Climate Change – News and Comments
Falsification tests of climate hypotheses
The trouble with clouds
Against the Danish physicist’s claim that cosmic rays influence the Earth’s low cloud cover and thereby the climate, there’s one contention that keeps turning up like the proverbial bad penny. During recent years, so the story goes, the Sun has been weak, cosmic rays have been relatively intense, and yet the expected increase in low clouds has not occurred. On the contrary, we’re told, low cloud cover has remained relatively sparse. That’s according the International Satellite Cloud Climatology Project, ISCCP, which pools data from the satellites of several nations,
The contention is repeated in a forthcoming paper in Journal of Climate by Ernest M. Agee, Kandace Kiefer and Emily Cornett of Purdue University, entitled “Relationship of Lower Troposphere Cloud Cover and Cosmic Rays: An Updated Perspective.” An advanced version of the full text is available from: http://curryja.files.wordpress.com/2011/09/agee-cosmic-rays.pdf A favourable commentary appears on the Ars Technica website: http://arstechnica.com/science/news/2011/09/do-cosmic-rays-set-the-earths-thermostat.ars?utm_source=rss&utm_medium=rss&utm_campaign=rss
Agee et al.’skey exhibit is their Fig. 2:
Over the period 1984-2008, cosmic ray variations (solid line) from a neutron counter at Kiel, Germany, are compared with cloudiness in the lower troposphere reported by ISCCP (broken line). Note the mismatch 2005-08. Agee et al. 2011, © American Meteorological Society.
And their abstract reads: An updated assessment has been made of the proposed hypothesis that “galactic cosmic rays (GCRs) are positively correlated with lower troposphere global cloudiness.” A brief review of the many conflicting studies that attempt to prove or disprove this hypothesis is also presented. It has been determined in this assessment that the recent extended quiet period (QP) between solar cycles 23-24 has led to a record high level of GCRs, which in turn has been accompanied by a record low level of lower troposphere global cloudiness. This represents a possible observational disconnect, and the update presented here continues to support the need for further research on the GCR-Cloud hypothesis and its possible role in the science of climate change.
There’s glory for you! – meaning a fine knock-down argument, as Humpty Dumpty said to Alice. To uninformed eyes (meaning, sadly, most climate scientists and commentators) the graph looks like a devastating falsification of the Svensmark hypothesis. Readers might even be surprised by the cautious language in the abstract, about a “possible observational disconnect”.
In fact the authors have every reason for caution. The conspicuous downward trend in the ISCCP cloud data is almost certainly unreal. An expert view is that it results from changes in the operational status of the satellites from which the data are pooled – see the references below to Campbell 2004, Campbell 2006 and Evan et al. 2007.
If a satellite views clouds from a slanting angle it sees more low clouds than when it’s looking straight down. Changes in the population and orbits of satellites contributing to ISCCP data have tended to narrow the viewing angle to nearer the vertical. That will have reduced the reported cloudiness even if, in the real world, the cloudiness were unchanging or even increasing. The effect is seen in these early maps from Campbell. 
Upper map: the trend in cloudiness from July 1983 to September 2001 across a grid box with 280 km squares, from the official ISCCP data with the annual cycle removed. Lower map: adjustment for the changing viewing angles of the satellites greatly reduces the areas of supposed loss of clouds (in blue). Campbell 2004.
The harsh fact is that supposedly real observations of clouds over the decades are in a state almost as parlous as the IPCC’s contradictory computer models of climate. Here is a summary of observed monthly cloud “anomalies” (i.e. variations) in five different data sets, published by the American Meteorological Society (ref. Arndt et al. 2010, see below). 
Black: ISCCP D2 Total cloud amounts from multiple satellites 1983-2008
Red: MISR Multi-angle Imaging Spectroradiometer on NASA’s Terra satellite 2000-2009
Blue: MODIS Moderate Resolution Imaging Spectroradiometer on NASA’s Terra and Aqua satellites 2000-2009
Brown: PATMOS-x Cloud data derived retrospectively from NOAA’s Advanced Very High Resolution Radiometer flown on a long succession of US spacecraft 1982-2009
Violet: SOBS Cloud amounts from surface weather observations 1971-1996.
In each case the solid lines are 12-month running means.
By cherry-picking favourable data (e.g. MISR and MODIS) I might try to claim that clouds have indeed increased with the high cosmic ray levels of the past decade. But judiciously one can only say that, as long as the data are so poor and contradictory, the jury must remain out, on what clouds have done and are doing. The last thing that Agee et al. or anyone should attempt with this shoddy stuff is to falsify the Svensmark hypothesis, for which plenty of other evidence exists. This includes variations in low clouds observed by satellites over days rather than decades, as in the Svensmark, Bondo and Svensmark 2009 paper summarized and referenced here https://calderup.wordpress.com/2010/05/03/do-clouds-disappear/ and in the recent Serbian paper that infers cloudiness from day-night temperature differences, as I reported here https://calderup.wordpress.com/2011/09/10/do-clouds-disappear-4/
The data on cloudiness over the longer term will be improvable by retrospective number-crunching, with PATMOS-x leading the way for total cloud. I’m very encouraged to see, in the last plot above, that PATMOS-x (brown) provides almost a mirror image of the ISCCP variations (black). But the low-level cloudiness may be more difficult to improve.
PS: For the philosophical importance of the falsifiability of hypotheses, according to Karl R. Popper, see https://calderup.wordpress.com/2010/05/01/falsification-intro/
References
It’s perhaps needless to say that none of these is to be found in Agee et al.’s paper.
G.G. Campbell, “View angle dependence of cloudiness and the trend in ISCCP cloudiness,” 13th AMS Conference on Satellite Meteorology and Oceanography, 2004
G.Garrett Campbell, “Diurnal and angular variability of cloud detection: consistency between polar and geosynchronous ISCCP products”, 14th AMS Conference on Satellite Meteorology and Oceanography, 2006
Amato T. Evan, Andrew K. Heidinger, and Daniel J. Vimont “Arguments against a physical long-term trend in global ISCCP cloud amounts” Geophysical Research Letters, 34, l04701, 2007
D.S. Arndt, M. O. Baringer, and M. R. Johnson, eds.: “State of the Climate in 2009”, Bull. Amer. Meteor. Soc., 91 (7), S1-S224, 2010
I’m promoting to the start of a new post a comment on an earlier post that came from Terry Sloan, together with my reply and his comment on my reply. I’ve included a graph that he sent in an e-mail because it wouldn’t upload into the Comments section. 
Posted by calderup 
Calder's Updates 
Further attempt to falsify the Svensmark hypothesis
05/10/2011Climate Change – News and Comments
Falsification tests of climate hypotheses
The trouble with clouds
The contention is repeated in a forthcoming paper in Journal of Climate by Ernest M. Agee, Kandace Kiefer and Emily Cornett of Purdue University, entitled “Relationship of Lower Troposphere Cloud Cover and Cosmic Rays: An Updated Perspective.” An advanced version of the full text is available from: http://curryja.files.wordpress.com/2011/09/agee-cosmic-rays.pdf A favourable commentary appears on the Ars Technica website: http://arstechnica.com/science/news/2011/09/do-cosmic-rays-set-the-earths-thermostat.ars?utm_source=rss&utm_medium=rss&utm_campaign=rss
Agee et al.’skey exhibit is their Fig. 2:
Over the period 1984-2008, cosmic ray variations (solid line) from a neutron counter at Kiel, Germany, are compared with cloudiness in the lower troposphere reported by ISCCP (broken line). Note the mismatch 2005-08. Agee et al. 2011, © American Meteorological Society.
And their abstract reads: An updated assessment has been made of the proposed hypothesis that “galactic cosmic rays (GCRs) are positively correlated with lower troposphere global cloudiness.” A brief review of the many conflicting studies that attempt to prove or disprove this hypothesis is also presented. It has been determined in this assessment that the recent extended quiet period (QP) between solar cycles 23-24 has led to a record high level of GCRs, which in turn has been accompanied by a record low level of lower troposphere global cloudiness. This represents a possible observational disconnect, and the update presented here continues to support the need for further research on the GCR-Cloud hypothesis and its possible role in the science of climate change.
There’s glory for you! – meaning a fine knock-down argument, as Humpty Dumpty said to Alice. To uninformed eyes (meaning, sadly, most climate scientists and commentators) the graph looks like a devastating falsification of the Svensmark hypothesis. Readers might even be surprised by the cautious language in the abstract, about a “possible observational disconnect”.
In fact the authors have every reason for caution. The conspicuous downward trend in the ISCCP cloud data is almost certainly unreal. An expert view is that it results from changes in the operational status of the satellites from which the data are pooled – see the references below to Campbell 2004, Campbell 2006 and Evan et al. 2007.
If a satellite views clouds from a slanting angle it sees more low clouds than when it’s looking straight down. Changes in the population and orbits of satellites contributing to ISCCP data have tended to narrow the viewing angle to nearer the vertical. That will have reduced the reported cloudiness even if, in the real world, the cloudiness were unchanging or even increasing. The effect is seen in these early maps from Campbell.
Upper map: the trend in cloudiness from July 1983 to September 2001 across a grid box with 280 km squares, from the official ISCCP data with the annual cycle removed. Lower map: adjustment for the changing viewing angles of the satellites greatly reduces the areas of supposed loss of clouds (in blue). Campbell 2004.
The harsh fact is that supposedly real observations of clouds over the decades are in a state almost as parlous as the IPCC’s contradictory computer models of climate. Here is a summary of observed monthly cloud “anomalies” (i.e. variations) in five different data sets, published by the American Meteorological Society (ref. Arndt et al. 2010, see below).
Black: ISCCP D2 Total cloud amounts from multiple satellites 1983-2008
Red: MISR Multi-angle Imaging Spectroradiometer on NASA’s Terra satellite 2000-2009
Blue: MODIS Moderate Resolution Imaging Spectroradiometer on NASA’s Terra and Aqua satellites 2000-2009
Brown: PATMOS-x Cloud data derived retrospectively from NOAA’s Advanced Very High Resolution Radiometer flown on a long succession of US spacecraft 1982-2009
Violet: SOBS Cloud amounts from surface weather observations 1971-1996.
In each case the solid lines are 12-month running means.
By cherry-picking favourable data (e.g. MISR and MODIS) I might try to claim that clouds have indeed increased with the high cosmic ray levels of the past decade. But judiciously one can only say that, as long as the data are so poor and contradictory, the jury must remain out, on what clouds have done and are doing. The last thing that Agee et al. or anyone should attempt with this shoddy stuff is to falsify the Svensmark hypothesis, for which plenty of other evidence exists. This includes variations in low clouds observed by satellites over days rather than decades, as in the Svensmark, Bondo and Svensmark 2009 paper summarized and referenced here https://calderup.wordpress.com/2010/05/03/do-clouds-disappear/ and in the recent Serbian paper that infers cloudiness from day-night temperature differences, as I reported here https://calderup.wordpress.com/2011/09/10/do-clouds-disappear-4/
The data on cloudiness over the longer term will be improvable by retrospective number-crunching, with PATMOS-x leading the way for total cloud. I’m very encouraged to see, in the last plot above, that PATMOS-x (brown) provides almost a mirror image of the ISCCP variations (black). But the low-level cloudiness may be more difficult to improve.
PS: For the philosophical importance of the falsifiability of hypotheses, according to Karl R. Popper, see https://calderup.wordpress.com/2010/05/01/falsification-intro/
References
It’s perhaps needless to say that none of these is to be found in Agee et al.’s paper.
G.G. Campbell, “View angle dependence of cloudiness and the trend in ISCCP cloudiness,” 13th AMS Conference on Satellite Meteorology and Oceanography, 2004
G.Garrett Campbell, “Diurnal and angular variability of cloud detection: consistency between polar and geosynchronous ISCCP products”, 14th AMS Conference on Satellite Meteorology and Oceanography, 2006
Amato T. Evan, Andrew K. Heidinger, and Daniel J. Vimont “Arguments against a physical long-term trend in global ISCCP cloud amounts” Geophysical Research Letters, 34, l04701, 2007
D.S. Arndt, M. O. Baringer, and M. R. Johnson, eds.: “State of the Climate in 2009”, Bull. Amer. Meteor. Soc., 91 (7), S1-S224, 2010