 = 1.5 GB), these are combined into a single compressed (gzip'd) archive 00cld.ukmo_hadcm3.tar.gz. I have also created monthly and seasonal-mean time series for every country, placed in separate compressed archives.
    • There is one additional data file that is not in a compressed archive: cld_obscru_ts_2_10_detrended.... This is the baseline file with no climate change pattern added.
  • 2008-09-01:
    • The directory has been reorganised so that there is a separate subdirectory for each GCM. Only the detrended observations remain in the main directory.
  • 2008-07-05:
    • Future scenarios for WET (wet day frequency) from the HadCM3 GCM for 9 prescribed amounts of global warming (0.5, 1.0, 1.5, 2.0, 2.5, 3.0, 4.0, 5.0 and 6.0 degrees Celsius above the 1961-1990 mean) have been added to the archive.
    • For ClimGen, as for the observations CRU TS 3.0, a wet-day is defined as a day with >0.1 mm of precipitation. However, note that the climate change scenarios for WET are estimated from an empirical function of the PRE scenario data, rather than from patterns of changes in wet-day frequency diagnosed directly from GCM simulations.
    • Due to the large file size (~170 MB per file, so a total of 9x170 = 1.5 GB), these are combined into a single compressed (gzip'd) archive 00wet.ukmo_hadcm3.tar.gz. I have also created monthly and seasonal-mean time series for every country, placed in separate compressed archives. The file 000contents_of_tar.gz_files lists the contents of each compressed archive.
    • There is one additional data file that is not in a compressed archive: wet_obscru_ts_3_00_detrended.... This is the baseline file with no climate change pattern added.
  • 2008-07-04:
    • Future scenarios for DTR (diurnal temperature range) from the HadCM3 GCM for 9 prescribed amounts of global warming (0.5, 1.0, 1.5, 2.0, 2.5, 3.0, 4.0, 5.0 and 6.0 degrees Celsius above the 1961-1990 mean) have been added to the archive.
    • Due to the large file size (~170 MB per file, so a total of 9x170 = 1.5 GB), these are combined into a single compressed (gzip'd) archive 00dtr.ukmo_hadcm3.tar.gz. I have also created monthly and seasonal-mean time series for every country, placed in separate compressed archives. The file 000contents_of_tar.gz_files lists the contents of each compressed archive.
    • There is one additional data file that is not in a compressed archive: dtr_obscru_ts_3_00_detrended.... This is the baseline file with no climate change pattern added.
    • Minimum and maximum temperature scenarios (TMN and TMX). In ClimGen output, the following rules are always true:
      • TMN = TMP - 0.5*DTR
      • TMX = TMP + 0.5*DTR
    • To reduce the volume of data provided, rather than generating TMN and TMX files, please just download the TMP and DTR files and use the rules above to obtain TMN and TMX.
  • 2008-07-03:
    • Future scenarios for PRE (precipitation) from the HadCM3 GCM for 9 prescribed amounts of global warming (0.5, 1.0, 1.5, 2.0, 2.5, 3.0, 4.0, 5.0 and 6.0 degrees Celsius above the 1961-1990 mean) have been added to the archive.
    • Due to the large file size (~170 MB per file, so a total of 9x170 = 1.5 GB), these are combined into a single compressed (gzip'd) archive 00pre.ukmo_hadcm3.tar.gz. I have also created monthly and seasonal-mean time series for every country, placed in separate compressed archives. The file 000contents_of_tar.gz_files lists the contents of each compressed archive.
    • There is one additional data file that is not in a compressed archive: pre_obscru_ts_3_00_detrended.... This is the baseline file with no climate change pattern added.
  • 2008-05-09:
    • Future scenarios for TMP from the 7 priority GCMs and for 9 prescribed amounts of global warming (0.5, 1.0, 1.5, 2.0, 2.5, 3.0, 4.0, 5.0 and 6.0 degrees Celsius above the 1961-1990 mean) have been added to the archive.
    • Due to the large file size (~170 MB per file, so a total of 7x9x170 = 10 GB), these are combined into a single compressed (gzip'd) archive per model. Let me know if you can't extract them using WinZip or an equivalent. The file 000contents_of_tar.gz_files lists the contents of each compressed archive.
    • There is one additional data file that is not in a compressed archive: tmp_obscru_ts_3_00_detrended.... Why do you need this observational file when I've already provided the CRU TS 3.0 observations? Well, the scenarios are made by combining detrended observations with the climate change pattern from each GCM; the observations file in this directory contains these detrended observations, with no climate change pattern added (the dtf0000 in the filename indicates that a fixed temperature change of zero has been added). The impact of a 2.5 degree global warming can be found by comparing the results under the 2.5 degree scenario with the results under these detrended observations.
    • All files contain years numbered arbitrarily as 2040-2069 (i.e., 30-year time series). These years are entirely arbitrary; the global-mean temperature change is prescribed and the climate scenarios (due to the pattern-scaling assumptions) are the same regardless of when such a level of warming might be reached. The data do not, therefore, really represent the period 2040-2069, they instead represent a 30-year sequence of climate under the prescribed amount of climate change.
    • The file names indicate the GCM used and the prescribed warming (e.g., ...dtf0250... is for fixed temperature change of 2.5 degrees Celsius above the 1961-1990 mean).
  
  Click here to access the files for prescribed-change time-slice scenarios from priority GCMs



• Global observations files for all QUEST-GSI users
  
  Click here to access the global observations files
  
  
  • 2008-10-23:
    • SST (sea surface temperature) observations derived from the HadISST1 data set have been added to the archive.
    • These have been interpolated onto the same 0.5 degree resolution grid as the land CRU TS data, and run from 1901-2007.
    • There are more than twice as many ocean boxes as land boxes, making this file very large (1.5 GB uncompressed) and so only a compressed file is available here 00sst.tar.gz.
    • In addition, seasonal-mean time series for some quasi-standard regional averages are also available. The definition of these regions is shown in this map.
  • 2008-06-10:
    • CRU TS 2.1 observations for all variables (except CLD which was already available) have been added in case anyone wishes to compare versions 2.1 and 3.0. Due to limited diskspace, these are only available as a single compressed file 00cru_ts_2_10.tar.gz.
  • 2008-06-10:
    • Monthly-mean time series averaged within each individual country have been added. Time series have been added for TMP, TMN, TMX, DTR, PRE, WET and VAP from CRU TS 3.0 (1901-2005) and for CLD from CRU TS 2.1 (1901-2002). You can save individual files for each variable, or get all variables in one go by saving 00countrymeans_monthly.tar.gz
  • 2008-05-08:
    • Seasonal-mean time series averaged within each individual country have been added (these are the data that I used to draw the country-mean time series graphs for comparing CRU TS 2.1 and CRU TS 3.0). Time series have been added for TMP, TMN, TMX, DTR, PRE, WET and VAP from CRU TS 3.0 (1901-2005) and for CLD from CRU TS 2.1 (1901-2002). You can save individual files for each variable, or get all variables in one go by saving 00countrymeans_seasons.tar.gz
  • 2008-05-07:
    • CRU TS 3.0 observations for 1901-2005 for WET (wet day frequency) added. Note that there is some variation between different studies in the threshold that determines whether a day is wet or not. The CRU TS 3.0 data, similar to previous versions (New et al.; Mitchell and Jones), attempts to count all days with >0.1 mm of precipitation as wet.
  • 2008-04-23:
    • CRU TS 3.0 observations for 1901-2005 for VAP (vapour pressure) added.
  • 2008-04-15:
    • Corrected CRU TS 3.0 observations for 1901-2005 for PRE (total monthly precipitation) have now replaced the preliminary data previously provided (see 2008-03-27 entry). The only change is that the erroneous data for Bangladesh from 1991-2000 has been corrected (this will also have some small affects on neighbouring countries). The winter 1991/2 data in Laos/Vietnam has not been modified because no obvious error could be found (at least one station did record unusually high precipitation during this season).
  • 2008-04-15:
    • File names for TMP, TMN, TMX, DTR and CLD have been slightly renamed to match the file name length that will be used for the future scenario data files. The contents of these files have not changed!
  • 2008-03-28:
    • Preliminary CRU TS 3.0 observations for 1901-2005 for PRE (total monthly precipitation) added. Note that these are preliminary data because I have identified an error in the precipitation data for Bangladesh from 1991-2000, and a possible error in the precipitation data for Laos/Vietnam in winter 1991/2. The CRU TS 3.0 project will provide corrected data in about 1 week from now, and I will replace this PRE data with the corrected version.
  • 2008-03-27:
    • Note that although these are "globally complete (for the land surface)" data sets, this has been achieved by interpolation from available station observations. Where no station observations were available within a reasonable distance, the values are instead adjusted towards their climatological "normal" value, usually the 1961-1990 mean for that location.
  • 2008-03-27:
    • CRU TS 3.0 observations for 1901-2005 for TMP (mean temperature), TMN (minimum temperature), TMX (maximum temperature) and DTR (diurnal temperature range) added. Note that they are not independent, because TMN=TMP-0.5*DTR and TMX=TMP+0.5*DTR.
  • 2008-03-20:
    • CRU TS 2.1 observations for 1901-2002 for CLD (Cloud Cover) added (the updated CRU TS 3.0 will not be available for CLD for some time, so use CRU TS 2.1 for this variable).
  
  Click here to access the global observations files



• Graphics files illustrating patterns of climate change from IPCC AR4 (i.e., CMIP3) GCMs
  
  Click here to access the precipitation pattern graphics
  Click here to access the temperature pattern graphics
  
  
  • 2008-03-25:
    • Maps showing patterns of precipitation change (for seasons DJF and JJA) diagnosed from all available CMIP3 GCMs have been aded. Each map is expressed as change in mm/day per degree of global warming. These maps, together with those shown in the IPCC AR4 WG1 report, supplementary information for chapter 10, have been used to assess GCM simulations for certain key regions/seasons, from which our selection of GCMs is guided.
  • 2008-03-25:
    • Maps showing patterns of temperature change (for seasons DJF and JJA) diagnosed from all available CMIP3 GCMs have been aded. Each map is expressed as change in local temperature per degree of global warming; so a value of 1 on the map indicates that, in this GCM, the local temperature changes at the same rate as the global-mean temperature; cooling is indicated by negative values.
  
  Click here to access the precipitation pattern graphics
  Click here to access the temperature pattern graphics



• Small region files for Richard Taylor
  
  Click here to access the files for Richard Taylor
  
  
  • 2008-03-25:
    • All variables for the observational record from CRU TS 2.1 have now been added for two further small regions.
  • 2008-03-25:
    • Correct VAP and WET files have been created for the observational record and are now available here. Note the new filename format, with the variable name at the start of the filename.
  • 2008-03-18:
    • All VAP and WET files deleted because found to be in error. Please delete all VAP and WET files downloaded before this date and wait for corrected versions.
  • 2008-01-23:
    • HadCM3 scenario data for 3 regions in Africa & China added. These data are preliminary and will be superseded when all scenarios are ready for all partners.
  • 2008-01-15:
    • Observed data for 3 regions in Africa & China added. Note that these data are from CRU TS 2.1 and run through to 2002 only. CRU TS 3.0 will be available later, running through to 2005 and also different to CRU TS2.1 in some regions through the 1901-2002 period.
  
  Click here to access the files for Richard Taylor