Satellite temperature measurements

Satellite temperature measurements have been obtained for troposphere since 1978. By comparison, the usable balloon ( radiosonde) record begins in 1958.

Satellites do not measure "temperature" as such. They measure radiances in various wavelength bands, which must then be mathematically inverted to obtain indirect inferences of temperature. 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 to calculate temperature trends have obtained a range of values. Among these groups are Remote Sensing Systems (RSS) and the University of Alabama in Huntsville (UAH).

To compare to the increase from the surface record (of approximately +0.07 °C/decade over the past century and +0.17 °C/decade since 1979) it is more appropriate to derive trends for the lower troposphere in which the stratospheric cooling is removed. Doing this, through January 2008:

• RSS v3.1 finds a trend of +0.180 °C/decade
• UAH analysis finds +0.14 °C/decade .

An alternative adjustment introduced by Fu et al finds trends (1979-2001) of +0.19 °C/decade when applied to the RSS data set . A less regularly updated analysis is that of Vinnikov and Grody with +0.20°C per decade(1978 - 2004) .

Using the T2 channel (which include significant contributions from the stratosphere, which has cooled), Mears et al of Remote Sensing Systems (RSS) find (through February 2006) a trend of +0.135 °C/decade. Spencer and Christy of the University of Alabama in Huntsville (UAH), find a smaller trend of +0.054 °C/decade.

The satellite records have the advantage of global coverage, whereas the radiosonde record is longer. There have been complaints of data problems with both records, and difficulty reconciling the observations with climate model predictions.

Reconciliation of satellites, radiosondes and climate 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 showed cooling globally. A longer data series and several corrections to the UAH method leaves the UAH series showing warming, though less than RSS version. In 2001, an extensive, but now somewhat dated, 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) identifed and corrected errors in the satellite data and other temperature observations.

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 humaninduced 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 Reconcilling 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-causedfingerprints 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 behaviour; 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 satellite temperature record

Since 1979, Microwave Sounding Units (MSUs) on NOAA polar orbiting satellites have measured the intensity of upwelling microwave radiation from atmospheric oxygen. The intensity is proportional to the temperature of broad vertical layers of the atmosphere, as demonstrated by theory and direct comparisons with atmospheric temperatures from radiosonde (balloon) profiles. Upwelling radiance is measured at different frequencies; these different frequency bands sample a different weighted range of the atmosphere. Channel 2 is broadly representative of the troposphere, albeit with a significant overlap with the lower stratosphere (the weighting function has its maximum at 350 hPa and half-power at about 40 and 800 hPa). In an attempt to remove the stratospheric influence, Spencer and Christy developed the synthetic "2LT" product by subtracting signals at different view angles; this has a maximum at about 650 hPa. However this amplifies noise (Christy et al., J. Clim., 1998), increases inter-satellite calibration biases and enhances surface contamination (Fu and Johanson, GRL, 2005). The 2LT product has gone through numerous versions as various corrections have been applied.

Records have been created by merging data from nine different MSUs, each with peculiarities (e.g., time drift of the spacecraft relative to the local solar time) that must be calculated and removed because they can have substantial impacts on the resulting trend.

The process of constructing a temperature record from a radiance record is difficult. The best-known, though controversial, record, from Roy Spencer and John Christy at the University of Alabama in Huntsville (UAH), is currently version 5.2, which corrects previous errors in their analysis for orbital drift and other factors. The record comes from a succession of different satellites and problems with inter-calibration between the satellites are important, especially NOAA-9, which accounts for most of the difference between the RSS and UAH analyses . NOAA-11 played a significant role in the August 2005 Mears et al paper 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.

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 line with that estimated by climate models. In April 2002, for example, an analysis of the satellite temperature data showed warming of only 0.04 °C per decade, compared with surface measurements showing 0.17 +/- 0.06 °C per decade. The correction of errors in the analysis of the satellite data, as noted above, have brought the two data sets more closely in line with each other.

Christy et al. (2007) find that the tropical temperature trends from radiosondes matches closest with his v5.2 UAH dataset. Furthermore, they assert there is a growing discrepancy between RSS and sonde trends beginning in 1992, when the NOAA-12 satellite was launched. This research found that the tropics were warming, from the balloon data, +0.09 (corrected to UAH) or +0.12 (corrected to RSS) or 0.05 K (from UAH MSU; ±0.07 K room for error) a decade.

Discussion of the satellite temperature records

In the late 1990s the disagreement between the surface temperature record and the satellite records was a subject of research and debate. The lack of warming then seen in the records was noted.. A report by the National Research Council that reviewed upper air temperature trends stated:

"Data collected by satellites and balloon-borne instruments since 1979 indicate little if any warming of the low- to mid-troposphere—the atmospheric layer extending up to about 5 miles from the Earth's surface. Climate models generally predict that temperatures should increase in the upper air as well as at the surface if increased concentrations of greenhouse gases are causing the warming."

However, the same panel then concluded that

"the warming trend in global-mean surface temperature observations during the past 20 years is undoubtedly real and is substantially greater than the average rate of warming during the twentieth century. The disparity between surface and upper air trends in no way invalidates the conclusion that surface temperature has been rising."

As noted earlier, these temperature data, misinterpreted from the satellite data, are now known to have been too low.

An important critique of the satellite record is its shortness—adding a few years on to the record or picking a particular time frame can change the trends considerably. The problems with the length of the MSU record is shown by the table below, which shows the UAH TLT (lower tropospheric) global trend (°C/decade) beginning with Dec 1978 and ending with December of the year shown.

  1992     0.014
  1993    -0.020
  1994    -0.013
  1995     0.019
  1996     0.023
  1997     0.031
  1998     0.101
  1999     0.091
  2000     0.081
  2001     0.091
  2002     0.109
  2003     0.118
  2004     0.116
  2005     0.13

Likewise, even though they began with the same data, each of the major research groups has interpreted it with different results. Most notably, Mears et al. at RSS find 0.193 °C/decade for lower troposphere up to July 2005, compared to +0.123 °C/decade found by UAH for the same period. There are ongoing efforts to resolve these differences. Some believe that much of the disparity may have been 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.

Satellite measurements of the stratospheric temperature

The satellites also measure the lower stratospheric temperature and show a decline in stratospheric temperatures, interspersed by warmings related to volcanic eruptions. Global Warming theory suggests that the stratosphere should cool while the troposphere warms. However, the lower stratospheric record is mostly explained by the effects ozone depletion, which has caused a cooling of the stratosphere.

Weather balloons (radiosondes)

The longest data sets of upper air temperature are derived from instruments carried aloft by balloons (radiosondes). The radiosonde data set becomes usably global in about 1958. Changes in balloon instrumentation and data processing over the years have been pervasive, however, resulting in discontinuities in these temperature records. The Sherwood et al. study, published August 2005, looked at solar heating issues and found a spurious trend of about -.16K per decade had been introduced into the record, asserting that this masked the true warming, particularly in the tropical regions. This is enough to make the trend compatible with surface warming. Christy et al (2007) acknowledge the spurious trends in the radiosondes and assert that "[w]hen the largest discontinuities in the sondes are detected and removed," a cooler heating trend (in line with their UAH v5.2 dataset) than the one previously held is found.

The radiosondes and the MSU were designed to detect short term changes in temperatures and not long term trends so it would be inappropriate to criticize them for being poor for long term trend detection. Other problems with the radiosondes in addition to the recently discovered solar heating issue could remain in the data.