Datasets and Hurricanes

Hello readers! Today we’ll be returning to the topics of hurricanes (or cyclones, or typhoons, if you prefer). We won’t be talking about a specific one in this, but about hurricanes in general–more specifically, two different datasets related to them.
Hurricanes are, of course, dangerous. The windstorms they can bring are not any nicer–they can cause damage to infrastructure and buildings, mess with the local economy, and are potentially lethal. There is less land for them to impact in the Southern Hemisphere, but can instead disrupt shipping in the nearby oceans.
Scientists have yet to completely understand the variation in storm frequency (amount per time period) from year to year and decade to decade. Some factors are known, but some pieces of the puzzle have yet to fall into place.
Hurricanes Madeline and Lester. Image credit NASA Earth Observatory; their Image of the Day for August 31st, 2016.
Hurricanes Madeline and Lester. Image credit NASA Earth Observatory; their Image of the Day for August 31st, 2016.
Enter today’s article, Different long-term trends of extra-tropical cyclones and windstorms in ERA-20C and NOAA-20CR reanalyses” by Daniel J. BefortSimon WildTim KruschkeUwe Ulbrich and Gregor C. Leckebusch. Published in 2016 in the Royal Meteorological Society’s journal Atmospheric Science Letters, this short article looks at two different datasets involving hurricanes; the ‘extra-tropical’ part just means ‘outside the tropics’, so not looking between 20 degrees North to 20 degrees South. One dataset is called NOAA-20CR, the other ERA-20C. 
These two datasets were combined with information of the average pressure at sea level, surface winds over the ocean, sea surface temperature, and sea ice extent. Using this data hurricanes were identified and tracked, along with windstorms. They were split into three time groups for easier analysis; 1901-1930, 1931-1960, and 1961 to 1990.
One would expect–or at least hope–that these two datasets show the general trend in frequency of hurricanes and windstorms. Unfortunately, that was not quite the case here. Regardless of which hemisphere and whether one looked at all hurricanes, or extreme hurricanes only, or windstorms, there was a notable difference between the two sets. Sometimes one set said the storms in question were becoming less frequent, while the other more frequent. Sometimes they agreed, but at different rates (for example, rapidly becoming more frequent or slowly increasing), and sometimes disagreed at different rates. Generally there was less agreement in the earlier two periods, with more agreement in the later time frame.
It should be noted that the NOAA-20CR data set involved a changing number of observations in the twentieth century, which may also be true of ERA-20C and the cause of some of the discrepancies. There are also some inherent differences in the data sets, including how winds near to the surface were found, that may also have been a factor. The differences between the two datasets, and the results thereof, will have to be the topic of another paper. Regardless, we will need to be careful it both data sets are being used in the same study, especially if a longer time period is in question.


Befort, Daniel J. et al. “Different long-term trends of extra-tropical cyclones and windstorms in ERA-20C and NOAA-20CR reanalyses”. Atmospheric Review Letters. 17: 586–595 (2016). DOI: 10.1002/asl.694
Today’s main article
Carlowicz, Mike. Hurricanes Madeline and Lester. NASA Earth Observatory. 31 Aug. 2016. Web.
A page from NASA Earth Observatory; the source of the image used in this post, it includes a brief discussion of the pictured storms.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s