NASA’s Aqua satellite captured this microwave image of the Earth including Arctic sea ice cap on 3 September 2010.
NASA’s Aqua satellite captured this microwave image of the Earth, including the Arctic sea ice cap, on 3 September 2010. Satellite images such as this provide a compelling visual record of seasonal and longer-term changes in sea ice extent partially associated with global surface temperature change. A new version of NOAA’s Global Surface Temperature Dataset, released this June, is designed to ensure the best possible representation of historical climate conditions using data from a wide variety of sources. Credit: NASA Earth Observatory

Government agencies, businesses, academic researchers, and members of the public rely on climate information to support informed decision-making. This information includes data obtained on the ground and at sea, satellite data, and computational models that help with interpreting the data and that allow climate scientists to construct forecasts and scenarios. One key indicator for Earth’s climate system, global surface temperature (GST), is widely used in climate monitoring and assessments.

One of the most widely used GST data sets is the National Oceanic and Atmospheric Administration’s (NOAA) Global Surface Temperature Dataset (NOAAGlobalTemp), formerly known as the Merged Land-Ocean Surface Temperature (MLOST) [Smith et al., 2008]. Version 5 of this data set was released on 18 June 2019. This new version of NOAAGlobalTemp uses more comprehensive data collection and increased spatial coverage over land and ocean surfaces, as well as improved treatment of historical changes in observing practice.

Identifying and Monitoring Anomalies and Trends

Reports on temperature trends and anomalies constructed using this data set provide policy makers, business leaders, and the general public with information that is essential for making decisions associated with climate variability. Hence, it is important for NOAAGlobalTemp to be kept up to date, using the best available observational data.

High-impact applications of this data set include annual climate reports from the World Meteorological Organization and American Meteorological Society and the monthly global climate reports from NOAA’s National Centers for Environmental Information (NCEI) for the previous month, season, and year.

NOAAGlobalTemp enables analyses of temperature anomalies in various ways. For example, global anomalies maps describe regions where temperatures are above or below averages and by how much. Global percentiles maps illustrate how the temperature anomaly for a given map grid point ranks in comparison to previous years. This comparison informs users of any grid points where warm or cold temperatures set records or fell into the upper or lower decile.

Global trend maps show the rates at which temperatures are changing for each grid point. Global and continental time series provide the changing trends and fluctuations for regions like North America, South America, Europe, Africa, Asia, and Oceania.

Filling in Gaps over Data-Sparse Regions

Air temperature data over land surfaces in NOAAGlobalTemp version 5 are taken from the Global Historical Climatology Network-Monthly data set (GHCNm), which was updated from version 3.3.0 to version 4 in October 2018 [Menne et al., 2018]. GHCNm version 4 consists of data from approximately 26,000 surface stations, roughly 4 times as many as its predecessor (Figure 1). The increase in the number of stations and the use of estimates for missing base period (30-year) averages expand the geographic coverage of temperature anomalies throughout the record period.

Fig. 1. NOAAGlobalTemp visualization of land and ocean observations for November 2015, version 4 (top left) and version 5 (top right). Surface-drifting buoy positions are subsampled every 5 days for easier readability. NOAAGlobalTemp temperature anomalies reconstructed for November 2015 for version 4 (bottom left) and version 5 (bottom right). CLIMAT data come from land-based meteorological surface observation sites. VOSCLIM is Voluntary Observing Ship–Climate. USHCN is the U.S. Historical Climatology Network. GHCNd is the Global Historical Climatology Network–Daily. Click image for larger version.

The new version greatly expands spatial coverage in the 5° × 5° gridded field. Figure 2 illustrates the impact of these updates on decadal climate trends. Quality control, bias adjustment, and gridding procedures are largely the same as in the previous version. GHCNm version 4 contains a newly added comprehensive uncertainty budget that broadly follows the approach of Morice et al. [2012] for the land component of the U.K. Met Office’s Hadley Centre Climatic Research Unit Temperature (HadCRUT) product. In GHCNm version 4, major sources of uncertainty, from the station level monthly averages up to the calculation of regional means, are estimated primarily via a 100-member ensemble. This ensemble was produced to quantify random, systematic, and correlated error structures in the monthly temperature data, as well as uncertainties associated with the GHCNm version 4 process [Menne et al., 2018].

Fig. 2. Surface air temperature trends over land areas from 1988 to 2017. Trends are based on station records binned into 5° × 5° boxes. The latest version (right) shows greatly increased spatial coverage over the previous version (left). Click image for larger version.

Argo Floats Improve Coverage in the Southern Ocean and Tropical Regions

Over the ocean, NOAAGlobalTemp version 5 uses updated sea surface temperature (SST) data from version 5 of the Extended Reconstructed Sea Surface Temperature (ERSST) data set [Huang et al., 2017], which has improvements in both data and methods. Specifically, the new version of ERSST incorporates Argo float observations for improvements over the Southern and tropical oceans.

Additionally, ship and buoy data in NOAAGlobalTemp were updated to release 3.0 of NOAA’s International Comprehensive Ocean-Atmosphere Data Set (ICOADS), and sea ice data were updated to version 2 of the Hadley Centre Sea Ice and Sea Surface Temperature (HadISST2) data set.

Improved methods include better quality control, interpolation, and bias adjustments using a new baseline reference from more accurate modern buoy observations. ERSST version 5 has improved the representation of spatial variability over the oceans, the magnitude of El Niño and La Niña events, and the accuracy of absolute SST.

Land and Ocean Data Show Continuing Warming Trends

NOAAGlobalTemp version 5 trends support earlier research findings, showing the robustness of the warming trends and no slowdown or warming hiatus on decadal scales.

Using the new land and ocean surface temperature data sets, NOAAGlobalTemp version 5 employs a statistical reconstruction method [Smith et al., 2008] to generate global surface temperature data with 5° × 5° grids and monthly resolution. Over the global domain, the NOAAGlobalTemp version 5 trends are statistically consistent with the previous version. These trends further support earlier research findings over decadal and longer timescales, showing the robustness of the warming trends and no slowdown or warming hiatus on decadal scales.

For the 1880–2018 centennial scale, the warming rates are roughly 0.07°C/decade in both data sets (Figure 3). Warming rates have increased in recent decades: Both versions 4 and 5 show warming rates of about 0.14°C/decade from 1950 to 2018. This warming has become more rapid since the mid-1970s (about 0.17°C/decade from 1975 to 2018 in version 4 and 0.18°C/decade in version 5).

Warming rates are even higher in the most recent period beginning in the late 1990s (0.18°C and 0.19°C/decade for 1990 to 2018 in versions 4 and 5, respectively) and early 2000s (0.19°C and 0.20°C/decade for 2000 to 2018 in versions 4 and 5). (Numbers are obtained from the version runs for the January 2019 update.)

Fig. 3. Annual mean temperature anomaly time series from NOAAGlobalTemp version 5 (solid line) and version 4 (dotted line). The line for version 4 was shifted down by a constant of 0.07°C for the whole time period, largely because of the reference baseline shift from ship sea surface temperature (SST) to buoy SST—readings from buoys tend to be lower than readings from ships for a given location in general [Huang et al., 2017]. Global trends are about the same (at the 90% linear regression error confidence level as in the Intergovernmental Panel on Climate Change’s Fifth Assessment Report), although version 5 shows a very slightly larger warming trend than version 4 for 1880 to 2017.

Worldwide, several internationally recognized organizations actively and continuously improve the GST data sets. Although these organizations employ somewhat different input data and technical approaches, all data sets depict widespread warming over the long term—roughly 1°C since 1901 [Intergovernmental Panel on Climate Change, 2013; Zhang et al., 2016].

Continuing Improvements for Informed Decisions

The most notable improvement currently in progress addresses the incomplete coverage in the Arctic, where evidence of climate change is greatest.

Further improvements are under development for future releases. The most notable improvement currently in progress addresses the incomplete coverage in the Arctic, where evidence of climate change is greatest; the lack of full coverage has been shown to underestimate the global warming rate. Increasing the spatial resolution from 5° to 2° is another potential improvement area, as is further improving global uncertainty estimates and incorporating additional observations collected through ongoing collaboration with the international community.

NOAAGlobalTemp is part of the suite of climate products and services that NOAA provides to government, business, academia, and the public to support informed decision-making. This latest release is designed to ensure the best possible representation of historical climate conditions across the globe by leveraging newly available data and the latest peer-reviewed scientific methods. Data are freely available from the National Centers for Environmental Information.


This global data set is only possible through the international collaboration under the auspices of the World Meteorological Organization and through significant contributions by the NOAA Climate Program Office’s Ocean Observation and Monitoring Division and the Earth System Science and Modeling Division’s networks. We thank Kevin O’Brien for providing some metadata for Figure 1. The reviews from the NCEI internal review process and Eos reviewers and editor have made this article more readable.


Huang, B., et al. (2017), Extended Reconstructed Sea Surface Temperature version 5 (ERSSTv5): Upgrades, validations, and intercomparisons, J. Clim., 30, 8,179–8,205,

Intergovernmental Panel on Climate Change (2013), Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change, edited by T. F. Stocker et al., Cambridge Univ. Press, Cambridge, U.K.,

Menne, M. J., et al. (2018), The Global Historical Climatology Network Monthly Temperature Dataset, version 4, J. Clim., 31, 9,835–9,854,

Morice, C. P., J. J. Kennedy, N. A. Rayner, and P. D. Jones (2012), Quantifying uncertainties in global and regional temperature change using an ensemble of observational estimates: The HadCRUT4 data set, J. Geophys. Res., 117, D08101,

Smith, T. M., R. W. Reynolds, T. C. Peterson, and J. Lawrimore (2008), Improvements to NOAA’s historical merged land–ocean surface temperatures analysis (1880–2006), J. Clim., 21, 2,283–2,296,

Zhang, H.-M., et al. (2016), Recent development on the NOAA’s Global Surface Temperature Dataset, Abstract GC53H-04 presented at 2016 Fall Meeting, AGU, San Francisco, Calif., 12–16 Dec.

Author Information

Huai-Min Zhang (, Jay H. Lawrimore, Boyin Huang, and Matthew J. Menne, National Centers for Environmental Information, National Oceanic and Atmospheric Administration (NOAA), Asheville, N.C.; Xungang Yin, Riverside Technology, Inc., Asheville, N.C.; Ahira Sánchez-Lugo, Byron E. Gleason, Russell Vose, and Derek Arndt, National Centers for Environmental Information, NOAA, Asheville, N.C.; J. Jared Rennie, Cooperative Institute for Climate and Satellites, Asheville, N.C.; and Claude N. Williams, National Centers for Environmental Information, NOAA, Asheville, N.C.


Zhang, H.-M.,Lawrimore, J. H.,Huang, B.,Menne, M. J.,Yin, X.,Sánchez-Lugo, A.,Gleason, B. E.,Vose, R.,Arndt, D.,Rennie, J. J., and Williams, C. N. (2019), Updated temperature data give a sharper view of climate trends, Eos, 100, Published on 19 July 2019.

Text © 2019. The authors. CC BY-NC-ND 3.0
Except where otherwise noted, images are subject to copyright. Any reuse without express permission from the copyright owner is prohibited.