USCRN TEMPERATURES 2005-2019

1. The chart above is seen in social media and denier blogs with the start year fixed at 2005 and the end year varying from 2017 to 2019 depending on when it was posted. It is meant to show that there is no global warming evident in the data. The chart shows mean temperatures from 114 USCRN station data across the USA. It is claimed that the unique property of this chart is that the start date of 2005 ensures that the UHI (urban heat island) effect has been removed from the data. It is thus argued based on this chart that AGW global warming is a faux creation of UHI because when UHI is removed the evidence for warming disappears from the data. It is further argued that therefore, all datasets that show warming, including the UAH satellite data, are flawed because they show warming when it is now clear that global warming is a creation of UHI. 
2. This post is a critical evaluation of this line of reasoning. First, mean USA temperature cannot be expected to mimic global mean temperature because of the geographical limitation and second, the time span of 15 years is too short for this kind of test as it has been shown in related posts that even strong warming trends do not show consistent monotonic warming year after year but rather that the long term trend is simply the net result of violent warming, cooling, and no trend cycles at decadal and multi-decadal time scales [LINK] [LINK] .
3. FIGURE 1: The chart and regression table in Figure 1 below show UAH global mean temperatures for the eight calendar months January to August over the time span, 2005-2019. The chart appears to show a warming trend and in the corresponding regression table we find p-values<0.05 in four of the eight months with an average p-value for all eight calendar months computed as p-value=0.0488 < 0.05. We conclude that there appears to be some evidence of a warming trend in the UAH global mean temperatures 2005-2019 though it does not appear to be a very strong warming trend.
4. Figure 2: A corresponding analysis of UAH temperatures for the USA lower 48 states is presented in Figure 2. Here the chart appears flat with no indication of a warming trend. The regression table shows very high p-values with an average p-value of 0.2637. The UAH data appear to agree with the USCRN data above that no trend in USA temperatures is evident for the time span 2005-2019.  A comparison of Figure 1 and Figure 2 indicates that the absence of trend in the USCRN temperatures cannot be ascribed to the removal of UHI but with peculiarities of the geographical limitation of the analysis. This comparison does not support the interpretation of the USCRN data in  terms of UHI. In other words, the data do not show that a warming trend was removed because of the exclusion of the UHI effect.
5. Figure 3: That the 15-year time span chosen for the UHI study is too short for drawing conclusions about AGW is shown in the four charts of Figure 3. Here we see large fluctuations in trends for a moving 15-year window as its end year moves through the full span from 1993 to 2019. The overall trend for the full span is therefore understood as a long term warming that accumulates from these violent swings. These charts also provide convincing evidence that 15-year temperature trends do not provide useful information nor information relevant to an empirical test of AGW theory.
6. Figure 4: Shows results for a full span trend analysis of the UAH USA48 data. Very little evidence for statistically significant trends is found in stark contrast with the UAH global temperatures presented in Figure 6 where robust warming trends are found for all eight calendar months January to August with an aveage p-value=0.0002. The comparison of Figure 4 and Figure 6 provide additional evidence that the difference in trends between the USA and the world is not due to UHI but to geographical limitations of the USA-only analysis. It is noted in this context that AGW theory does not claim nor require that all geographical locations should warm at the same rate but only that a long term warming trend should exist in the global mean temperature.
7. SUMMARY AND CONCLUSION: The data presented in the chart at the top of this post are too limited by geography and time span in its trend analysis to have significant implications for AGW where only long term trends in deseasonalized mean global temperature are relevant. Also, the claim that the absence of trend can be attributed to the removal of the UHI effect in 2005 is not supported by the trend profile showing large changes in trends in a moving 15-year window where trends as low or lower are seen prior to 2005. The comparison of USA data from USCRN and UAH does not indicate that the absence of trend since 2005 is due to the removal of the UHI effect because no such intervention exists in the UAH data. 

FIGURE 1: UAH GLOBAL TEMPERATURES 2005-2019 

FIGURE 2: UAH TEMPERATURES FOR USA48 2005-2019

FIGURE 3: TRENDS IN A MOVING 15-YEAR WINDOW: UAH-USA48 1979-2019

FIGURE 4: UAH USA48 FULL SPAN TRENDS: 1979-2019

FIGURE 5: UAH GLOBAL TRENDS IN MOVING 15-YEAR WINDOW 1979-2019

FIGURE 6: UAH GLOBAL FULL SPAN TRENDS 1979-2019

[HOME PAGE]