Weather Anomalies and Natural Disasters



Results


The yearly number of all natural disasters and weather anomalies since 1900 is shown in Figure 7 below:



Figure 7. Yearly number of earthquakes (black), volcanoes (blue), droughts (green), hurricanes (solid red), and sunspots (dashed red). The number of sunspots is scaled to match other graphs.


As noted before, up until ~1965 many data were impartial and incomplete, possibly due to the absence of satellite technology and two major war conflicts. The detailed analysis of each event is given below and the correlation between them is shown in Table 1 (largest correlation coefficients that are close to 0.3 or larger are shown in red).




Earthquakes


Figure 8. Yearly number of earthquakes (black) and sunspots, scaled, (red) since 1900.

As can be seen in Figure 8, the number of earthquakes before approximately 1965 is significantly lower than after 1965, probably due to the smaller number of seismic stations at that time. Therefore, below is a graph that closely shows the more reliable time period after 1965.



Figure 9. Yearly number of earthquakes (black) and sunspots (red) since 1965.


It is hard to see an obvious correlation between earthquakes and sunspots, although there seems to be a peak in earthquakes around each sunspot maximum. The calculated correlation between earthquakes and sunspots with the lag of [-2, -1, 0, 1, 2] respectively is given below:

0.158648 ; 0.316627 ; 0.335183 ; 0.237568 ; 0.122465


Volcanoes


Figure 10. Yearly number of eruptions (black) and sunspots, scaled, (red) since 1900


As can be seen in Figure 10, the number of events until approximately 1950 is less frequent. This is probably due to the lesser number of scientific programs to study volcanoes. Also there are two gaps in data in 1915-1920 (corresponds to WW-I) and 1940-1946 (corresponds to WW-II) Therefore, Figure 11 below shows volcanic eruptions for the same time as reliable earthquake data - after 1965.

It looks like volcanoes do not correlate with other events, including earthquakes (which disproves a part of my original hypothesis). But there seem to be a small increase in the number of eruptions every 5-8 years.



Figure 11. Yearly number of eruptions (black) and sunspots, scaled, (red) since 1965


The correlation between volcanic eruptions and sunspots with the lag of [-2, -1, 0, 1, 2] respectively is given below:

-0.124338 ; -0.0580344 ; -0.0819768 ; -0.120825 ; -0.0692361


Droughts


Figure 12. Yearly number of droughts (black) and sunspots, scaled, (red) since 1900


As can be seen in Figure 12, the yearly number of droughts until approximately 1965 is very low. Therefore, below is a graph that shows droughts and sunspots after 1965.



Figure 13. Yearly number of droughts (black) and sunspots, scaled, (red) since 1965


It seems that there is a general increase in the number of droughts around the sunspot maximum. The correlation between droughts and sunspots with the lag of [-2, -1, 0, 1, 2] respectively is given below:

0.203110 ; 0.291821 ; 0.315697 ; 0.163417 ; 0.0600165


Hurricanes


Figure 14. Yearly number of hurricanes (black) and sunspots, scaled, (red) since 1900



Figure 15. Yearly number of hurricanes (black) and sunspots, scaled, (red) since 1965.


From Figure 15 it looks like hurricanes correlate with sunspots in a rather unusual way. There are approximately 2 hurricane cycles in one sunspot cycle, and the maximum of one hurricane cycle is close to the sunspots maximum while the maximum of the next hurricane cycle roughly coincides with the sunspotsí minimum. However, the correlation coefficients between hurricanes and sunspots with the lag of [-2, -1, 0, 1, 2] are close to zero:

-0.0749556 ; 0.0566299 ; 0.167551 ; 0.187455 ; 0.0153343

This is probably because two hurricane cycles in every sunspot cycle can effectively cancel each other out.


Therefore, I separated years where it seems to be a correlation between hurricanes and sunspots from those where it seems to be an anti-correlation: Correlation coefficients for years 1966-1973, 1978-1982, 1986-1993, 1998-2002 (where it seems to be a correlation):

-0.319380 ; 0.0530675 ; 0.297760 ; 0.489824 ; -0.0222017


Correlation coefficients for years 1973-1978, 1982-1986, 1993-1998 (where it seems to be an anti-correlation):

0.0846662 ; -0.314165 ; -0.190513 ; -0.348391 ; 0.0481177


It looks like the separation of the hurricane data into two parts produced a good correlation of 0.49 (with the 1 year time lag) for the first set of data, but smaller anti-correlation, -0.35, for the second set of data (also with the 1 year time lag).


Sum of All Events

I also wanted to find out whether there was any increase or decrease in all of the events added together, and if it had any connection with the 11-year sunspot cycle. Figure 16 shows all events added together (excluding sunspots) as black curve; and the sunspot numbers, scaled, as red curve. During the last 3 sunspot cycles, there seems to be a general increase in the total number of all events as the number of sunspots goes up, and when the sunspot numbers decline, there is a general decrease in the sum of all events.



Figure 16. All events, excluding sunspots, added together since 1965 are represented by the black curve.
The red curve shows sunspot numbers, scaled.


Also it seems that the total number of events around the last solar cycle maximum (1999-2002) increased dramatically, nearly a factor of 2, compared to previous years. If we look at each individual event after 1965, there is no indication of any increase in the number of eruptions and hurricanes (although both have the maximum of their mini-cycles at this time), but there is about 25% increase in the number of eruptions in 2002 and a significant, 2.5 times, increase in the number of droughts in 1999-2000. Similar increase in the number of droughts was also observed in 1982, but the total number of events in 1982 is still much less than in 2000-2002 (Figure 16) because some other events were at their minimum during this time.


Back to Top