The Earth System Research Laboratories, National Oceanic and Atmospheric Administration, collect atmospheric CO2 concentration data at the Barrow Atmospheric Baseline Observatory, latitude 71.323̊North, longitude 156.611̊West. The Observatory is at Utqiagvik (pronounced uut-kee-ah-vik) on the Northern-most point of Alaska, formally known as Barrow until 2016, at an elevation of 11 metres above sea level. The climate is classified as Arctic Tundra. Its extreme location means Utqiagvik receives 24-hour daylight between sunrise on May 12 (day 132) and sunset on August 2 (day 214) and is in darkness from November 19 (day 323) to January 22 (day 22) every year.

The data may be downloaded from their web site at :

https://www.esrl.noaa.gov/gmd/dv/data/

The two files of monthly CO2 concentration measurements:

co2_brw_surface-flask_1_ccgg_month.txt, 1 May 1971 to 1 December 2018,

and co2_brw_surface-insitu_1_ccgg_MonthlyData.txt, 1 July 1973 to 1 December 2019,

were combined to provide more complete coverage as each had missing data points and covered different periods. This was considered appropriate as a t-test of the difference between the common points of the two time series revealed that statistically the difference was not significantly different from zero.

Figure 1 displays the monthly CO2 concentration for the combined time series. There are two inflections in the graph which correspond to the time of the volcanic eruptions at Tolbachik on the Kamchatka Peninsular, 6 July 1975 to 10 December 1976, and Mount Pinatubo, Philippines, 12 June 1991.

Figure 1.

The Barrow Observatory is 2651 km from the Tolbachik volcano on a bearing of 32.8̊East and distance 8171 km from the Mount Pinatobu volcano on a bearing of 19.4̊East.

A linear trend fitted to the combined time series revealed the rate of increase for the CO2 concentration over the 50 year period May 1971 to December 2019 was 1.72 ppm per annum. For the five year period May 1971 to 1976, the average rate of increase was 1.6 ppm per annum and for December 2014 to 2019, the rate had steadily increased to 2.14 ppm per annum. This was similar to the statistics from the Mauna Loa Observatory as shown in the comparative graphs in Figure 2. Other than the amplitude of the seasonal variation, the time series are remarkably similar given that the Barrow Observatory is 5756 km North of the Mauna Loa Observatory.

Figure 2.

The estimated seasonal variation at Barrow ranged in amplitude from 13.1 ppm to 19.3 ppm with the range increasing over time, as has been seen in other CO2 time series. Cross correlation between the seasonally adjusted CO2 time series for the two Observatories showed that the Barrow series lagged that for Mauna Loa by almost 3 months.

The relationship between the monthly CO2 concentration and the satellite lower troposphere temperature for the North Pole zone ( Ref.[1] ) is shown in Figure 3 for the period December 1978 to December 2019. The temperature range is from -1.98̊C to 2.47̊C relative to a 30 year average base. The range for the monthly CO2 concentration is from 317.71 ppm to 417.81 ppm. The correlation coefficient for the CO2 concentration relative to the satellite Arctic zone as a whole was 0.49, relative to the land component of the temperature it was 0.43 and for the ocean component, 0.48. The same values were obtained by taking the relationship between the temperature series and the time, while the CO2 concentration relative to time was 0.96. That is, these correlation coefficients merely reflect the fact that the CO2 concentration series is almost linear with respect to time and do not indicate any causal relationship between the two entities temperature and CO2 concentration.

Figure 3.

A 13 point moving average function was applied to the CO2 concentration series in order to minimise the seasonal variation as the satellite temperature measurements are supplied after being adjusted for seasonal variation. This reduced the time span to 572 monthly CO2 values over the period from November 1971 to June 2019. The correlation values were unchanged to the nearest 0.01 on comparing the temperature zones to the averaged CO2 values.

Detrending of the averaged CO2 concentration, as the independent variable, and the Arctic zone temperature, as the dependent variable, gave a correlation coefficient of 0.11 and a Durbin Watson statistic of 1.54 indicating positive autocorrelation. The autocorrelation coefficient was estimated to be 0.232 which when applied to the two detrended series gave a correlation coefficient of 0.087. Applying a test of independence between the transformed variables resulted in a Spearman Rank statistic of 0.096 and probability of 3% thereby rejecting the null hypothesis that the populations were independent at the 5% level.

The detailed relationship between temperature and CO2 concentration is shown in Figure 4. The daily CO2 concentration data for the period 01 January 2018 to 31 December 2019 was taken from the file:

co2_brw_surface-insitu_1_ccgg_DailyData.txt

and the daily temperature data for that period was calculated as the mean of the hourly data from the files:

met_brw_insitu_1_obop_hour_2018.txt

and met_brw_insitu_1_obop_hour_2019.txt.

Extended linear sections in the graphs mark periods for which there was no data.

There are 703 values in the temperature graph and 662 in the CO2 concentration graph.

Figure 4.

Annotations show the periods for which there is 24 hour sunlight and those with no sunlight. There is an obvious negative correlation between the temperature and the CO2 concentration which is the complete opposite to the claim by the UN IPCC that increased CO2 concentration causes an increase in atmospheric temperature. During the period of 24 hour sunshine, the CO2 concentration falls rapidly to a minimum, attributed to photosynthesis, while the temperature rises to a maximum. During the period of no sunshine, the CO2 concentration rises gradually while the temperature approaches a minimum. The same cycle occurs every 24 hours in the lower latitudes across much of the Earth’s surface in phase with the more common 24 hour day-night cycle.

Figure 3 shows that at the end of each yearly cycle the CO2 concentration has marginally increased but there is no such variation apparent in the atmospheric temperature series. This pattern is broken for the monthly CO2 concentration series by volcanic eruptions whereby the annual CO2 increase does not occur contrary to the fact that considerable quantities of CO2 gas are released during an eruption. Presumably the ash cloud from the eruption reduces the Sun’s radiation to a significant degree.