HadCRUT and CRUTEM Surface (Air) Temperature Anomalies



This data set is only available for a restricted user group, please contact us if you want to access these data.

RESTRICTED only accessable in CEN/MPI net or via CliSAP login  What does that mean?

View data via LAS: CRUTEM3 | CRUTEM4 | HadCRUT3 |


Use data via OPeNDAP: CRUTEM | HadCRUT

Access data via file system:





Long-term observations of near surface air temperature and sea surface temperature play a key role for climate system analysis.

CRUTEM3 and CRUTEM4, HadCRUT3 and HadCRUT4, and the 1961-1990 temperature climatology (Land: CRU, Jones et al. 1999, Ocean: HadSST. which are offered via this website are gridded data sets of global historical land surface (and sea surface) temperature anomalies. HadCRUT4 and CRUTEM4 are from version Data are available for each month since January 1850, on a 5 degree grid. The data sets are collaborative products of the Met Office Hadley Centre and the Climatic Research Unit at the University of East Anglia.

CRUTEM3 and CRUTEM4 are limited to land stations; HadCRUT3 (HadCRUT4) is a blend of the CRUTEM3 (CRUTEM4) land-surface air temperature data set and the HadSST2 (HadSST3) sea-surface temperature data set.

Common to all these data sets is that these display the anomaly of the respective observations relative to the period 1961-1990. First the anomaly of the station values is computed and then all anomalies are gridded into a 5° x 5° grid.

Last update of this data set at ICDC: October 10, 2019.

To top


Name Unit Data Set
Surface air temperature anomaly [2m] K CRUTEM3 and 3v, CRUTEM4 and 4v
  HadCRUT3 and 3v, HadCRUT4
Sea surface temperature anomaly [0m] K HadCRUT3 and 3v, HadCRUT4
Number of observations per grid cell CRUTEM3 and CRUTEM4
Sampling error K CRUTEM3 and CRUTEM4
Station error K CRUTEM3 and CRUTEM4, HadCRUT3 and HadCRUT4
Combined measurement & sampling error K CRUTEM3 and CRUTEM4, HadCRUT3 and HadCRUT4
Absolute 2 m air temperature (1961-1990 Climatology) °C CRU
Absolute SST [0 m] (1961-1990 Climatology) °C HadSST.

To top

Coverage, spatial and temporal resolution

Period and temporal resolution:

  • 01/1850 - 07/2019
  • Monthly

Coverage and spatial resolution:

  • Global
  • Spatial resolution: 5° x 5°, cartesian grid
  • Geographic longitude: -177.5°E to 177.5°E
  • Geographic latitude: -87.5°N to 87.5°N
  • Dimension: 72 columns x 36 rows
  • Altitude: 2 m (CRUTEM), 0 m & 2 m (HadCRUT)


  • NetCDF

To top

Data quality

Estimates are made of the uncertainties arising from thermometer accuracy, homogenisation, sampling grid boxes with a finite number of measurements available, large-scale biases such as urbanisation and estimation of regional averages with non-complete global measurement coverage.

Because the number of observations forming each grid-box average in CRUTEM3 or CRUTEM4 is not constant in space or time, the variability in the gridded anomalies also changes: grid-box averages made from many observations have a smaller variance than those made from few. We have made a "variance adjusted" version of the dataset, where the anomalies have been statistically processed to homogenise the variances. These datasets are denote with an appended "v", i.e. CRUTEM3v and CRUTEM4v or HadCRUT3v. These data sets do not come with additional uncertainty estimates (of course).

For HadCRUT4 no variance adjusted version of the data set is available because the processing changed. Instead, the temperature anomaly given in the main data set is the median of 100 ensembles of the temperature anomaly distribution. We only offer this median value here; however all 100 ensembles as well as the error covariance matrices are available on request.

For more information we refer to the references.

To top


John Kennedy
E-Mail: john.kennedy (at) metoffice.gov.uk
Telefon: +44-1392-885105

Colin Morice
E-Mail: colin.morice (at) metoffice.gov.uk
Telefon: +44-1392-884566

Stefan Kern
ICDC / CEN / University of Hamburg
E-Mail: stefan.kern (at) uni-hamburg.de
Telefon: +49-40-42838-2415

To top

###BACKLINK###     ###TOPLINK###