[LIST OF POSTS ON THIS SITE]

CLIMATE CHANGE IN AUSTRALIA AS SEEN IN DAILY STATION DATA

A month by month trend analysis of more than 100 years of daily maximum (TMAX) and daily minimum (TMIN) temperatures from three weather stations in Australia is presented. The results show warming trends in TMIN for all twelve calendar months at all three stations with observed warming rates ranging from 0.4 to 2.3 degrees Celsius per century. The TMAX data show a combination of warming trends, cooling trends, and no trends with significant differences among stations and among the calendar months so that no coherent conclusion can be drawn with respect to the long term trend in TMAX. Temperature trends in a moving 30-year window indicate that long term linear OLS trends in temperature are the residual product of violent multi-decadal cycles of warming and cooling at rates that are an order of magnitude greater. Detrended correlation analysis failed to establish a relationship between emissions and warming. The strong evidence of warming found in the TMIN data is confounded by its absence in TMAX as no theoretical basis exists for fossil fuel emissions to cause warming in TMIN and not in TMAX.

RELATED POST ON STATION DATA:  [LINK]

1. ACRONYMS: C=degrees Celsius, OLS=ordinary least squares, WMO=World Meteorological Organization, AGW=anthropogenic global warming, TMIN=nighttime daily minimum temperature, TMAX=daytime daily maximum temperature, STDEV=standard deviation,
2. SUMMARY: Over a hundred years of daily minimum (TMIN) and daily maximum (TMAX) temperature measurements at three weather stations in Australia (90015, 26026, & 30045) taken by the Bureau of Meteorology (BOM) are studied separately on a month by month basis for long term trends. The procedure is designed to maintain data integrity in terms of the diurnal and seasonal cycles that involve temperature changes orders of magnitude greater than long term trends. The results for TMAX are inconclusive with one station showing cooling and the other two showing mixed results with warming trends for some months and no evidence of trends for other months. We conclude that the TMAX data do not provide convincing evidence of long term warming trends. Dramatically different results are found for TMIN. All three stations show strong and statistically significant warming trends in TMIN. However, detrended correlation analysis could not relate these warming trends to fossil fuel emissions. These results are anomalous. No theoretical framework exists for anthropogenic global warming acting through the greenhouse effect of atmospheric CO2 by way of fossil fuel emissions to cause warming in nighttime temperatures without affecting the maximum daytime temperature. It is proposed that anomalies of this kind may be used in conjunction with other factors in the evaluation of the integrity of the instrumental record of surface temperature.
3. DETAILS BELOW:
4. Most of the variance in temperature is contained in the diurnal cycle, the seasonal cycle, in multi-decadal trend cycles within the full span of the data, and in random variability. A very small portion, usually in the order of 5% or less, can be attributed to long term trends (Munshi J. , 2015). For example, in Cape Otway, Australia, located at latitude of -38.86, the diurnal range is ≈7C on average and the range in the seasonal cycle is ≈8C on average. Also trends in a 30-year moving window show multi-decadal trends in the range of -8C to +8C per century. By comparison, the conventional aggregated long term OLS analysis of daily minimum temperatures shows full span linear OLS warming rates 1865-2016 of less than≈1C per century, a rather insignificant number in comparison. For instance, the annual seasonal cycle of ≈8C is equivalent to more than 800 years of long term OLS trend in the full span of the data. Yet, Information about diurnal and seasonal cycles is lost when the data are aggregated prior to trend analysis. To maintain data integrity we propose a procedure in which trend analysis is carried out separately for extreme daytime and nighttime temperatures (daily maximum and daily minimum) and for each calendar month as a way of preserving the integrity of variable diurnal and seasonal cycles throughout the sample period. The procedure yields 24 different trend values. These trends are then understood and interpreted in their context and combined as necessary and as appropriate to answer specific research questions.
5. A linear OLS trend line across the full span of the time series yields an estimate of the overall trend across the entire sample period but assumes linearity. Long term non-linearity in the data can be detected by drawing a quadratic or cubic curve across the full span and comparing R-squared values. Non-linearity is inferred if the value of R-squared for the non-linear OLS curve is much higher than that of the linear OLS line. Conversely, the linearity assumption is supported if the difference is not significant or a higher value of R-squared is seen in the linear OLS line. Greater insight into the nonlinear trend cycles within the full span of the data is achieved by constructing what may be termed a “trend profile”. It is designed to discover multi-decadal trend cycles that are known to occur in long temperature series (Parker, 2007). A generational (30-year) moving window is used for this purpose. The significance of the 30-year time span derives from the WMO position that climate is 30 years of weather. It moves through the data series one year at a time computing generational trends. A plot of generational trends against time reveals the multi-decadal cycles if they exist in the data. Greater information about non-linearity in the data is gained.
6. An important consideration in the study of temperature trends in the context of AGW is whether these trends may be related to the rate of fossil fuel emissions. Empirical evidence in support of the AGW hypothesis that fossil fuel emissions cause a measurable warming trend in surface temperature has been presented as a correlation between cumulative emissions and cumulative warming in various global and regional homogenized temperature series as well as in station data (Allen, 2009) (IPCC, 2000) (Matthews, 2009) (Gillett, 2013) (Zickfeld, 2009) (Solomon, 2009) (Davis, 2010) (Meinshausen, 2009) (Karoly, 2006) (IPCC, 2007). However, it is shown in a related post that correlations between cumulative emissions and cumulative warming are spurious because a time series of cumulative values of observations in another time series contains neither time scale nor degrees of freedom [LINK] . This finding implies that only correlations at finite time scales between emissions and warming can serve as empirical evidence for AGW. Katherine Ricke and Ken Caldeira had proposed that the optimal time scale is a decade based on the response characteristics of warming to emissions in the CMIP5 climate model (Ricke, 2014). However, it was not possible to replicate this result outside of climate models and it appears that longer time scales yield better correlations. For example, correlations at a generational time scale (30 years) are higher than those at a decadal time scale. As well, a generation (30 years) time scale is recognized by the WMO as the appropriate time scale in the study of climate (WMO, 2016) (Ackerman, 2006). The generational time scale within a moving 30-year span is therefore used in this work. (A comparison of time scales is presented in a related post on this site [LINK] .
7. Daily maximum (TMAX) and daily minimum (TMIN) temperatures are provided by the Government of Australia Bureau of Meteorology (BOM) for a large number of weather stations located throughout Australia (BOM, 2017). Five of these weather stations are listed below. They were selected for the study based on data availability of data for more than a century. Large gaps of a decade or more without data were found in the Boulia Airport and Hobart datasets and these stations were removed from the study. In the remaining three stations, all data from incomplete calendar years at the beginning and end of the time series were removed. The remaining sample period for full calendar years are 1865-2016 for Station-90015, 1885-2016 for Station-26026, and 1893-2016 for Station-30045. For each station the daily maximum temperature and daily minimum temperature are reported for each day of the year. Data are missing for 3% to 5% of the days. Missing data were replaced with the most recent data available – typically separated by one to four days. The error introduced by this procedure is assumed to be negligible. Measuring station details are shown in the table below for the five stations considered in the study. Three of these stations contained a sufficient span of data with low missing data counts and the data from these stations were selected for study. The three selected stations are: 90015-Cape-Otway, 26026-Robe, and 30045-Richmond. TMIN and TMAX and the twelve calendar months are studied separately as different phenomena of nature and not combined. This procedure requires twenty four separate sets of trend analysis for each weather station. The procedure maintains the integrity of both the diurnal cycle and the seasonal cycle. 
8. DESCRIPTIVE STATISTICS ALL THREE STATIONS 
9. TREND ANALYSIS FOR 90015 CAPE OTWAY: TMAX AND TMIN
10. TREND ANALYSIS FOR 26026 ROBE: TMAX AND TMIN 
11. TREND ANALYSIS FOR 30045 RICHMOND: TMAX AND TMIN 
12. SUMMARY OF RESULTS: TMAX AND TMIN: ALL THREE STATIONS 

CONCLUSION

Over a hundred years of daily minimum (TMIN) and daily maximum (TMAX) temperature measurements at three weather stations in Australia (90015, 26026, & 30045) taken by the Bureau of Meteorology (BOM) are studied separately on a month by month basis for long term trends. The procedure is designed to maintain data integrity in terms of the diurnal and seasonal cycles that involve temperature changes orders of magnitude greater than long term trends. The results for TMAX are inconclusive with one station showing cooling and the other two showing mixed results with warming trends for some months and no evidence of trends for other months. We conclude that the TMAX data do not provide convincing evidence of long term warming trends. Dramatically different results are found for TMIN. All three stations show strong and statistically significant warming trends in TMIN. However, detrended correlation analysis could not relate these warming trends to fossil fuel emissions. These results are anomalous. No theoretical framework exists for anthropogenic global warming acting through the greenhouse effect of atmospheric CO2 by way of fossil fuel emissions to cause warming in nighttime temperatures without affecting the maximum daytime temperature. It is proposed that anomalies of this kind may be used in conjunction with other factors in the evaluation of the integrity of the instrumental record of surface temperature.

[SOURCE DOCUMENT DOWNLOAD LINK]

REFERENCES

1. Ackerman, S. (2006). Meteorology: Understanding the Atmosphere. Jones and Barlett Titles in Physical Science.
2. Allen, M. (2009). Warming caused by cumulative carbon emissions towards the trillionth tonne. Nature , 458.7242 (2009): 1163-1166.
3. BOM. (2017). Climate Data. Retrieved 2017, from Bureau of Meteorology, Government of Australia: http://reg.bom.gov.au/climate/data/
4. Bowley, A. (1928). The standard deviation of the correlation coefficient. Journal of the American Statistical Association , 23.161 (1928): 31-34.
5. Box, G. (1994). Time series analysis: forecasting and control. Englewood Cliffs, NJ: Prentice Hall.
6. Callendar, G. (1938). The artificial production of carbon dioxide and its influence on temperature. Quarterly Journal of the Royal Meteorological Society , 64.275 (1938): 223-240.
7. CarbonBrief. (2016). What global emissions in 2016 mean for climate change goals. Retrieved 2017, from Carbon Brief: https://www.carbonbrief.org/what-global-co2-emissions-2016-mean-climate-change
8. CDIAC. (2014). Global Fossil-Fuel CO2 Emissions. Retrieved 2017, from CDIAC / ORNL: http://cdiac.ornl.gov/trends/emis/tre_glob_2013.html
9. Davis, S. (2010). Future CO2 emissions and climate change from existing energy infrastructure. Science , 329.5997 (2010): 1330-1333.
10. Draper&Smith. (1998). Applied Regression Analysis. Wiley.
11. Easterling, D. (2009). Is the climate warming or cooling? Geophysical Research Letters , 36.8 (2009).
12. Gillett, N. (2013). Constraining the ratio of global warming to cumulative CO2 emissions using CMIP5 simulations. Journal of Climate , 26.18 (2013): 6844-6858.
13. Hansen, J. (1988). 1988 Hansen Senate Testimony. Retrieved 2016, from Procon.org: http://climatechange.procon.org/sourcefiles/1988_Hansen_Senate_Testimony.pdf
14. Hansen, J. (2016). Ice melt, sea level rise and superstorms: evidence from paleoclimate data, climate modeling, and modern observations that 2 C global warming could be dangerous. Atmospheric Chemistry and Physics , 16.6 (2016): 3761-3812.
15. Hansen, J. (1981). Impact of Increasing Atmospheric Carbon Dioxide. Science , 213: 957-66.
16. Holm, S. (1979). A simple sequentially rejective multiple test procedure. Scandinavian Journal of Statistics , 6:2:65-70.
17. Hu, K. (2001). Effect of trends on detrended fluctuation analysis. Physical Review , E 64.1 (2001): 011114.
18. IPCC. (2007). AR4 WG1 Chapter 7: Couplings between changes in the climate system and biogeochemistry. Geneva: IPCC.
19. IPCC. (2014). Climate Change 2013 The Physical Science Basis. Geneva: IPCC/UNEP.
20. IPCC. (2000). Special report on emissions scenarios (SRES), a special report of Working Group III. Cambridge: Cambridge University Press.
21. Johnson, V. (2013). Revised standards for statistical evidence. Retrieved 2015, from Proceedings of the National Academy of Sciences: http://www.pnas.org/content/110/48/19313.full
22. Jones, P. (2001). The evolution of climate over the last millennium. Science , 292.5517 (2001): 662-667.
23. Kantelhardt, J. (2001). Detecting long-range correlations with detrended fluctuation analysis. Physica A: Statistical Mechanics and its Applications , 295.3 (2001): 441-454.
24. Karl, T. (2015). Possible artifacts of data biases in the recent global surface warming hiatus. Science , 348.6242 (2015): 1469-1472.
25. Karoly, D. (2006). Anthropogenic warming of central England temperature. Atmospheric Science Letters , 7.4 (2006): 81-85.
26. Lacis, A. (2010). Atmospheric CO2: Principal control knob governing Earth’s temperature. Science , 330.6002 (2010): 356-359.
    Matthews, D. (2009). The proportionality of global warming to cumulative carbon emissions. Nature , 459.7248 (2009): 829-832.
27. Meinshausen, M. (2009). Greenhouse-gas emission targets for limiting global warming to 2 C. Nature , 458.7242 (2009): 1158-1162.
28. Munshi, J. (2015). A Robust Test for OLS Trends in Daily Temperature Data. SSRN , http://dx.doi.org/10.2139/ssrn.2631298.
29. Munshi, J. (2015). Decadal Fossil Fuel Emissions and Decadal Warming: A Note. SSRN Electronic Journal , http://dx.doi.org/10.2139/ssrn.2662870.
30. Munshi, J. (2016). Effective Sample Size of the Cumulative Values of a Time Series. SSRN Electronic Journal , http://dx.doi.org/10.2139/ssrn.2853163.
31. Munshi, J. (2016). Generational Fossil Fuel Emissions and Generational Warming. SSRN Electronic Journal , http://dx.doi.org/10.2139/ssrn.2845972.
32. Munshi, J. (2016). Illusory Statistical Power in Time Series Analysis. SSRN Electronic Journal , http://dx.doi.org/10.2139/ssrn.2878419.
33. Munshi, J. (2017). OLS Trend Analysis of CET Daily Mean Temperatures 1772-2016 . SSRN Electronic Journal , https://ssrn.com/abstract=2951507.
34. Munshi, J. (2017). Oz Temperature Trends Archive. Retrieved 2017, from Google Drive: https://drive.google.com/open?id=0ByzA6UNa41ZfTng1RlgyWmF4YkU
35. Munshi, J. (2016). Some Methodological Issues in Climate Science. SSRN Electronic Journal , http://dx.doi.org/10.2139/ssrn.2873672.
36. Munshi, J. (2016). Spurious Correlations in Time Series Data. SSRN Electronic Journal , http://dx.doi.org/10.2139/ssrn.2827927.
37. Munshi, J. (2016). The Spuriousness of Correlations between Cumulative Values. SSRN Electronic Journal , http://dx.doi.org/10.2139/ssrn.2725743.
38. Netherlands Environmental Assessment Agency. (2016). Trends in global CO2 emissions. Retrieved 2017, from European Commission Joint Research Centre: http://edgar.jrc.ec.europa.eu/news_docs/jrc-2016-trends-in-global-co2-emissions-2016-report-103425.pdf
39. Parker, D. (2007). Decadal to multidecadal variability and the climate change background. Journal of Geophysical Research: Atmospheres , 112.D18 (2007).
40. Parker, D. (2005). Uncertainties in central England temperature 1878–2003 and some improvements to the maximum and minimum series. International Journal of Climatology , 25.9 (2005): 1173-1188.
41. Plaut, G. (1995). Interannual and interdecadal variability in 335 years of central England temperatures. Science , 268.5211 (1995): 710.
42. Podobnik, B. (2008). Detrended cross-correlation analysis: a new method for analyzing two nonstationary time series. Physical review letters , 100.8 (2008): 084102.
43. Revelle, R. (1957). Carbon dioxide exchange between atmosphere and ocean and the question of an increase of atmospheric CO2 during the past decades. Tellus , 9.1 (1957): 18-27.
44. Ricke, K. (2014). Maximum warming occurs about one decade after a carbon dioxide emission. Environmental Research Letters , 9.12 (2014): 124002.
45. Siegfried, T. (2010). Odds Are, It’s Wrong. Retrieved 2016, from Science News: https://www.sciencenews.org/article/odds-are-its-wrong
46. Solomon, S. (2009). Irreversible climate change due to carbon dioxide emissions. Proceedings of the national academy of sciences , (2009): pnas-0812721106.
47. The Conversation. (2016). Fossil fuel emissions have stalled. Retrieved 2017, from The Conversation: https://theconversation.com/fossil-fuel-emissions-have-stalled-global-carbon-budget-2016-68568
48. Trenberth, K. (2013). An apparent hiatus in global warming? Earth’s Future , 1.1 (2013): 19-32.
49. Trenberth, K. (2014). Seasonal aspects of the recent pause in surface warming. Nature Climate Change , 4.10 (2014): 911-916.
50. Tung, K. (2013). Using data to attribute episodes of warming and cooling in instrumental records. Proceedings of the National Academy of Sciences , 110.6 (2013): 2058-2063.
51. WMO. (2016). WMO. Retrieved 2017, from Climate: https://public.wmo.int/en/our-mandate/climate
52. Zickfeld, K. (2009). Setting cumulative emissions targets to reduce the risk of dangerous climate change. Proceedings of the National Academy of Sciences , 106.38 (2009): 16129-16134.

[LIST OF POSTS ON THIS SITE]