UAH vs RSS – Better with that?

news7g03/25/2022

31 7 minutes read

By Andy May

My recent review posts Related to HadCRUT5 surface and UAH Lower troposphere temperatures often degrade the accuracy of . RSS and UAH satellite temperature. Some seem to believe that RSS is more accurate than UAH, when radiosonde data shows that the opposite is true. So here is a short post to briefly present the problem.

My main source is blog post by Roy Spencer and a peer-reviewed article^[1] by John Christy, Roy Spencer, William Braswell and Robert Junod in 2018 this. The 2018 paper, here titled CSBJ18, presents a comprehensive analysis of the satellite’s four mid-troposphere temperature records. Three are global datasets: UAH (version 6), RSS (version 4) and NOAA (version 4). The fourth is the newer UW (University of Washington, 1st edition) series, which covers only the significant tropics between 30°S and 30°N.

Comparing satellite datasets is difficult, because while they share the same data, they use very different processes to produce their respective temperature records. In particular, the UAH procedure is quite different from the other three procedures. Whereas the traditional NOAA dataset was called “STAR”, they now have a new set, called RMTMT for the mid troposphere, which we will discuss near the end of the article.

CSBJ18 compares all satellite datasets with the 564 stations of the Integrated Global Radiosonde Archive (IRGA). The stations used have data from 1979 to 2016. The monthly satellite data is compared with the average monthly radiosonde (balloon) data using the global grid. CSBJ18 explains in detail the comparisons they made. Their procedures and methods are very thorough. Nearly all radiosonde records are on land, so comparisons with weather reanalysis datasets are also made because warming rates vary between land and ocean.

In all comparisons, both globally and in the tropics, the UAH satellite temperature record correlates best with radioisotopes. In addition, the UAH global temperature trend is lower than that of other data sets from 1979 to 2015 globally and for the tropics. The radiosonde data is not perfect, it also has false data, but it is independent of satellite records and provides a neutral, unbiased check of various methods of processing satellite data. There is no such test for different sets of surface temperature data, they all share the same data and most use the same methods to process it.

One of the reasons why UAH has a lower warming tendency than other datasets is that UAH corrected the apparent spurious data in the NOAA-11 satellite instruments to the older NOAA-14, and the other datasets do not. are not. These satellites used a previous MSU instrument and had orbital issues that needed to be fixed. NOAA-14 overlaps NOAA-15 for three years, and NOAA-15 has a much more advanced MSU ( AMSU), plus it has a better orbit in the overlapping period. Comparison of data from NOAA-15 to NOAA-14 demonstrated a problem with NOAA-14 equipment, details on how UAH fixed this problem and some more data are explained in a 2017 article by Spencer, Christy and Braswell.^[2] After all corrections were applied to NOAA-14 and NOAA-15 data, NOAA-14 still showed 0.2°C more warming per decade than NOAA-15. After applying only basic date correction, the two satellites differed by as much as 0.34°C/decade.^[1]

The full story of all the corrections and adjustments required to produce a satellite troposphere temperature record is too involved to be explained here but is fully documented in the references. references are cited. Here, we will only show the comparison results of Christy, et al. between satellite data sets and radiosonde data. Figure 1 compares global radiosonde data with satellite dates at similar locations. In this comparison, the UAH data correlated best with the unadjusted radiosonde data. The Y-axis is the correlation between satellite data and radio data, and the higher the better. Solid bars represent unadjusted data, fixed bars use the same satellite data, but adjusted radio data includes only radio co-waves that fit the data set The corresponding satellite has a correlation coefficient of 0.7 or higher. No changes have been made to the satellite dataset.

Figure 1. Correlation of three global satellite data sets and radiosonde temperature. Solid bars include all radioactive isotopes, unadjusted. Anti-interference bars use only radio results that correlate well with the corresponding satellite data set. The higher the correlation, the better, the UAH dataset correlates better with radioisotopes than other data sets. Source: (Christy, Spencer, Braswell, & Junod, 2018).

As Christy, et al. explained, while the UAH dataset correlates better than the other two satellite datasets, the difference in correlation approaches statistical significance, but does not reach that level. Figure 2 compares temperature trends from IGRA radioisotopes with global satellite datasets. Trends of unadjusted radio wave data, from 1979 to 2015, are shown in gray. Satellite data, for radio positions, is shown in green. The corrected radiosonde wave data is shown in red, and the entire global grid trend is shown in pink.

Figure 2. Comparison of warming trends from 1979 to 2015 for global three satellite and radio wave data. Source CSBJ18.

The UAH dataset fits radio warming trends better than the RSS and NOAA datasets. Better match applies to both adjusted and unadjusted data. At the locations of the radionuclides, the UAH trend is within 0.01°C/decade compared with the uncorrected radioacoustic trend. The RSS and NOAA datasets show more warming.

View from NOAA

It’s interesting and revealing that CSBJ18 hasn’t been refuted in any paper or blog post I could find. In fact, they seem to avoid discussing the earlier satellites, which are NOAA-14 and earlier satellites whose issues are discussed in CSBJ18 and later. Cheng-Zhi Zou, who helps maintain the NOAA satellite data set, wrote an article^[3] compares RSS, NOAA and UAH temperature records, but he started the comparison in 2002, shortly after the NOAA-15 was released with the more advanced AMSU microwave unit in it. This avoids comparing satellite records from NOAA-14 and earlier satellites containing more primitive MSU data. His Figure 4 (our Figure 3) compares records from 2002 to 2020.

Figure 3. Comparison of UAH v6, STAR (NOAA) v4.1, RSS v4 and NOAA RFTMT. The left plots record all four anomalous profiles at the same scale, the right plots are the difference from RFTMT. Note how similar RSS and UAH are in both sets of graphs. The NOAA STAR and NOAA RFTMT datasets are outliers.

The left-hand plots in Figure 3 compare the generated temperature anomalies and show that from 2002 to 2020, the different temperature series cover quite well, with occasional NOAA STAR exceptions. Red. The graph on the right shows the difference between the three satellite records listed and the new NOAA RFTMT dataset introduced by Zou, et al. NOAA’s new RFTMT is warmer than RSS or UAH on both land and ocean, and roughly equal. It is remarkable how close RSS and UAH are. Zou, et al. I want RFTMT to be the new standard satellite dataset and claim to be more accurate than the others, but I don’t like that.

Zou, et al. indicated that NOAA-15 had a problem at the end of its life and that STAR used it until 2015. This may cause some problems for NOAA STAR. UAH stopped using NOAA-15 in 2007 and RSS stopped using it in 2010. Zou, et al. recommends that their new RFTMT satellite record be used as a standard, but it appears this is just another attempt by NOAA to increase the warming trend with unwarranted corrections and satellite data collection. widely refined. Zou, et al. estimates that a satellite with the most advanced AMSU instrument, in a perfectly stable orbit and under ideal conditions, has a trend uncertainty of less than 0.04°C/decade over the period from 2002-2020, in line with previous estimates. The orbital features add to the uncertainty and the total global warming trend he reports is UAH: 0.17°C/decade, STAR: 0.17°C/decade, RSS: 0 .18°C/decade and RFTMT: 0.20°C/decade. The total uncertainty associated with all warming trends is about ±0.1°C/decade. Uncertainties by satellite are listed in Figure 3, you may need to click on Figure 3 to go to full scale to read values.

Calculating climate sensitivity with UAH

In addition to verifying that UAH radio temperature data is better than Christy and McNider’s RSS or NOAA datasets^[4] used the data to calculate the lower tropospheric climate sensitivity to greenhouse squeezing. This is best characterized as a TCR (transient climate response) estimate, and Christy and McNider refer to it as TTCR for tropospheric transient climate response.

They accept the IPCC’s assumptions that the only significant influences on climate since 1979 have been volcanic eruptions, anthropogenic greenhouse gas emissions, other human activities, and ENSO activity. They removed volcanism and ENSO effects from their lower tropospheric UAH record, and the baseline trend, without these effects, was about +0.1°C/decade.

They have been using IPCC AR5 values to force GHG since 1979 (1.45 W/m²) and total force (1.24 W/m²). With a corrected warming of 0.1°C/decade, they have a total warming time of 0.368°C. The time span is 38 years and the warming they calculate is 0.096°C, they I rounded to 0.1°C. Apply values to calculate TTCR, Christy, et al. calculation 1.1°C / 2xCO₂ ± 0.26, where “2xCO₂“Meaning double CO₂. This should be a more accurate estimate of the TCR that can be obtained from any surface profile, as more atmospheres are used in the calculation.

Conclude

Satellite temperature measurements are more useful than surface measurements in climate studies for a number of reasons:

The data used is all collected in the same way and using similar tools.
Extra atmospheric volume is included.
The measured temperature is mainly above the turbulent boundary layer of the atmosphere and is more stable.
Radiosonde data is available as an independent check of calculations.

Regarding #4 above, the UAH temperature calculations correlate best with radio wave data, suggesting that this is the best satellite temperature record. RSS’s decision to include NOAA-14 data with obvious flaws in their calculations is highly questionable when clearly superior NOAA-15 data is available for the time period in question. NOAA-15 developed problems later and both RSS and UAH dropped it early, but NOAA-STAR continued to use it for many more years.

Since the UAH temperature record is probably the higher one, the TCR estimate from it is reasonable. Christy and McNider did this and obtained a TCR of 1.1°C/2xCO₂. While this is much lower than the AR6, AR5 and AR4 values of 1.8°C/2xCO₂ (AR6: 1,2 to 2.4),^[5] it is in agreement with the estimates of Lewis and Curry, Lindzen and Choi, Alexander Otto et al., and others.^[6]

The 39% reduction in TCR is significant, especially because of the estimate by Christy, et al. drops below even AR6 probable range (1.4 to 2.2) and theirs most likely range (1,2 to 2.4).^[7] Sound very potential to me that AR6 is intentionally ignoring valid data and analytics to sell their political agenda.

References can be downloaded this.