CRU : Projects : SO&P : Data
SO&P: Instrumental data
Gridded climate observations with data from 1851 or later
- Many of these data sets are available from the CRU data web page
- Temperature
- The gridded temperature data sets developed and maintained by CRU and the Met Office/Hadley Centre underwent a significant update at the beginning of 2003. SO&P work should use the new versions (Version '2') wherever possible.
- Data are accessible from the CRU gridded temperature data webpage on a 5° by 5° degree grid.
- Absolute = absolute temperature mean climatology (1961-1990 means).
- CRUTEM2 & CRUTEM2v = air temperature over land only.
- HadCRUT2 & HadCRUT2v = land and sea , a combination of air temperatures over land with sea surface temperatures.
- The 'v' indicates that artificial
changes in variance due to changes (over time) in the number of
observations in each grid box have been removed. Thus, CRUTEM2v and HadCRUT2v are preferred (but note that HadCRUT2v only extends back to 1870 because the
adjustment of variance was not made before 1870 for sea surface
temperatures).
- Precipitation
- Precipitation observations are only available over land for a long enough period for SO&P.
- There are two main choices here:
- Mike Hulme's global land precipitation available for 1900-1998 (though poorer coverage after 1995!) on either a 2.5° by 3.75° or a 5° by 5° grid. Despite what it says about the files being 'binary', I think you will find that they are ASCII files once they have been unzipped.
- CRUTS2.0
available for 1901-2000 on a 0.5° by 0.5° grid. CRUTS2.0
is a high resolution gridded data set of 5 climate variables over
land, including precipitation. They have been developed by a
number of CRU and Tyndall Centre staff, with the most recent
version being produced by Tim Mitchell. The use of
interpolation and infilling with average values means that great
care should be taken when using this data set to estimate climate
trends. The interpolation/infilling means that this data set
is less suitable for applications in SO&P, especially for
temperature (where CRUTEM2v and HadCRUTv are preferred). But it
has the advantage that it is at a very fine resolution and has
been adjusted for elevation effects - which might be useful for
precipitation.M
- Mean sea level pressure
- Monthly mean sea level pressure (SLP) data are available from
operational analyses from various national meteorological services, from
various re-analyses of observed data, and from combinations of
analyses/re-analyses with station records of SLP. Links are provided here
to three sources:
- The monthly SLP analyses of the UK Met Office from 1873 to 2000 are
available for the extra-tropical Northern Hemisphere only, from the CRU
gridded pressure data webpage on a 5° latitude by 10° longitude
grid. There are a number of problems with the data, especially (but not
only) in high
elevation areas prior to the 1920s (Rockies, Appalachians, Atlas and
Himalayas).
- Global monthly SLP data have been constructed by the Hadley Centre for
Climate Prediction and Research, by combining information from analyses,
reanalyses and station records, with additional homogenisation where
possible. The current version has removed some of the problems with the UK
Met Office analyses, and is called GMSLP3.2. It is available from 1871 to
1998 and can be requested from the Hadley Centre (information
here). A newer version which has further improvements and more recent
updates will be available soon called HadSLP1.
- NCEP/NCAR re-analyses provide many variables (mixture of observed and
model-derived information) globally for the period from 1948-present. Monthly SLP
are available from the Climate Diagnostics Center's NCEP/NCAR
reanalysis data pages.
Early instrumental records with data before 1851
- Gridded temperature 1781-1850
- Phil Jones has created an earlier version of his global gridded
temperature data set, using only stations that have data back before
1851. Coverage is almost completely limited to Europe, and this data set
goes back to 1781. It can be merged with his standard post-1851 land
temperature data set (CRUTEM2 -- see above).
- The reference for this early gridded data set is:
Jones PD, Moberg A, Osborn TJ and
Briffa KR (2003) Surface climate responses to explosive volcanic
eruptions seen in long European temperature records and mid-to-high
latitude tree-ring density around the Northern Hemisphere. In
Volcanism and the Earth's atmosphere (ed. Robock A and Oppenheimer
C), American Geophysical Union, Washington, DC (in press).
- Data format. The file contains data fields for every month from January
1781 to December 1850. Each field begins with a 1-line header giving the
month and the year, followed by 2592 data values (temperature anomalies
from the 1961-1990 mean, multiplied by 100, -9999 is missing),1s
followed by 2592 station count
values (number of stations with data in each grid box). The 2592 values
are for Jones' standard 5° by 5° grid (72 columns by 36 rows). The
order of values in this file is different to Jones' standard layout:
they begin, as usual, with the northern-most row (grid boxes centred at
87.5°N) and move along each row from west to east, and then down through
the rows until the southern-most row is reached; the difference is that each
row starts with the box centred at 2.5°E (i.e., immediately to the
east of the Greenwich meridian) and ends with the box centred at 357.5°E
(=2.5°W, i.e., immediately to the west of the Greenwich meridian).
-
Click here
to download the gzip'd crutem_17811850.dat.gz data file (size 0.2 Mb).
Climate indices
- North Atlantic Oscillation
- An NAO index based on the pressure difference between Gibraltar and
SW Iceland is available from the
CRU NAO data page.
- The raw station pressure observations are also available there, and
I have explained how they are used to compute the NAO index
in a document available here.
- Global temperature
- Monthly and annual global-mean temperature time series, and northern
and southern hemisphere time series, are available from the
CRU temperature data page.
Last updated: February 2003, Craig Wallace