Climatic Research Unit
Google Earth interface for CRUTEM4 land temperature data
CRUTEM is a dataset derived from air temperatures near to the land surface
recorded at weather stations across all continents of Earth. It has been developed
and maintained by the Climatic Research Unit since the early 1980s, with funding
provided mostly by the US Department of Energy. The lead scientist for most of this
work was Professor Phil Jones, though many colleagues have also contributed. In
recent years, the Met Office Hadley Centre (MOHC) have also been involved, especially in
the regular updating of the operational version of CRUTEM (current version CRUTEM4).
CRUTEM has been combined with the MOHC's dataset of sea surface temperatures to provide
a near-global dataset of temperatures across Earth's surface, called HadCRUT. For example,
the current version HadCRUT4 combines CRUTEM4 and HadSST3. These datasets have been
widely used for assessing the possibility of anthropogenic climate change.
The latest version of CRUTEM is called CRUTEM4 and is available in text and netCDF
formats at the
Climatic Research Unit and at the
Met Office Hadley Centre.
To access CRUTEM4 through Google Earth, download the following KML file and open it in
New version available (17 August 2015):
<-- CLICK HERE to access CRUTEM4 in Google Earth
- Data up to June 2015 are included (note that annual temperature anomalies for 2014 are based the incomplete year January to June and will change when the remaining data become available).
From version CRUTEM22.214.171.124-2015-06 [see
Older versions still available:
- Data up to August 2014 (note that 2014 annual temperature anomalies are based on incomplete data, January to August only)
from version CRUTEM126.96.36.199-2014-08 [see
- Data up to December 2012
from version CRUTEM188.8.131.52-2013-03 [see
Also view this verion in Google Maps here (but limited functionality, Googe Earth is better!)
To avoid red/green colour problems, try this version with a grey checkerboard: CRUTEM4-2013-03_gridboxes_grey.kml
To facilitate direct access to visualisations and the underlying data values,
the current version is also made available via
An online rendition of the dataset -- the weather station monthly temperature data
and their locations, the grid-box monthly temperature anomalies, and seasonal and
annual timeseries graphs of all these data -- is accessed via an overlay written in
Keyhole Markup Language (KML). The KML file overlays these locations onto the
three-dimensional representation of the Earth provided by the Google Earth software,
enabling the dataset to be explored and accessed interactively and graphically.
This significantly enhances the accessibility of this key climate
dataset, whether for exploring the data and extracting regional information for
research and teaching, or for identifying any errors or limitations, without the
need to develop bespoke software to analyse the data. Given that the Google Earth
software is freely available and has been
downloaded more than one billion times
this represents an important additional dissemination route for the CRUTEM4 dataset.
IMPORTANT: on a Mac, running Mac OS X 10.10.0 or later, choosing to view
a larger annual or seasonal plot will crash Google Earth when you try to go "Back to Google Earth". There
is no fix for this as yet (September 2015)
How do I use it?
When you have clicked the KML file above, choose to open it with Google Earth. You will then see a dark and light grey (previously green and red)
checkerboard over the Earth. These are the boxes that we have weather station data in (note that sea surface temperatures
are not included in CRUTEM, only in HadCRUT). You can click any grey box you choose and it will show the
annual temperature anomaly timeseries for that grid box, plus clickable links to seasonal images and the grid-box data.
There is also a link called "Stations": clicking this will show the approximate locations of all weather stations in
our archive for that grid box with marker pins and the ID/name of the station. You can click any station pin to see
the annual temperature timeseries for that station, and to access the seasonal images and the station data itself.
The data files are in
a suitable format to import into a spreadsheet: save them and then open as comma-deliminated (CSV) files.
The primary reference for the CRUTEM4 dataset, which should be cited whenever CRUTEM4
data are used, is:
If you use the information provided by this Google Earth interface, then please also
acknowledge our work by citing:
- Jones P.D., Lister D.H., Osborn T.J., Harpham C., Salmon M. and Morice C.P., 2012:
Hemispheric and large-scale land surface air temperature variations: an extensive revision and an update to 2010.
Journal of Geophysical Research 117, D05127.
- Osborn T.J. and Jones P.D., 2014:
The CRUTEM4 land-surface air temperature dataset: construction, previous versions and dissemination via Google Earth.
Earth System Science Data 6, 61-68.
Notes (Questions & Answers)
- Q. What if I haven't got Google Earth on my computer/device?
- A. You can view some of the information using the Google Maps webpage. The link for this is given earlier.
You can click the red and green boxes to view the grid-box temperature anomalies, and access the other images
and the grid-box data. However, to access the stations in the grid box, right-click the "Stations" link in
the grid-box balloon and paste this link into the search box of the Googe Maps page.
The stations will appear, but all the grid-box elements will disappear. Google Earth is better!
- Q. Is the entire weather station database included?
- A. All weather stations that are used in the final CRUTEM4 dataset are included. Some weather stations
do not have sufficient measurements during the 1961-1990 reference period to be used. Of these unused stations,
those that are in grid boxes for which CRUTEM4 does have data (from other stations in the box with longer records)
are included, while those in grid boxes with no CRUTEM4 data are not. The full station database can be obtained
- Q. Why are weather stations in the wrong locations?
- A. The information we currently have about the latitude/longitude of each station is limited
to 1 decimal place, so the station markers could be a few kilometres from the actual location.
This is adequate for the construction of the CRUTEM gridded and global temperature records, because they
do not depend on the precise location of each station. WMO/GCOS have asked member states to provide more
accurate location details in future initiatives.
- Q. Why do the grid-box temperature graphs appear to contain more years with data
than the station temperature graphs within the same grid box?
- A. For the station temperature graphs, complete data are required to form an annual
or seasonal average (i.e. 12 full months or 3 full months, respectively). For the grid-box
temperature anomaly graphs, annual or seasonal averages are shown even if up to one third
of the individual monthly values are missing.
The reason for the different approach is
that seasonal variations can be large and therefore an annual-mean value could be biased
significantly higher if a winter month value was missing. For the absolute values shown
for the station temperature graphs, complete data are required to avoid such artefacts.
For the grid-box data, all values are temperature anomalies and the conversion to anomalies
removes the mean seasonal cycle, so any bias in the results due to a small number of
missing values is likely to be much smaller.
- Q. For a small number of stations, there is no annual temperature graph shown when the station marker
is clicked (for example, some stations in the grid box containing Baghdad, Iraq).
- A. If there is only one year or less with complete data, then an annual temperature graph has not been
drawn for that station.
The "Data" link shows the raw data, where this can be verified. The "Seasonal image" link may show
seasonal temperature graphs if there are complete data for more than one year in at least one season.
These stations are not used for constructing the gridded CRUTEM dataset because they do not have sufficient
observations during the reference period.
Last updated: August 2015 by Tim Osborn