UAH finds a warming error in satellite data, lowers global temperature trend, contradicts IPCC models

UAH finds a warming error in satellite data, lowers global temperature trend, contradicts IPCC models

From the University of Alabama, Huntsville via email from Dr. John Christy.

Weather Satellite Wanders Through Time, Space, Causing Stray Warming to Contaminate Data

In the late 1990s, the NOAA-14 weather satellite went wandering through time and space, apparently changing the record of Earth’s climate as it went.

Designed for an orbit synchronized with the sun, NOAA-14’s orbit from pole to pole was supposed to cross the equator at 1:30 p.m. on the sunlit side of the globe and at 1:30 a.m. on the dark side, 14 times each day. One of the instruments it carried was a microwave sounding unit (MSU), which looked down at the world and collected data on temperatures in Earth’s atmosphere and how those temperatures changed through time.

By the time NOAA-14 was finishing its useful life in 2005, however, it had strayed eastward from its intended orbit until it was crossing the equator not at 1:30 but at about 8:00. That pushed its early afternoon passage until after dark and its middle of the night measurements until well after dawn.

Diagram showing wander orbit of NOAA-14. Credit: University of Alabama Huntsville

Because local temperatures typically change between 1:30 and 8:00, this introduced spurious temperature changes that had to be calculated and removed from long-term temperature datasets that use data from satellite instruments.

The drift also changed the satellite’s orientation relative to the sun, so that instead of instruments being shielded from sunlight in a consistent way, the sun’s rays peaked into unshielded and open crevices and other places where intense sunlight could influence the sensors. It warmed the MSU, which caused it to look at the atmosphere’s mid-troposphere (from the surface up to about 40,000 feet) as very slightly warmer than it actually was relative to its initial 1:30 crossing time.

Using data from weather satellites that stayed closer to home than NOAA-14, scientists in the Earth System Science Center (ESSC) at The University of Alabama in Huntsville (UAH) have calculated how much false warming NOAA-14 reported so the false warming could be removed from a long-term global atmospheric temperature record collected by MSU’s on satellites since mid-November 1978.

Details of that research have been published in the International Journal of Remote Sensing, and are available online at:

The wandering satellite became an issue when Dr. John Christy, director of UAH’s ESSC and lead author of the study, and Dr. Roy Spencer, an ESSC principal research scientist, were updating and revising UAH’s satellite-based global temperature dataset. (Version 6.0 was completed in 2016.) While they knew NOAA-14 had strayed from its path, a closer look showed the warming reported by the MSU on NOAA-14 was out of kilter with temperature data collected by instruments on other NOAA satellites. This seemed to be especially true in the tropical mid-troposphere.

NOAA-14 was “drifting more than any other spacecraft used in this dataset,” said Christy.

“We were looking at 39 years of a temperature trend, and this stray satellite affected the trend by about 0.05 degrees Celsius (about 0.09° F) per decade,” said Christy said. “Over 39 years, that would be a total warming of about 0.2 C, or more than one-third of a degree Fahrenheit. And this problem occurred, almost all of it, in the 1990s and the early 2000s.

“An important piece of evidence pointing to a problem with the NOAA-14 satellite was its warming relative to the new NOAA-15 satellite that came in at the end of the 1990s,” Christy said.

To measure the scale of the problem after the UAH satellite dataset was finalized, the ESSC team started with a subset of U.S. weather balloons that hadn’t changed either instruments or software during at least a major part of the period NOAA-14 was in orbit. Weather balloons are a useful tool for benchmarking against the satellite data because both collect temperatures from the surface up through deep layers of the atmosphere.

When the U.S. balloons showed less warming than the MSU on NOAA-14, Christy expanded the study to include a group of Australian weather balloons that also hadn’t changed instruments during a major part of the time NOAA-14 was in orbit.

“This gave us two reputable datasets that are widely separated across the Earth’s surface,” Christy said. “Then we also looked at homogenized data from independent groups that correct balloon datasets.”

That data from NOAA, the University of Vienna and the University of New South Wales was added to data from three other groups — the European Center for Medium Range Forecasts, the Japan Climate Research Group and NASA — that “reanalyze” global weather data, correcting for flaws and problems.

UAH even created its own homogenized balloon dataset from raw balloon data archived by NOAA, which came from 564 stations around the world.

“We tried to understand the situation by inter-comparing against as many individual, independent datasets as possible,” Christy said. “We know no dataset is perfect, so comparing against various sources is a key part of dataset analysis. This allows us to zero in on places with the greatest discrepancy. We found the largest difference between the UAH dataset versus other satellite datasets — and even some with balloons — was found in the 1990s and early 2000s, the period of NOAA-14.”

And NOAA-14 was showing more warming than any of the balloon datasets, as well as more warming than NOAA-12 or NOAA-15, each of which overlapped NOAA-14 at some point during its time in service. They also compared temperature data from NOAA-15 to data from NASA’s orbit-stabilized AQUA satellite.

When the UAH team, led by Spencer, Christy and W. Daniel Braswell, built the Version 6.0 UAH dataset in 2016, they serendipitously did two things that limited the drift’s influence.

“First, we stopped using NOAA-14 in 2001, when it had drifted to 5 p.m.,” Christy said. “In our view, that was just too much drift to have confidence an accurate correction could be found.”

They also applied an algorithm that minimized the differences between the satellites, largely removing the NOAA-14 drift relative to the other satellites.

This resulted in a long-term mid-troposphere warming trend in the tropics of about 0.082 C (about 0.15° F) from late 1978 to 2016. This compares well with the +0.10 C (±0.03 C) per decade trend found by other sources that weren’t exclusively satellite based.

The way UAH built its dataset and accounted for these issues is unique among the four major satellite temperature datasets. The other three datasets still include all of the NOAA-14 data and show warming trends greater than the trend shown in the UAH dataset.

Other satellite-only temperature datasets report tropical mid-troposphere warming trends ranging from +0.13 to +0.17 C per decade.

“Not realizing it at the time, the methods we used to build the newest dataset appear to have dealt with a discrepancy that came to light through this inter-comparison study,” Christy said.

Note: This is not the lower tropospheric data or long-term warming trend reported each month for more than 27 years in UAH’s Global Temperature Report. The lower troposphere extends from the Earth’s surface to an altitude of about eight kilometers (more than 26,000 feet).

Because most low Earth orbiting satellites tend to stray somewhat from their intended orbits, new NOAA polar-orbiting weather satellites scheduled for launch in coming years will carry the extra fuel needed to keep them closer to home throughout their time in space.


The paper is fully open-access and available here:

Examination of space-based bulk atmospheric temperatures used in climate research


The Intergovernmental Panel on Climate Change Assessment Report 5 (IPCC AR5, 2013) discussed bulk atmospheric temperatures as indicators of climate variability and change. We examine four satellite datasets producing bulk tropospheric temperatures, based on microwave sounding units (MSUs), all updated since IPCC AR5. All datasets produce high correlations of anomalies versus independent observations from radiosondes (balloons), but differ somewhat in the metric of most interest, the linear trend beginning in 1979. The trend is an indicator of the response of the climate system to rising greenhouse gas concentrations and other forcings, and so is critical to understanding the climate. The satellite results indicate a range of near-global (+0.07 to +0.13°C decade−1) and tropical (+0.08 to +0.17°C decade−1) trends (1979–2016), and suggestions are presented to account for these differences. We show evidence that MSUs on National Oceanic and Atmospheric Administration’s satellites (NOAA-12 and −14, 1990–2001+) contain spurious warming, especially noticeable in three of the four satellite datasets.

Comparisons with radiosonde datasets independently adjusted for inhomogeneities and Reanalyses suggest the actual tropical (20°S-20°N) trend is +0.10 ± 0.03°C decade−1. This tropical result is over a factor of two less than the trend projected from the average of the IPCC climate model simulations for this same period (+0.27°C decade−1).

From the paper, here is the main conclusion:

One key result here is that substantial evidence exists to show that the processed data from NOAA-12 and −14 (operating in the 1990s) were affected by spurious warming that impacted the four datasets, with UAH the least affected due to its unique merging process.

RSS, NOAA and UW show considerably more warming in this period than UAH and more than the US VIZ and Australian radiosondes for the period in which the radiosonde instrumentation did not change. Additionally the same discrepancy was found relative to the composite of all of the radiosondes in the IGRA database, both global and low-latitude. While not definitive, the evidence does support the hypothesis that the processed satellite data of NOAA-12 and −14 are characterized by spurious warming, thus introducing spuriously positive trends in the satellite records. Comparisons with other, independently constructed datasets (radiosonde and reanalyses) support this hypothesis (Figure 10).

Figure 10. Magnitude of the relative difference between two periods for the respective satellite datasets (colored bars) and the respective radiosonde-based datasets (i.e. positive value indicates satellite warmed more than the radiosonde-based data between defined periods.).

Given this result, we estimate the global TMT trend is +0.10 ± 0.03°C decade−1. The rate of observed warming since 1979 for the tropical atmospheric TMT layer, which we calculate also as +0.10 ± 0.03°C decade−1, is significantly less than the average of that generated by the IPCC AR5 climate model simulations. Because the model trends are on average highly significantly more positive and with a pattern in which their warmest feature appears in the latent-heat release region of the atmosphere, we would hypothesize that a misrepresentation of the basic model physics of the tropical hydrologic cycle (i.e. water vapour, precipitation physics and cloud feedbacks) is a likely candidate.

Superforest,Climate Change

via Watts Up With That?

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s