Confirmation of the IPCC Prediction re: Increased Storminess

Rob Kuhlman

Montgomery County Community College

Pottstown, PA. 19464


Confirmation of IPCC Prediction, Pt. 1


You're going to be working on this project in four groups; I'll allow you folks to choose the groups -- just make sure that they're approximately equal in size. We'll continue working in these groups on Friday as well.

When we began this unit several weeks ago, we looked at Table 1 on page 9 of the Intergovernmental Panel for Climate Change (IPCC) report; within that table we noted that the following predictions were considered "likely, over some areas":

  • Increase in tropical cyclone peak wind intensities
  • Increase in tropical cyclone mean and peak precipitation intensities
For the past several weeks we've been working through, largely in the form of group projects, some fundamental meteorological concepts: volume/temperature relationships, pressure/temperature relationships, capacity/temperature relationships, latent heat, the role of angle of incidence in Earth's heat budget, pressure gradients, behavior of cyclones. You've also studied one specific tropical cyclone in some detail, derived some understanding of the behaviors of tropical cyclones in general, and, most recently, utilized these concepts in developing the structure of the formation and demise of tropical cyclones. With all that background it's appropriate that we now attempt to test the validity of the IPCC prediction.
Procedure I -- Setting up the Seasonality

Click on link to the AVHRR website ( administered by Johns Hopkins University's Advanced Physics Lab, to open the AVHRR home page in a new window. AVHRR is the abbreviation for Advanced Very High Resolution Radiometer, a satellite which measures the intensity of infrared radiation being radiated by the surface of the oceans. The AVHRR satellite makes direct measurements, in other words, of sea surface temperature (SST).

Once at the AVHRR home page, go to the left margin frame and click on the link for 'Gulf of Mexico' and follow the instructions below:

  • Click on Multi-Day Merged Data, From April 1996 to the Present
  • Click on Jan under 2005's listings
  • Scroll down to the column labeled 'Seven-Day Composite' and click on JAN 11 2005 (The reason we're using 7 day composite images is that daily anomalies are smoothed out and cloud cover is erased.)
  • When the image loads, enlarge it to full size (remember the 'enlarge' icon in the lower right corner of the image?)
  • Notice the temperature scale just below the right bottom edge of the image. We're going to use it to answer the following questions:

  1. Notice the Florida Keys. What's the temperature of the water just south of the Florida Keys? This region, the Straits of Florida, is the source region for the Gulf Stream.
  2. What is the temperature of the water on the north side of the Keys?
  3. What is the temperature of the Gulf of Mexico just offshore of Tampa Bay (the larger and more northerly of the embayments on Florida's west coast)?
  4. Now, swing counterclockwise up to the Mississippi Delta region. What is the temperature of the Gulf waters offshore of the delta?
  5. Over to the Texas coast, what's the temperature of the mid-Texas Gulf waters?

Now, back out of this image, and click on the Browser's 'back' icon to display the calendar for 2005, but this time choose April. Then, click on 'Seven-Day Composites', APR 1, 2005. Enlarge this image and answer the following questions:

  1. Temperature of water in the Florida Straits (south of the Keys)?
  2. Temperture north of the Keys?
  3. Temperature west of Tampa Bay?
  4. Temperature south of the Mississippi Delta?
  5. Temperature off the mid-Texas Gulf Coast?

Now, back out of the April image, choose July, then choose JUL 11, 2005. Same questions:

  1. Temperature south of the Keys?
  2. Temperature north of the Keys?
  3. Temperature off Tampa Bay?
  4. Temperature south of the Delta?
  5. Temperature off the Texas coast?

Back out of July, and choose September, then choose SEP 15, 2005 (around Rita time). Once again, same questions:

  1. Temperature south of the Keys?
  2. ...north of the Keys?
  3. Tampa Bay?
  4. the Delta?
  5. the Texas coast?

You've examined SSTs for the Gulf at roughly the four seasonal change positions of the calendar. Consider the following questions:

  • At which time of the year did you find the warmest water temperatures?
  • And at what time of year do we find the greatest number of tropical cyclones?

Procedure II -- Storm Tracks

Back out of the September image, then click on Hurricanes in the left frame. Click on 2005 Atlantic Hurricane Season Track Maps & Satellite Images, then click on Katrina. Over on the right side, click on the upper image thumbnail depicting Katrina's track over the Atlantic Ocean. Enlarge this image. Consider the following questions:

  1. What was the average wind speed when the storm first appears on this image, down in the Bahamas?
  2. What is the wind speed when it leaves the image northwest of the Keys?
  3. What is the SST underneath the storm track?
Now, back out of this image and then open and enlarge the image of Katrina over the Gulf.
  1. What is the wind speed where the track is northwest of the Keys?
  2. What is the maximum wind speed that Katrina attains?
  3. What is the wind speed when landfall in the Delta occurs?
  4. What is the wind speed when the storm moves inland and the storm reaches the northern margin of the image?
  5. What is the SST in the Gulf under the storm track?

...content to be continued on Friday...


Please respond to all the bulleted and/or numbered questions above. Today's exercise will count for 5 (scored on a 5-4-3-2-0 basis) of the total of 15 points for the complete two-day IPCC project. Given that we have a pattern of absences from one day to the next, I'll assess each day's work separately. Please sign each person's name in the group to the group report and then submit your group report to our course's Digital Dropbox so I can access it while I'm away. Each person present will earn the same score.

Confirmation of IPCC Prediction


Procedure III -- Is the Gulf getting Warmer?

Before we get into this project, be alert to saving your work on a regular basis. Our network tends to get 'jumpy' around this time of the semester.

  • Go back to the AVHRR site which we used in Procedure I by clicking on this link: AVHRR website (
  • Click on the Gulf of Mexico region in the left frame, and then click on Multi-Day Merged Data, From April 1996 to the Present.
  • Scroll down the list to 1997, (1996, the first year of the satellite's operations, doesn't display data for us) and click on Sep.
  • Now click on Sep 01, 1997 in the '3 day composites' column. Enlarge this image.
  • Choose four separate points within the Gulf (it's your choice about which points to choose, but at least one should be out in the middle somewhere) and record the temperatures for those four points in the table below.
  • Now back out of the 1997 data, choose 1998, and choose Sep 01, 1998. Record the temperatures for the same four points.
  • Continue recording the temperatures for the same four points for the earliest date in September right up to 2007 and enter the temperature data into the table below.

YearTemp #1Temp #2Temp #3Temp #4

  • Next, click on Applications from the bottom toolbar, click on Microsoft Office, then open up Excel.
  • Transfer your year and temperature data into a spreadsheet.
  • Create another column within which you'll compute the average of your four locations for each year.
  • Click on the top cell in your average column to highlight it, then go to the 'Formulas' tab, then the 'Insert Function' ribbon block (to the left), click on 'AutoSum' and select 'Average'. You'll see that Excel is trying to include the calendar year cell in your computation, so you'll need to adjust the cell range (highlighted in black) to restrict the computation to just your four data cells.
  • Now, highlight your entire column of data cells in the average column. Click on the 'Home' tab, go over to the right 'Editing' ribbon block, click on the little blue arrow icon (which points down), and select 'Down'. Excel will now compute each year's average.

Next we want to create a graph.

  • Highlight your column of averages, including the column heading.
  • Click on the 'Insert' tab.
  • From the chart type options, click on 'Line' and then click on the 'Line with Marker' option (middle left icon).
  • Now, click on your graph's line to highlight it, and, without moving the cursor, 'Right click' with your mouse.
  • In the new dialog box, choose Add Trendline
  • Make sure the 'Linear' radio button is highlighted, then look further down in this dialog box, and click on the box labeled 'Display R-squared value...' and click on 'Close'. You may need to click and drag on the R2 equation to drag it into an empty location on the graph.
  • Click on the 'Design' tab, click on the 'Move Chart Location' box, and click on the radio button for 'New Sheet'.
  • Since we may be pressed for time during this project, don't worry too much about formatting the graph -- you can always quickly add information with a pen after printing it.
  • Print out both your Excel data table and this graph.

Consider the following questions:

  1. Does your trend line have a negative, zero, or positive slope?
  2. Given that there's a very limited range of data on Gulf temperatures (the satellite has only been measuring for a decade) does this data support an assertion that SSTs should be getting progressively warmer?

Procedure IV -- Are Tropical Cyclones Becoming More Numerous or Nastier?

We'll next consider whether the storms have become more numerous, and also we'll try to assess whether they're becoming more severe.

  • Back on the AVHRR website, click on the left frame link for Hurricanes.
  • Click on Atlantic Tropical Activity Summary, 1967-2007. Enlarge this image.
  • Simultaneously open up Excel and a new spreadsheet.
  • For each year indicated in the chart, record the following three items in the Excel spreadsheet:
    • 1st column = Year
    • 2nd column = Total number of storms
    • 3rd column = Number of storms Category 2 through 5 (green, yellow, orange, red)
  • Get Excel to draw two line graphs (use the instructions above) -- one for the total number of storms and another for the # of Cat 2-5 storms.
  • Get Excel to create a trend line through the data points of each of your two graphs. Also, have Excel print the R2 value. Print out both of your graphs, and label them by hand so you don't get them mixed up.
Consider the following questions:
  1. Do either of the trend lines show a positive slope? (i.e., slants up to the right?)
  2. How much 'noise' is there in your data? Another way to ask the question is to consider to what extent there are numerous peaks and valleys above and below the trend line? (This is the purpose of the R2 value.)

Let's try another data source. Click on the following link to a Unisys database of hurricane activity ( and follow the directions:
  • You'll see a large array of links to annual data sets; this data extends back as far as 1851. To the extent that we can trust data which extends back to the 'dark ages', it's the best we've got. There's too much for us to analyze in our alloted time, so what we're going to do is sample the data at decadal intervals. Group 1 will examine 1855, 1865, 1875, etc.; Group 2 will do 1856, 1866, 1876, etc.; Group 3 will do 1857, 1867, 1877, etc.; and Group 4 will do 1868, 1878, 1888, etc.; .
  • For each year in your group's record, you'll be noting the number of storms, the maximum wind speed measured for that year, and the minimum atmospheric pressure (if indicated).
  • Open up Excel again, create a new spreadsheet, and enter the following data:
    • Column 1 = Year
    • Column 2 = # of storms
    • Column 3 = maximum wind speed for that year
    • Column 4 = minimum atmospheric pressure for that year
    If a particular data item is missing for a storm, don't enter a '0' in the spreadsheet cell; rather, leave the cell blank.
  • Get Excel to plot up a line graph for each of your three data variables and then get Excel to create trend lines, including the R2 value, through each set of data points. Print out all three graphs. Label them so you don't get them mixed up.

Consider the following questions for the Unisys data:

  1. Do any of your trend lines show a change to either a greater number of storms, or an increase in maximum wind speed, or a reduction in minimum central atmospheric pressure? (If you do see a change in maximum wind speed with reduction in minimum pressure, do they correlate with one-another?)
  2. How much noise is there in the data? That is, to what extent does decade-to-decade variation overwhelm any pervasive change in the variable?


The fundamental question around which this unit has been centered is whether the IPCC prediction of 'increased storminess' as an outcome of global warming (as expressed by an increase of SST) has any legitimacy. It's a testable hypothesis. To what extent did the hypothesis survive the testing you have done today? In response to my query, consider the following questions:

  1. Did you see any evidence that SST (represented by a small sample, the Gulf of Mexico, over a short period of time, ~10 years) has been rising?
  2. Did you see any evidence in the two data searches that you conducted that storms have either gotten more numerous...?
  3. ...with higher wind velocities...?
  4. ...or with lower minimum central pressures?
  5. To the extent that some or all of your graphs show no convincing trends, please explain why this might be so?


Please respond to all the numbered questions above (there are 11 of them) and submit them along with all your graphs. This exercise will count for the remaining 10 of the total of 15 points for the complete IPCC project (scored on a 10-8-6-4-0 basis). Given that we have a pattern of absences from one day to the next, I'll assess each day's work separately. Please sign each person's name in the group to the group report which gets turned in. Each person present will earn the same score.