National Climatic Data Center
6 February 2006
In February 2006, NCDC transitioned to the use of an improved Global Land and Ocean data set (Smith and Reynolds analysis (2005)) which incorporates new algorithms that better account for factors such as changes in spatial coverage and evolving observing methods.
NCDC's long-term mean temperatures for the Earth were calculated by processing data from thousands of world-wide observation sites on land and sea for the entire period of record of the data. Many parts of the globe are inaccessible and therefore have no data. The temperature anomaly time series presented here were calculated in a way that did not require knowing the actual mean temperature of the Earth in these inaccessible areas such as mountain tops and remote parts of the Sahara Desert where there are no regularly reporting weather stations. Using the collected data available, the whole Earth long-term mean temperatures were calculated by interpolating over uninhabited deserts, inaccessible Antarctic mountains, etc. in a manner that takes into account factors such as the decrease in temperature with elevation. By adding the long-term monthly mean temperature for the Earth to each anomaly value, one can create a time series that approximates the temperature of the Earth and how it has been changing through time.
The Smith and Reynolds analysis (2005) merges a new analysis of in situ SST anomalies [Smith and Reynolds, 2004] with an analysis of Land Surface Temperature (LST) anomalies from a gridded version of the Global Historical Climatology Network (GHCN) [Peterson and Vose, 1997]. The LST analysis is produced using the same methods as the SST analysis. The analysis method is briefly discussed here, but interested readers should see Smith and Reynolds [2004, 2005] for details.
Both the SST and LST components of the SR05 are created by separately analyzing the low- and high-frequency anomalies. Low-frequency anomalies are analyzed by spatial and temporal filtering when enough data are available. Spatial filtering is done by averaging anomalies over 10-15 degree latitude-longitude regions, and temporal filtering is done by averaging and median filtering over 15 year running periods. Separate low-frequency analysis is done to minimize the damping of those signals. Damping of the low-frequency may occur if it is analyzed by projecting it onto a set of stationary modes that do not fully resolve all of its variations.
The high-frequency residuals from this low-frequency analysis are analyzed separately by fitting them to a set of screened covariance modes representing large-scale monthly temperature patterns. It is assumed that the base period for the modes is long enough to resolve the high frequency variations. The sum of the low- and high frequency anomalies gives the total anomaly. In addition to the anomaly, an error estimate is also computed for the merged analysis.
Although the SR05 analysis is spatially complete, regions such as the Polar Latitudes are nearly always sampled poorly and the anomalies there are damped toward zero in the SR05 analysis. As expected, the SR05 sampling error estimate is large for the poorly-sampled regions. To prevent poorly-sampled regions from damping the global average, regions with large sampling errors are excluded from the global average. Sampling error, normalized by anomaly standard deviation, is used to define excluded regions. After testing several cut offs, it was decided to exclude regions with a normalized sampling error of 0.5. The amount of global area excluded is greatest in the 19th century, when it is 20%-30%. For the 20th century the area excluded is 20% or less, and after 1950 it is less than 15%.
Also, since the reconstruction is designed to resolve the large-scale variations, it does not always have as much spatial structure as the observations. Therefore, the land component of the merged temperature is adjusted to make it more consistent with the GHCN temperature. In 5-degree regions with no GHCN value available there is no adjustment. In other regions the strength of the adjustment towards the GHCN temperature is a function of the number of individual stations used to produce the GHCN. With one station in the 5-degree region the adjustment is to about 60% of the GHCN value, while for three or more stations the adjustment is an almost complete replacement with the GHCN value.
Peterson, T. C., and R. S. Vose (1997), An Overview of the Global Historical Climatology Network Temperature Database, Bull. Am. Meteorol. Soc., 78, 2837-2849.
Smith, T. M., and R. W. Reynolds (2004), Improved extended reconstruction of SST (1854-1997), J. Climate, 17, 2466-2477.
Smith, T. M., and R. W. Reynolds (2005), A global merged land air and sea surface temperature reconstruction based on historical observations (1880-1997), J. Climate, 18, 2021-2036.
Global Means Last Updated: 8 December 2006
Estimates of mean monthly global surface temperatures are given below with respect to the 20th century average (1901-2000). The figures are based on 1961-1990 estimates from the University of East Anglia's Climate Research Unit (UEA-CRU). The recently derived 1961-1990 global monthly surface temperature averages represent, in our opinion, the best absolute estimates of global mean temperature and were compiled at UEA-CRU by M. New, P.D. Jones, D.E. Parker and others . The data and methods used are described in the Monthly Gridded Time Series and in current and forthcoming publications (see below).
The UEA-CRU 1961-1990 estimates have been separated into land and sea components and adjusted using the longer-term global temperature anomaly time series from NCDC. The figures presented below therefore are mean monthly global surface temperature estimates for the 20th century, 1901 to 2000. Estimates for land (including Antarctica) and sea surface areas for the period 1901 to 2000 are given separately and in combined form.
Absolute estimates of global mean surface temperature are difficult to compile for a number of reasons. Since some regions of the world have few temperature measurement stations (e.g., the Sahara Desert), interpolation must be made over large, data sparse regions. In mountainous areas, most observations come from valleys where the people live so consideration must be given to the effects of elevation on a region's average as well as to other factors that influence surface temperature. Consequently, the estimates below, while considered the best available, are still approximations and reflect the assumptions inherent in interpolation and data processing. Time series of monthly temperature records are more often expressed as departures from a base period (e.g., 1961-1990, 1901-2000) since these records are more easily interpreted and avoid some of the problems associated with estimating absolute surface temperatures over large regions. For a brief discussion of using temperature anomaly time series see the Climate of 1998 series.
The global monthly surface temperature averages in the table below can be added to a given month's anomaly (departure from the 1901 to 2000 base period average) to obtain an absolute estimate of surface temperature for that month. (Files of absolute estimates are provided below.)
| Land Surface Mean Temp. |
J |
F |
M |
A |
M |
J |
J |
A |
S |
O |
N |
D |
Annual |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1901 to 2000 (°C) |
2.8 |
3.2 |
5.0 |
8.1 |
11.1 |
13.3 |
14.3 |
13.8 |
12.0 |
9.3 |
5.9 |
3.7 |
8.5 |
| 1901 to 2000 (°F) |
37.0 |
37.8 |
40.8 |
46.5 |
52.0 |
55.9 |
57.8 |
56.9 |
53.6 |
48.7 |
42.6 |
38.7 |
47.3 |
| Sea Surface Mean Temp. |
J |
F |
M |
A |
M |
J |
J |
A |
S |
O |
N |
D |
Annual |
| 1901 to 2000 (°C) |
15.8 |
15.9 |
15.9 |
16.0 |
16.3 |
16.4 |
16.4 |
16.4 |
16.2 |
15.9 |
15.8 |
15.7 |
16.1 |
| 1901 to 2000 (°F) |
60.5 |
60.6 |
60.7 |
60.9 |
61.3 |
61.5 |
61.5 |
61.4 |
61.1 |
60.6 |
60.4 |
60.4 |
60.9 |
| Combined Mean Surface Temp. |
J |
F |
M |
A |
M |
J |
J |
A |
S |
O |
N |
D |
Annual |
| 1901 to 2000 (°C) |
12.0 |
12.1 |
12.7 |
13.7 |
14.8 |
15.5 |
15.8 |
15.6 |
15.0 |
14.0 |
12.9 |
12.2 |
13.9 |
| 1901 to 2000 (°F) |
53.6 |
53.9 |
54.9 |
56.7 |
58.6 |
59.9 |
60.4 |
60.1 |
59.0 |
57.1 |
55.2 |
54.0 |
57.0 |
Erratum: Please note that prior to 26 June 2000, the mean values added to the land and ocean anomalies were incorrect. These data are now correct. Analysis of trends in the time series would not be impacted by this error since the error involved adding a constant to the entire period of record.
The complete land-sea surface climatology from the Climate Research Unit is described in:
Jones, P. D., M. New, D. E. Parker and S. Martin, submitted: Surface air temperature and its changes over the past 150 years. Rev. Geophys.
This climatology is actually a combination of four separate data sets:
Global land areas, excluding Antarctica, described in:
New, M. G., M. Hulme and P. D. Jones, in press: Representing 20th century space-time climate variability. I: Development of a 1961-1990 mean monthly terrestrial climatology. J. Climate.
Global oceans, 60S-60N, described in:
Parker, D. E., M. Jackson and E. B. Horton, 1995: The GISST2.2 sea surface temperature and sea-ice climatology. Climate Research Technical Note, CRTN 63, Hadley Centre for Climate Prediction and Research, Bracknel, UK.
Arctic sea areas, described in:
Rigor, I. G., R. L. Colony and S. Martin, submitted: Statistics of surface air temperature observations in the Arctic. J. Climate.
Martin, S. and E.A. Munoz: Properties of the Arctic 2-Meter Air temperature field for 1979 to the present derived from a new gridded data set. J. Climate, 10, 1428-1440.
NOTE: From February 2006 through April 14, 2006, the anomalies provided from the links below were inadvertently provided as departures from the 1961-1990 average. Anomalies are now provided as departures from the 20th century average (1901-2000).
Monthly and annual global anomalies are available through the most recent complete month and year, respectively.
http://www.ncdc.noaa.gov/oa/climate/research/anomalies/anomalies.html 