Combining the Sources of Temperature Change

By Gerard Gilliland

Introduction:

This paper ties natural and manmade events and processes to temperature change. This is a curve relationship study. I am using existing data: the numbers gathered by accountants, and scientists that work in diverse fields. I combine these diverse data to 'reflect' the shape of the temperature anomaly curve. With time and additional data, I hope they will be shown to 'drive' the temperature curve.

I chose the following for analysis:

I compare the direction of Temperature change with the above Combination for each month from January 1880 through December 2016 with a matching Result of 0.6650.

Process:

To compare the Combined curve (sum of the components) with the Temperature curve, I use the following process: The comparison is done on a monthly basis. I look at each month of this Combined number and set a combined flag (CombFlag) equal to 1 if Combined increased from the previous month, or -1 if it decreased from the previous month. I am looking at direction only not magnitude. I make the same test with temperature (TempFlag). Then I compare the two flags (CombFlag and TempFlag) for each month. If they match, I set the MatchFlag for that month to 1, else I set it to 0. I do this for each month for the entire period. Then I average all the MatchFlags for the entire period. The Result would be 1.0 if every Combined period and Temperature period matched. It would be zero if none matched, and 0.5 if half of the match flags were the same. To build the Combined curve, I use the following process: Some of the data is monthly. I convert all the five year and one year data to monthly data using a straight line between known data points. Some of the early data is US only and most of the later data is Global. When converting, I scale the US data up to Global. Each data curve is normalized, averaged, multiplied, and then summed in the Combined curve.

To keep track of the process, I use a Multiply table and a Average table for each curve. In addition to the MatchFlag, I also use an accumulated MatchFlag. If the CombFlag and TempFlag match, I set the accumulated MatchFlag for that month to 1 plus the previous month's value, else I set it to 0. The Average best fit (highest sum of the accumulated MatchFlags) are stored in each row of the Multiply table.

An example will clarify the process. The volcanoes main source of Global Cooling is stratospheric sulfate aerosols. The Sato Index (Figure1) represents the stratospheric aerosol optical thickness. The Y axis is measuring aerosols at the 550 nm range which is about in the middle of visible light. The white or non-absorbing aerosols in the stratosphere reflect the sunlight back to space. Thus an increase in the Sato Index reflects a decrease in heating by the sun. Choosing one volcano for this example, the magmatic phase of the eruption of Mount Pinatubo began June 3, 1991. The peak Sato Index was recorded in February 1992.

Figure1: Sato Index
Missing Sato Index graph

Figure2: Process Example
Missing Process Example graph

An example of the calculation process for the SatoIndex can be seen in Figure2. For each month, the process collects the previous 56 months of data, Inverts it (because the Sato Index is associated with cooling), Normalizes, Averages, Multiplies, and Adds it to the Combined curve. You may view a GIF animation of the above graph. The animation is just the multiply part of the process and adding that result to the Combined curve.

Figure3: SatoIndex Average Curve is a view of the Average table. Recognize that Figure3 was captured when the Average was 56 months and the Multiply table was 0.359375. The Average table and associated curves get repopulated with each Multiply value and repopulated a final time with the best Result.

Figure3: SatoIndex Average Curve
Missing Sato Index Average Curve
You may view a GIF animation of the above graph. The animation shows the best fit for the multiply values from the following table.

Table1: SatoIndex Multiply Table

MultiplyResultLocationMax
    
1.00010603P48M0 
0.50011103P55M0 
0.75010342P53M0 
0.2511067P56M0 
0.37511352P56M0 
0.12511004P56M0 
0.312511180P56M0 
0.437511213P55M0 
0.3437511141P56M0 
0.35937511391P56M0P56M0
    

Figure4: Combined Manmade and Natural Components of Temperature
Missing Combined Components graph

Figure4 displays all the components. You may view a GIF animation of the above graph.

The Component Details are listed in Table2. Component is the name of the curve. The Temperature Anomaly is from the mean of January 1901 through December 2000. I use that same period to find the Maximum for Normalization of the other curves. Multiply is the fraction of the normalized curve. The larger the fraction the more the component affects the Combined curve. Months Ave is Months the component is averaged.

Table2: Component Details

ComponentMultiplyMonths AveComments
Temperaturen/a50Monthly from 1880: http://www.ncdc.noaa.gov/cag/time-series/global Temperature Monthly from 1880; select Globe; All Months; January; click Plot; Download
What I am currently trying to match. (And ultimately what I'm saying is being driven by these other curves.)
Sato Index0.35937556Inverted Monthly data through September 2012: http://data.giss.nasa.gov/modelforce/strataer/ Select ASCII Data
Oil0.9687535Five year US data scaled to global from 1880 to 1945: http://www.eia.doe.gov/emeu/aer/txt/ptb1701.html (Fossil Fuels / Petroleum column).
Yearly US data scaled to Global for 1949 through 1969: http://www.eia.doe.gov/emeu/aer/txt/stb0103.xls (Fossil Fuels / Petroleum column).
Yearly Global data from 1970: http://www.bp.com/statisticalreview Select download the workbook (Oil Consumption – tonnes tab).
Coal0.937515Inverted Five year US data scaled to global from 1880 to 1945: http://www.eia.doe.gov/emeu/aer/txt/ptb1701.html (Fossil Fuels / Coal column).
Inverted Yearly US data scaled to Global for 1949 through 1969: http://www.eia.doe.gov/emeu/aer/txt/stb0103.xls (Fossil Fuels / Coal column).
Inverted Yearly Global data from 1970: http://www.bp.com/statisticalreview Select download the workbook (Coal - Consumption Mtoe tab).
Gas0.7529Five year US data scaled to global from 1885 to 1945: http://www.eia.doe.gov/emeu/aer/txt/ptb1701.html (Fossil Fuels / Natural Gas column).
Yearly US data scaled to Global for 1949 through 1969: http://www.eia.doe.gov/emeu/aer/txt/stb0103.xls (Fossil Fuels / Natural Gas column).
Yearly Global data from 1970: http://www.bp.com/statisticalreview Select download the workbook (Gas Consumption - tonnes tab)
Wood (Biomass)0.04687553Inverted Five year US data from 1880 to 1945: http://www.eia.doe.gov/emeu/aer/txt/ptb1701.html (Renewable Energy / Wood column). Fuel wood only.
Inverted Yearly US data for 1949 through 2009: http://www.eia.doe.gov/emeu/aer/txt/stb0103.xls (Renewable Energy / Biomass column). Wood and derived wood fuels.
Events0.62555June 30, 1908 Tunguska, Russia meteoroid: http://en.wikipedia.org/wiki/Tunguska_event I didn't have any data so I used the 1883 Krakatoa eruption Sato index.

Figure5: Detail
Missing Combined Detail graph

Figure5 is using the same curves, averages, and offsets as listed in Table2. Note: I added a curve of CO2 from the bp spread sheet - Carbon Dioxide Emissions tab. This is million metric tonnes per year normalized and scaled by 0.5. This is not part of the calculations. It is only here for comparison with the combined curve and the temperature curve. Read into it what you want. You may view a GIF animation of the above graph.

Comments:

The conversion to monthly data using a straight line between known data points is adversely affecting the averaging and matching processes. The curves with yearly data are averaged for much shorter periods. They are already quite stiff. I think I will have more success looking for recent monthly data to replace the yearly data I am using instead of searching for older data that I believe does not exist.

I had not been able to find a reason to explain the cooler temperatures in the first half of the study. I read an article on Coal (http://americanradioworks.publicradio.org/features/coal/) where people had to turn on their lights during the day and the trees had all died because the pollution was so bad. So I inverted the coal curve and that improved the analysis. Could coal (at least before scrubbers and precipitators) be so polluting that haze, and acid rain from SO2 cause more cooling than CO2 causes warming? (I’m not recommending the solution to Global Warming is acid rain and recessions.) In 2011, another paper came to this same conclusion. I have heard that the Natural Gas production increase in the US through fracking (hydraulic fracturing) has reduced electricity produced by Coal to the 1940's level. If this trend continues world-wide we should expect the largest increase in temperature we have seen to date.

Other components have been included at different times in the past but have not survived when included in combination curve analysis:

Sunspots. Monthly data: http://solarscience.msfc.nasa.gov/greenwch/spot_num.txt however Wikipedia showed the Sun's non-attributability and removing Sunspots improved the accumulated MatchFlag

SO2. Inverted yearly data 1850 to 2011: http://www.atmos-chem-phys.net/11/1101/2011/acp-11-1101-2011.html
Inverted Yearly data 1971 through 2008: http://edgar.jrc.ec.europa.eu/ does definately affect the temperature. But is it double counting the SO2 in coal? When SO2 is added, the coal line drops back. And it is hard to tie SO2 to the economy other than through coal. I think the reduction of SO2 in the 1990’s was a major part of the temperature increase during that period. (Sometimes it's hard to recognize progress.)

Industrial Production. US Monthly data scaled to global from 1919 to 1970: http://research.stlouisfed.org/fred2/series/INDPRO?cid=3
Yearly data from 1970 to 1990: http://search.worldbank.org/data?qterm=industrial%20production&language=EN&format=html Select "Industry, value added (constant 2000 US$)
Monthly data from 1990: http://www.cpb.nl/en/number/cpb-world-trade-monitor-including-april-2013 Download Spreadsheet and Select the InPro tab.

I attempted to graph Recessions (GDP Decline in Percent) but the Results were mixed. I failed to graph War. I was looking for Tons of TNT per month. It seems most war histories are heavy on strategy, heroics, and tragedies, and light on logistics and economics. I was overwhelmed with the volume of literature listed on http://www.airforcehistory.hq.af.mil/Publications/titleindex.htm, but they responded they "... could find no overall statistics of munitions expended covering friend and foe for wars and conflicts." I now view War and the Economy as drivers of Source Energy and Industrial Production. However, the largest discrepancies between the temperature and combined curves occur during World War Two and during the Vietnam war. Was coal gasification, or agent orange, or other processes not modeled part of the discrepancy?

I have been updating this process since 2009 (with data through the previous year) and it continues to track. I am certainly open to additional processes or events. The paper What's Really Warming the World? written in June 24, 2015 uses the same concept and philosophy I have been using. However they use "green house gasses" rather than consumption of the products producing those gasses.

Summary:

I want to stress this is a curve relationship study: A graphic association. I think the same process could be used to show relationships in other fields. Here the purpose is to show a relationship exists between nature, mankind, the economy (that drives the production lines in good times and unemployment lines in bad times), and global temperature. If you can see that relationship you might agree that there is something to the saying: The economy is hot, or the economy is cooling off.

My thanks to American Radio Works, BP, CPB (Centraal Planbureau), EDGAR, EIA, EPA, Federal Reserve Bank of St. Louis, Google, NOAA, PNNL, Wikipedia, World Bank, and World Resources Institute.

My thanks to you for reading it.

Excel Spreadsheet with the calculations.


mailbox
27Oct2017 Creative Commons License
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.
Email: