The influence of anthropogenic surface processes and inhomogeneities

The influence of anthropogenic surface processes and inhomogeneities on gridded global climate data Ross Mc. Kitrick Department of Economics University of Guelph ON Canada Presentation to the American Chemical Society Denver CO via Webinar August 28 2011

Surface Climate Data l The “global temperature” ross. mckitrick. weebly. com 2

Summary l l Climate data is the output of a model l Raw data: daily T-Min and T-Max readings from inhabited places l This isn’t what the climate analyst is interested in: it must be converted into “climate data” using a statistical adjustment model. How do we know the adjustment model “works”? l Many papers merely describe the adjustment steps in enthusiastic detail l I have focused on devising statistical tests of the results ross. mckitrick. weebly. com 3

Conclusions l Based on analysis of multiple data sets, and after addressing a long list of statistical rebuttals, I find the evidence convincing that: l The adjustment models are inadequate l The resulting climate record over land is contaminated with patterns of socioeconomic development l This adds a net warming bias to the global trend and may lead to misattribution of spatial patterns to greenhouse gases l A valid empirical model of the spatial pattern of observed warming must include anthropogenic surface processes ross. mckitrick. weebly. com 4

Papers l Mc. Kitrick, Ross and Patrick J. Michaels (2004). “A Test of Corrections for Extraneous Signals in Gridded Surface Temperature Data” Climate Research 26 pp. 159 -173. l Mc. Kitrick, Ross R. and Patrick J. Michaels. (2007) “Quantifying the influence of anthropogenic surface processes and inhomogeneities on gridded surface climate data. ” Journal of Geophysical Research-Atmospheres 112, D 24 S 09, doi: 10. 1029/2007 JD 008465. l Mc. Kitrick, Ross R. and Nicolas Nierenberg (2010) “Socioeconomic Patterns in Climate Data. ” Journal of Economic and Social Measurement, 35(3, 4) pp. 149175. DOI 10. 3233/JEM-2010 -0336. l Mc. Kitrick, Ross R. (2010) “Atmospheric Oscillations do not Explain the Temperature-Industrialization Correlation. ” Statistics, Politics and Policy, Vol 1 No. 1, July 2010. l rossmckitrick. weebly. com ross. mckitrick. weebly. com 5

Core Methodology l There is a spatial pattern of warming and cooling trends since 1980 l Climate models predict the pattern as a response to GHG’s, solar changes, etc. l The predicted pattern is uncorrelated with spatial pattern of socioeconomic development l But raw weather data is known to be correlated with socioeconomic development l The adjustment models are supposed to remove these effects. l Therefore: If the adjustments are adequate, the climate data should be uncorrelated with socioeconomic patterns ross. mckitrick. weebly. com 6

Core Methodology l There is a spatial pattern of warming and cooling trends since 1980 l Climate models predict the pattern as a response to GHG’s, solar changes, etc. l The predicted pattern is uncorrelated with spatial pattern of socioeconomic development l But raw weather data is known to be correlated with socioeconomic development l The adjustment models are supposed to remove these effects. l Therefore: If the adjustments are adequate, the climate data should be uncorrelated with socioeconomic patterns ross. mckitrick. weebly. com 7

Core Methodology l There is a spatial pattern of warming and cooling trends since 1980 l Climate models predict the pattern as a response to GHG’s, solar changes, etc. l The predicted pattern is uncorrelated with spatial pattern of socioeconomic development l But raw weather data is known to be correlated with socioeconomic development l The adjustment models are supposed to remove these effects. l Therefore: If the adjustments are adequate, the climate data should be uncorrelated with socioeconomic patterns ross. mckitrick. weebly. com 8

Core Methodology l There is a spatial pattern of warming and cooling trends since 1980 l Climate models predict the pattern as a response to GHG’s, solar changes, etc. l The predicted pattern is uncorrelated with spatial pattern of socioeconomic development l But raw weather data is known to be correlated with socioeconomic development l The adjustment models are supposed to remove these effects. l Therefore: If the adjustments are adequate, the climate data should be uncorrelated with socioeconomic patterns ross. mckitrick. weebly. com 9

Core Methodology l Hypothesis: l l {spatial pattern of trends in surface climate data} is uncorrelated with {spatial pattern of socioeconomic development} In a series of papers I have shown that this hypothesis is strongly rejected ross. mckitrick. weebly. com 10

Sources of climate data l CRU, NOAA, NASA all produce “global climate data” products l All rely on same underlying archive l l Global Historical Climatology Network (run by NOAA) The 3 data products are very similar since they all use the same input data and similar, though not identical, averaging methods ross. mckitrick. weebly. com 11

Sources of observational error: l l l l Changing sample size Changing sample locations Build up of surrounding landscape Equipment changes Poor quality control Local air pollution Waste heat from buildings and traffic, etc. ross. mckitrick. weebly. com 12

GHCN sample 1885 l Locations of weather stations ross. mckitrick. weebly. com 13

GHCN sample 1925 l Locations of weather stations ross. mckitrick. weebly. com 14

GHCN sample 1945 ross. mckitrick. weebly. com 15

GHCN sample 1965 l Locations of weather stations ross. mckitrick. weebly. com 16

GHCN sample 1985 l Locations of weather stations ross. mckitrick. weebly. com 17

GHCN sample 2005 l Locations of weather stations ross. mckitrick. weebly. com 18

GHCN sample size over time ross. mckitrick. weebly. com 19

GHCN fraction of sample from urban airports ross. mckitrick. weebly. com 20

“Climate” data: the record as if the land surface was never modified and equipment never varied Temp data from cities adjustment algorithm + “True” record = ross. mckitrick. weebly. com 21

Structure of data set l l l Cross-sectional Observational unit is a 5 ox 5 o grid cell Dependent variable is 1979 -2002 trend ross. mckitrick. weebly. com 22

Measurement Model Where qi = observed climatic trend o. C/decade Ti = “true” trend f (Si) = surface processes like urbanization and agriculture g (Ii) = data inhomogeneities ross. mckitrick. weebly. com 23

For gridcell i l Ti (ideal temperature trend) represented by l TROPi = trend in troposphere over same gridcell as measured by satellites ross. mckitrick. weebly. com 24

For gridcell i l Surface processes f (Si) measured by pi = % growth in population density mi = % growth in real average income yi = % growth in real national GDP ci = % growth in national consumption ross. mckitrick. weebly. com 25

For gridcell i l Inhomogeneities g (Ii) measured by gi = GDP density (GDP per square km) ei = availability of educated workers (sum of literacy + postsecondary education) xi = rate of missing observations (# missing months in cell) ross. mckitrick. weebly. com 26

Regression equation Surface proc. l Inhom. GLS with clustering-robust std error matrix ross. mckitrick. weebly. com 27

First pair of studies: l Mc. Kitrick and Michaels (2004) l Tested 218 raw series and corresponding CRU gridded data l Both exhibited significant imprint of socioeconomic data with v. similar coefficients l ‘Adjustment’ hypothesis rejected at high confidence level l Mc. Kitrick and Michaels (2007) l Complete sample of (available) surface grid cells l ‘Independence’ hypothesis again rejected at high confidence level l Both studies: nonclimatic signals likely add up to a net warming bias in global average ross. mckitrick. weebly. com 28

2007 Results Probability that effects are zero: l Joint P = 0. 0000 (7 x 10 -14) ross. mckitrick. weebly. com 29

Specification tests l l l Bootstrap resampling Remove outliers, re-estimate RESET test Cross-validation tests Hausman endogeneity test (P = 0. 9962) ross. mckitrick. weebly. com 30

Generating ‘clean’ trends l l l Set GDP density and education to US levels Set all other surface and inhomogeneity effects to 0 Use model coeff’s to generate adjusted predicted values Observed average surface trend: MSU average: 0. 23 Adjusted average surface trend: 0. 17 0. 30 o. C/decade ross. mckitrick. weebly. com 31

IPCC Report l How did the IPCC deal with this? l IPCC AR 4 page 244: l l Mc. Kitrick and Michaels (2004) and De Laat and Maurellis (2006) attempted to demonstrate that geographical patterns of warming trends over land are strongly correlated with geographical patterns of industrial and socioeconomic development, implying that urbanisation and related land surface changes have caused much of the observed warming. However, the locations of greatest socioeconomic development are also those that have been most warmed by atmospheric circulation changes (Sections 3. 2. 2. 7 and 3. 6. 4), which exhibit large-scale coherence. Hence, the correlation of warming with industrial and socioeconomic development ceases to be statistically significant. No supporting citation given ross. mckitrick. weebly. com 32

IPCC Report l I obtained correlation fields between gridded temperatures and AO, ENSO and PDO ross. mckitrick. weebly. com 33

IPCC Report I augmented data sets for M&M 2004 and M&M 2007 with circulation terms l 2004 Model: l l l Circulation index effects are insignificant Including them anyway does not remove the significance of the conclusions 2007 Model l l Circulation index effects are jointly barely significant Including them increases size and significance of socioecononomic terms Conclusion: IPCC claim is false. (Mc. Kitrick 2010, Statistics Politics and Policy July 2010) ross. mckitrick. weebly. com 34

IPCC Report I augmented data sets for M&M 2004 and M&M 2007 with circulation terms l 2004 Model: l l l Circulation index effects are insignificant Including them anyway does not remove the significance of the conclusions 2007 Model l l Circulation index effects are jointly barely significant Including them increases size and significance of socioecononomic terms Conclusion: IPCC claim is false. (Mc. Kitrick 2010, Statistics Politics and Policy July 2010) ross. mckitrick. weebly. com 35

IPCC Report I augmented data sets for M&M 2004 and M&M 2007 with circulation terms l 2004 Model: l l l Circulation index effects are insignificant Including them anyway does not remove the significance of the conclusions 2007 Model l l Circulation index effects are jointly barely significant Including them increases size and significance of socioecononomic terms Conclusion: IPCC claim is false. (Mc. Kitrick 2010, Statistics Politics and Policy July 2010) ross. mckitrick. weebly. com 36

IPCC Report I augmented data sets for M&M 2004 and M&M 2007 with circulation terms l 2004 Model: l l l Circulation index effects are insignificant Including them anyway does not remove the significance of the conclusions 2007 Model l l Circulation index effects are jointly barely significant Including them increases size and significance of socioecononomic terms Conclusion: IPCC claim is false. (Mc. Kitrick 2010, Statistics Politics and Policy July 2010) ross. mckitrick. weebly. com 37

Schmidt (2009) “Spurious correlation between recent warming and indices of local economic activity. ” International Journal of Climatology 10. 1002/joc. 1831 l 3 arguments against our findings l surface temperature field exhibits spatial autocorrelation (SAC) so results are insignificant l Use of RSS satellite series rather than UAH series removes significance of results l Data generated by climate model yields apparent correlations with socioeconomic data, yet is uncontaminated by construction, so effects must be a fluke ross. mckitrick. weebly. com 38

Schmidt (2009) “Spurious correlation between recent warming and indices of local economic activity. ” International Journal of Climatology 10. 1002/joc. 1831 l 3 arguments against our findings l surface temperature field exhibits spatial autocorrelation (SAC) so results are insignificant l Use of RSS satellite series rather than UAH series removes significance of results l Data generated by climate model yields apparent correlations with socioeconomic data, yet is uncontaminated by construction, so effects must be a fluke ross. mckitrick. weebly. com 39

Schmidt (2009) “Spurious correlation between recent warming and indices of local economic activity. ” International Journal of Climatology 10. 1002/joc. 1831 l 3 arguments against our findings l surface temperature field exhibits spatial autocorrelation (SAC) so results are insignificant l Use of RSS satellite series rather than UAH series removes significance of results l Data generated by climate model looks correlated with socioeconomic data, yet is uncontaminated by construction, so effects must be a fluke ross. mckitrick. weebly. com 40

Mc. Kitrick & Nierenberg “Socioeconomic patterns in climate data” J Econ Soc Measurement 2010 l Responses l Schmidt did not actually test SAC. We do, and show that while depvar is AC’d, regression residuals are not, as long as socioecon variables are included in model. l Use of RSS data diminishes individual significance but effect due to a small number of outliers. Once these removed, RSS yields strongest results of all data sets l Model-based data cannot replicate observed patterns; predicts opposite signs ross. mckitrick. weebly. com 41

Mc. Kitrick & Nierenberg “Socioeconomic patterns in climate data” J Econ Soc Measurement 2010 l Responses l Schmidt did not actually test SAC. We do, and show that while depvar is AC’d, regression residuals are not, as long as socioecon variables are included in model. l Use of RSS data diminishes individual significance but effect due to a small number of outliers. Once these removed, RSS yields strongest results of all data sets l Model-based data cannot replicate observed patterns; predicts opposite signs ross. mckitrick. weebly. com 42

Mc. Kitrick & Nierenberg “Socioeconomic patterns in climate data” J Econ Soc Measurement 2010 l Responses l Schmidt did not actually test SAC. We do, and show that while depvar is AC’d, regression residuals are not, as long as socioecon variables are included in model. l Use of RSS data diminishes individual significance but effect due to a small number of outliers. Once these removed, RSS yields strongest results of all data sets l Model-based data cannot replicate observed patterns; predicts opposite signs ross. mckitrick. weebly. com 43

Data variations l Surface l l l Observed: CRU, CRU 2 v, CRU 3 v Modeled: GISS-E; GCM average Troposphere l l Observed: UAH, RSS Modeled: GISS-E; GCM average ross. mckitrick. weebly. com 44

Spatial Autocorrelation Tests OBSERVED: SAC DISAPPEARS MODELS: SAC REMAINS ross. mckitrick. weebly. com 45

Estimation with SAC model ross. mckitrick. weebly. com 46

Estimation with SAC model OBSERVATIONS: MODELS: SIGNIFICANT INSIGNIFICANT ross. mckitrick. weebly. com 47

GCM Counterfactual l Schmidt 2009, p. 2: There is a relatively easy way to assess whethere is any true significance to these correlations. We can take fully consistent model simulations for the same period and calculate the distribution of the analogous correlations. Those simulations contain no unaccounted-for processes (by definition!) but plenty of internal variability, locally important forcings and spatial correlation. If the distribution encompasses the observed correlations, then the null hypothesis (that there is no contamination) cannot be rejected. ross. mckitrick. weebly. com 48

Results ross. mckitrick. weebly. com 49

Results 1 = climate model reproduces observed effect, 0 = failure to do so ross. mckitrick. weebly. com 50

Filtering results on surface data l l l Set GDP density and education to US levels Set all other surface and inhomogeneity effects to 0 Use model coeff’s to generate adjusted predicted values ross. mckitrick. weebly. com 51

Filtering results on surface data l l l Set GDP density and education to US levels Set all other surface and inhomogeneity effects to 0 Use model coeff’s to generate adjusted predicted values ross. mckitrick. weebly. com 52

Filtering results on surface data l l l Set GDP density and education to US levels Set all other surface and inhomogeneity effects to 0 Use model coeff’s to generate adjusted predicted values This method should not reduce mean trend in GISS data ross. mckitrick. weebly. com 53

Conclusions l In general, I reject the null hypothesis that adjustment models yield “climate” data l l l socioeconomic patterns are highly significant across wide variety of specifications and data combinations socioeconomic data are necessary for well-specified error term This suggests a causal interpretation of the regression results ross. mckitrick. weebly. com 54

Responses to critiques: l IPCC claim that the results were statistically insignificant & due to natural circulation patterns was a fabrication l l The claim was both unsubstantiated and untrue Various critiques have not held up l l l SAC is not a source of bias Results hold up across numerous data sets Climate models cannot reproduce results ross. mckitrick. weebly. com 55

Thank you l ross. mckitrick@uoguelph. ca l rossmckitrick. weebly. com ross. mckitrick. weebly. com 56