UAH satellite temperature dataset

Satellite temperature measurements

Satellites do not measure temperature directly. They measure radiances in various wavelength bands, from which temperature may be inferred. The resulting temperature profiles depend on details of the methods that are used to obtain temperatures from radiances. As a result, different groups that have analyzed the satellite data have obtained different temperature data. Among these groups are Remote Sensing Systems (RSS) and the University of Alabama in Huntsville (UAH). The satellite series is not fully homogeneous - it is constructed from a series of satellites with similar but not identical instrumentation. The sensors deteriorate over time, and corrections are necessary for satellite drift and orbital decay. Particularly large differences between reconstructed temperature series occur at the few times when there is little temporal overlap between successive satellites, making intercalibration difficult.

Description of the data

UAH provide data on three broad levels of the atmosphere.

The Lower troposphere - TLT (originally called T2LT).

The mid troposphere - TMT

The lower stratosphere - TLS

Data are provided as temperature anomalies against the seasonal average over a past basis period, as well as in absolute temperature values.

All the data products can be downloaded from the UAH server.

Geographic coverage

Data are available as global, hemispheric, zonal, and gridded averages. The global average covers 97-98% of the earth's surface, excluding only latitudes above +85 degrees, below -85 degrees and, in the cases of TLT and TMT, some areas with land above 1500 m altitude. The hemispheric averages are over the northern and southern hemispheres 0 to +/-85 degrees. The gridded data provide an almost global temperature map.

Temporal coverage

Daily global, hemispheric and zonal data are available. Monthly averages are available in gridded format as well as by hemisphere and globally.

Each set has data back to December 1978.

Comparison with other data and models

Climate models predict that as the surface warms, so should the global troposphere. Globally, the troposphere should warm about 1.2 times more than the surface; in the tropics, the troposphere should warm about 1.5 times more than the surface. For some time the only available satellite record was the UAH version, which (with early versions of the processing algorithm) showed a global cooling trend for its first decade. Since then, a longer record and a number of corrections to the processing have revised this picture: the UAH dataset has shown an overall warming trend since 1998, though less than the RSS version. In 2001, an extensive comparison and discussion of trends from different data sources and periods was given in the Third Assessment Report of the Intergovernmental Panel on Climate Change (IPCC) (section 2.2.4).

A detailed analysis produced by dozens of scientists as part of the US Climate Change Science Program (CCSP) identified and corrected errors in a variety of temperature observations, including the satellite data.

The CCSP SAP 1.1 Executive Summary states:

"Previously reported discrepancies between the amount of warming near the surface and higher in the atmosphere have been used to challenge the reliability of climate models and the reality of human induced global warming. Specifically, surface data showed substantial global-average warming, while early versions of satellite and radiosonde data showed little or no warming above the surface. This significant discrepancy no longer exists because errors in the satellite and radiosonde data have been identified and corrected. New data sets have also been developed that do not show such discrepancies."

The IPCC Fourth Assessment Report Summary for Policymakers states:

"New analyses of balloon-borne and satellite measurements of lower- and mid-tropospheric temperature show warming rates that are similar to those of the surface temperature record and are consistent within their respective uncertainties, largely reconciling a discrepancy noted in the TAR."

However, as detailed in CCSP SAP 5.1 Understanding and Reconciling Differences, neither Regression models or other related techniques were reconcilable with observed data. The use of fingerprinting techniques on data yielded that "Volcanic and human-caused fingerprints were not consistently identifiable in observed patterns of lapse rate change." As such, issues with reconciling data and models remain.

A potentially serious inconsistency has been identified in the tropics, the area in which tropospheric amplification should be seen. Section 1.1 of the CCSP report says:

"In the tropics, the agreement between models and observations depends on the time scale considered. For month-to-month and year-to-year variations, models and observations both show amplification (i.e., the month-to-month and year-to-year variations are larger aloft than at the surface). This is a consequence of relatively simple physics, the effects of the release of latent heat as air rises and condenses in clouds. The magnitude of this amplification is very similar in models and observations. On decadal and longer time scales, however, while almost all model simulations show greater warming aloft (reflecting the same physical processes that operate on the monthly and annual time scales), most observations show greater warming at the surface. "These results could arise either because "real world" amplification effects on short and long time scales are controlled by different physical mechanisms, and models fail to capture such behavior; or because non-climatic influences remaining in some or all of the observed tropospheric data sets lead to biased long-term trends; or a combination of these factors. The new evidence in this Report favors the second explanation."

The lower troposphere trend derived from UAH satellites (+0.128 °C/decade) is currently lower than both the GISS and Hadley Centre surface station network trends (+0.161 and +0.160 °C/decade respectively), while the RSS trend (+0.158 °C/decade) is similar. The surface station data indicate a trend of around 0.194 °C/decade, making the UAH and RSS trends 66% and 81% of the surface station derived value respectively.

For some time, the UAH satellite data's chief significance was that they appeared to contradict a wide range of surface temperature data measurements and analyses showing warming. In 1998 the UAH data showed a cooling of 0.05 K per decade (at 3.5 km - mid to low troposphere). Wentz & Schabel at RSS in their 1998 paper showed this (along with other discrepancies) was due to the orbital decay of the NOAA satellites. Once the orbital changes had been allowed for the data showed a 0.07 K per decade increase in temperature at this level of the atmosphere.

Corrections made

The table below summarizes the adjustments that have been applied to the UAH TLT dataset. The 'trend correction' refers to the change in global mean decadal temperature trend in degrees celsius/decade as a result of the correction.

The UAH TLT dataset was a source of controversy in the 1990s as, at that time, it showed little increase in global mean temperature, at odds with surface measurements. Since then a number of errors in the way the atmospheric temperatures were derived from the raw radiance data have been discovered and corrections made by Christy et al. at UAH.

The largest of these errors was demonstrated in a 1998 paper by Frank Wentz and Matthias Schabel of RSS. In that paper they showed that the data needed to be corrected for orbital decay of the MSU satellites. As the satellites' orbits gradually decayed towards the earth the area from which they received radiances was reduced, introducing a false cooling trend.

Even after the correction for satellite decay UAH continued to infer lower TLT temperatures than RSS based on the same raw data. For example Mears et al. at RSS found 0.193 °C/decade for lower troposphere up to July 2005, compared to +0.123 °C/decade found by UAH for the same period.

Much of the remaining disparity was resolved by the three papers in Science, 11 August 2005, which pointed out errors in the UAH 5.1 record and the radiosonde record in the tropics.

NOAA-11 played a significant role in a 2005 study by Mears et al. identifying an error in the diurnal correction that leads to the 40% jump in Spencer and Christy's trend from version 5.1 to 5.2.

Christy et al. asserted in a 2007 paper that the tropical temperature trends from radiosondes matches more closely with their v5.2 UAH-TLT dataset than with RSS v2.1.

Much of the difference, at least in the Lower troposphere global average decadal trend between UAH and RSS, has been removed with the release of RSS version 3.3 in January 2011. RSS and UAH TLT are now within 0.003 K/decade of one another. Significant differences remained, however, in the Mid Troposphere (TMT) decadal trends.

A beta version of 6.0 of the dataset was released on April 28, 2015 via blog post. It has not yet been published. This dataset has higher spacial resolution and uses new methods for gridpoint averaging.