skierinvermont Posted April 12, 2011 Share Posted April 12, 2011 A recent (March 24 2011) article at Skeptical Science provides a good basic summary of the various methods and problems associated with satellite temperature analyses. It's easy to read, and provides a decent summary of all the various methods that I have been posting about for the last few months. It's balanced and thoughtful. As you will see, there are many problems with satellite temperature analyses which produce large uncertainties when scientists try to compensate for these problems in various ways. Each method yields different results. The author suggests the most balanced approach at this point is to select an average of the cooler UAH/RSS estimates, and the warmer Fu, V&G, and Zou (STAR) methods. It seems likely to me that over the next few years we will find the Zou (STAR) raw T2 method become favored. One must then apply either the UAH, RSS, or Fu methods of removing stratospheric contamination since Zou 2010 is for T2 not TLT. http://www.skeptical...-Satellite.html Link to comment Share on other sites More sharing options...
BethesdaWX Posted April 12, 2011 Share Posted April 12, 2011 Good read, I have it added to my favs. Did you read the paper by John Christy and Roy Spencer? The trend on UAH was found to be at most +/- 0.05C/decade, most likely smaller after calibration, in the 2010 update. They also address having to correct for RAOBCORE & RICH error, that was not only picked up by ECMWF, but many other suites, regarding an "infrared channel". With the installation of AQUA, the trend/per decade will be even smaller. Since UAH will only improve as the years go on, we're probably set as far as monitoring temps. Link to comment Share on other sites More sharing options...
BethesdaWX Posted April 18, 2011 Share Posted April 18, 2011 Ok I've solved this. Amazing what happens when you read. Zou et al ( http://www.agu.org/journals/ABS/2006/2005JD006798.shtml ) his analysis is NOT dealing with direct error! He is removing the calibrations & correlations between satellites leaving the unadjusted error (before Chisty & Spencer adjust for it) and making a New Analysis Method, not addressing D-drift. Spencer/Christy's +/-0.05C/decade is after it has been corrected for...although there is still room for error at +/- 0.05C/decade ALSO: The debate ended a Month ago. This SkepticalScience post (that you linked) ( http://www.skepticalscience.com/Primer-Tropospheric-temperature-measurement-Satellite.html ) was Addressing the Discrepancy between RSS and UAH....guess what, the debate ended a Month ago! RSS has been adjusted downwards due to errors I've xplained...and are widely accepted throughout the climate science community! There ya go Link to comment Share on other sites More sharing options...
skierinvermont Posted April 18, 2011 Author Share Posted April 18, 2011 Ok I've solved this. Amazing what happens when you read. Zou et al ( http://www.agu.org/j...5JD006798.shtml ) his analysis is NOT dealing with direct error! He is removing the calibrations & correlations between satellites leaving the unadjusted error (before Chisty & Spencer adjust for it) and making a New Analysis Method, not addressing D-drift. Spencer/Christy's +/-0.05C/decade is after it has been corrected for...although there is still room for error at +/- 0.05C/decade ALSO: The debate ended a Month ago. This SkepticalScience post (that you linked) ( http://www.skeptical...-Satellite.html ) was Addressing the Discrepancy between RSS and UAH....guess what, the debate ended a Month ago! RSS has been adjusted downwards due to errors I've xplained...and are widely accepted throughout the climate science community! There ya go First of all that is not what the article is discussing, if you take the time to read it. Second of all, discrepancies between UAH and RSS have not been resolved.. this is simply fabrication on your part. Third, the error estimates for Zou are calculated in the same way they are by Spencer and Christy, and they find smaller error estimates but much more warming. Link to comment Share on other sites More sharing options...
BethesdaWX Posted April 18, 2011 Share Posted April 18, 2011 First of all that is not what the article is discussing, if you take the time to read it. Second of all, discrepancies between UAH and RSS have not been resolved.. this is simply fabrication on your part. Third, the error estimates for Zou are calculated in the same way they are by Spencer and Christy, and they find smaller error estimates but much more warming. RSS has been adjusted down matching UAH... it is now actually colder than UAH post 1998! NOw RSS V 3.3 instead of the old V3.2 http://www.ssmi.com/...msu_browse.html Changes from 3.2 to 3.3: Additional satellites are now included in the merge. Version 3.2 only used data from one AMSU instrument, NOAA-15. For TLT, TMT, and TLS, Version 3.3 includes data from the AMSU instruments on NOAA-15, AQUA, NOAA-18, and METOP-A. AMSU channel 7 exhibits unexplained drifts in METOP-A, so for TTS, data from METOP-A is not used. V3.2 is documented thoroughly in “Construction of the Remote Sensing Systems V3.2 atmospheric temperature records from the MSU and AMSU microwave sounders,” Carl A. Mears and Frank J. Wentz, accepted for publication by the Journal of Atmospheric and Oceanic Technology, December 2008, and "Construction of the RSS V3.2 lower tropospheric temperature dataset from the MSU and AMSU microwave sounders," Carl A. Mears and Frank J. Wentz, submitted to the Journal of Atmospheric and Oceanic Technology, September 2008. Version 3.3 is a direct extension of the methods described there. See the data description for details Link to comment Share on other sites More sharing options...
BethesdaWX Posted April 18, 2011 Share Posted April 18, 2011 Watch Skier deny that RSS has even been changed. this could be hilarious. Link to comment Share on other sites More sharing options...
skierinvermont Posted April 18, 2011 Author Share Posted April 18, 2011 The update to version 3.3 doesn't even begin to alter the long-term discrepancies between UAH and RSS. The 32 year RSS trend remains .163C/decade, while the UAH trend remains around .14C/decade. http://www.ssmi.com/...escription.html Link to comment Share on other sites More sharing options...
BethesdaWX Posted April 18, 2011 Share Posted April 18, 2011 Zou is removing the adjustements that create accuracy.....since he is leaving out Diurnal Drift, the of course it will show more warming. He isn't correcting for it, because that is not the point of his paper! Link to comment Share on other sites More sharing options...
BethesdaWX Posted April 18, 2011 Share Posted April 18, 2011 The update to version 3.3 doesn't even begin to alter the long-term discrepancies between UAH and RSS. The 32 year RSS trend remains .163C/decade, while the UAH trend remains around .14C/decade. http://www.ssmi.com/...escription.html The trend on RSS is now lower in V3.3 than in V3.2...again, RSS has been adjusted down! AKA: less warming. Link to comment Share on other sites More sharing options...
skierinvermont Posted April 18, 2011 Author Share Posted April 18, 2011 Zou is removing the adjustements that create accuracy.....since he is leaving out Diurnal Drift, the of course it will show more warming. He isn't correcting for it, because that is not the point of his paper! Removing diurnal drift doesn't create smaller error bars. In fact, the process of removing diurnal drift creates potential error. Link to comment Share on other sites More sharing options...
BethesdaWX Posted April 18, 2011 Share Posted April 18, 2011 Removing diurnal drift doesn't create smaller error bars. In fact, the process of removing diurnal drift creates potential error. Do you know what "excluding" diurnal drift means? Link to comment Share on other sites More sharing options...
skierinvermont Posted April 18, 2011 Author Share Posted April 18, 2011 good response. Since the statistics of these studies is clearly over your head, I recommend you take a statistics class. Link to comment Share on other sites More sharing options...
skierinvermont Posted April 18, 2011 Author Share Posted April 18, 2011 The trend on RSS is now lower in V3.3 than in V3.2...again, RSS has been adjusted down! AKA: less warming. Indeed, but the change is so small as to imperceptibly affect the long term trend, meaning that it remains significantly warmer than RSS. As do Fu, VG, and STAR, as well as the various radiosonde sources. Link to comment Share on other sites More sharing options...
BethesdaWX Posted April 18, 2011 Share Posted April 18, 2011 Indeed, but the change is so small as to imperceptibly affect the long term trend, meaning that it remains significantly warmer than RSS. As do Fu, VG, and STAR, as well as the various radiosonde sources. The adjustement is only after 1998....Post is what we're focusing on. It is now Colder than UAH post 1998. Significantly. Tell me when you're ready to stop with the Glaring errors and relax. Link to comment Share on other sites More sharing options...
BethesdaWX Posted April 18, 2011 Share Posted April 18, 2011 Do you know what "excluding" diurnal drift means? (nice edit to my above post) Although I'm not surprised you've resorted to those tactics at this point. Link to comment Share on other sites More sharing options...
skierinvermont Posted April 18, 2011 Author Share Posted April 18, 2011 Do you know what "excluding" diurnal drift means? (nice edit to my above post) Although I'm not surprised you've resorted to those tactics at this point. I didn't change it, you did. Nice try. Yes I know what excluding diurnal drift means. It means using side-view measurements or modelling techniques to remove the bias created by changing time of observation. Since these side-view measurements or modelling methods are not perfect, they introduce potential new error. Link to comment Share on other sites More sharing options...
BethesdaWX Posted April 18, 2011 Share Posted April 18, 2011 I didn't change it, you did. Nice try. Yes I know what excluding diurnal drift means. It means using side-view measurements or modelling techniques to remove the bias created by changing time of observation. Since these statistical or modelling methods are not perfect, they introduce potential new error. No, it means taking Diurnal drift out of the Equation quote-on-quote "exclude". I will post the direct quote if you wish. Link to comment Share on other sites More sharing options...
skierinvermont Posted April 18, 2011 Author Share Posted April 18, 2011 The adjustement is only after 1998....Post is what we're focusing on. It is now Colder than UAH post 1998. Significantly. Tell me when you're ready to stop with the Glaring errors and relax. oh really? Who says that's what we're focusing on? It's certainly not what I am or ever have focused on, or anybody else for that matter. The discrepancy is a long-term 32-year discrepancy, and it is not even remotely removed by the update to version 3.3. Link to comment Share on other sites More sharing options...
BethesdaWX Posted April 18, 2011 Share Posted April 18, 2011 oh really? Who says that's what we're focusing on? It's certainly not what I am or ever have focused on, or anybody else for that matter. The discrepancy is a long-term 32-year discrepancy, and it is not even remotely removed by the update to version 3.3. The adjustement is only post 1998....Data was not measured/calibrated the same in the 80's/90's. I don't see why you're focusing on the 80's/90's...the Cooling trend began after that anyway. RSS is now much colder than UAH post 1998. Link to comment Share on other sites More sharing options...
skierinvermont Posted April 18, 2011 Author Share Posted April 18, 2011 The adjustement is only post 1998....Data was not measured/calibrated the same in the 80's/90's. I don't see why you're focusing on the 80's/90's...the Cooling trend began after that anyway. RSS is now much colder than UAH post 1998. I'm not focusing on the 80s and 90s, I am looking at the full 32 year trend, for which RSS has a higher trend than UAH substantially. The adjustments from version 3.2 to 3.3 do not affect prior to 2002, because the adjustment was to include several other modern satellites that were not previously included in version 3.2. These satellites did not exist prior to 2002. As such, their inclusion did not alter the data prior to 2002. Link to comment Share on other sites More sharing options...
BethesdaWX Posted April 18, 2011 Share Posted April 18, 2011 I'm not focusing on the 80s and 90s, I am looking at the full 32 year trend, for which RSS has a higher trend than UAH substantially. The adjustments from version 3.2 to 3.3 do not affect prior to 2002, because the adjustment was to include several other modern satellites that were not previously included in version 3.2. These satellites did not exist prior to 2002. As such, their inclusion did not alter the data prior to 2002. Well I'm focusing on the time period after 1998, where data is MUCH better and less error prone in all satellites, mainly because there are many more satellites and better analysis methods today. Data in all Satellites gets better every decade as we discover new things, and continue to learn. In the 80's/90's, there is a higher error bar than in the 2000's. Link to comment Share on other sites More sharing options...
skierinvermont Posted April 18, 2011 Author Share Posted April 18, 2011 Well I'm focusing on the time period after 1998, where data is MUCH better and less error prone in all satellites, mainly because there are many more satellites and better analysis methods today. Data in all Satellites gets better every decade as we discover new things, and continue to learn. In the 80's/90's, there is a higher error bar than in the 2000's. Indeed, they may become more accurate over time, but as it stands the 32-year trends are prone to large errors and discrepancies between various satellite data sources. Link to comment Share on other sites More sharing options...
BethesdaWX Posted April 18, 2011 Share Posted April 18, 2011 Indeed, they may become more accurate over time, but as it stands the 32-year trends are prone to large errors and discrepancies between various satellite data sources. I agree, I never argued that. My argument is the fact that over the past decade, since the Lauch of AQUA in 2002, the error bar is +/- 0.05C/decade at most. Before then, in the 1980's/90's, error bars are likely greater. RSS is the "middle road" solution for that time period, so that may be a safer bet during that timeframe. Link to comment Share on other sites More sharing options...
Recommended Posts
Archived
This topic is now archived and is closed to further replies.