Atlantic Hurricanes Part II

Recap:  Atlantic Hurricanes Part I

In the earlier assignment, we noted that the 2020 hurricane season broke many records. We wondered whether this indicated a long-term trend, perhaps due to global climate change. 

We used data and definitions from HURDAT (available at Wikipedia Links to an external site.) and the National US Hurricane Center. Links to an external site.

Using data from 1851-1930, we calculated the mean and standard deviation for the number of named storms, the number of hurricanes, and the number of severe hurricanes.  The results were:

We learned that from 1981 to 2010, the yearly average was 12 named storms, 6.5 hurricanes, and 2 major hurricanes.  We thought about whether this supported the claim that extreme weather events have become more frequent.  Some of you may have investigated (in the Discussion) how this might be related to global climate change.

Testing Hypotheses about a Population Mean

Now you have tools of statistical inference available to revisit this question.  Here we look at a sample of 5 years selected randomly from 1960-2020.  The results are:

1. Find the sample mean number of Named Storms.
2. What statistical test should be used, and why?
3. What are the conditions for performing this test? Do you think that they have been met?  Why or why not?
4. Question #3 is not obvious. The sample is supposed to come from a population that is normally distributed.  What is the population of interest?  Do we know whether or not it is normally distributed?  If not, how could we find out?
5. Test the hypothesis that named storms have become more frequent than they were between 1851 and 1930. Use a 5% level of significance and show all steps of the hypothesis test.  State your conclusion in a complete, correct sentence.

Now let’s look at the data for Hurricanes.  The distribution for 1851-1930 was also roughly mound-shaped and symmetrical, though not perfectly normal.  You can confirm (if you find the data on the web) that the distribution for 1960-2020 is similar, though it is somewhat skewed right.  We will assume that it’s close enough to normal to conduct the hypothesis test.

6. Test the hypothesis that Hurricanes have become more frequent than they were between 1851 and 1930. Use a 5% level of significance and show all steps of the hypothesis test.  State your conclusion in a complete, correct sentence.

Finally, let’s consider Major Hurricanes.  The distribution for 1851-1930 was strongly skewed right, but maybe that has changed?  You can find the data on the Internet, but we already checked.  The distribution for 1960-2000 is still skewed right, if a little less strongly.  A t-test is probably not appropriate.  But you can do this:

7. Find the sample mean number of Major Hurricanes.  Then, using the “reference” data, find its z-score.  What does that suggest to you?

Now return to the Discussion:  Hurricanes and Global Climate Change.  Whether or not you posted in this Discussion earlier, please post again based on the work you did in this activity.