WeatherNation Blog

October Tornado Tendencies

Much ‘todo’ has been made in recent years about the frequency and location of tornadoes in the United States, not to mention the “Enhanced” Fujita Scale. Since I have been asked to blog on items of interest related to current weather ‘happenings,’ and provide synopses of my various research endeavors, today we will briefly discuss October tornado tendencies.

An example of your tax dollars actually providing sound benefit is the work of John Hart at the Storm Prediction Center (SPC). A man with excellent forecasting abilities and a constant desire to provide added benefit through the creation and maintenance of additional tools for those in the research community. I often use his Severe Plot application, or the “csv” archive at the SPC to retrieve textual data of past tornado events that can be seamlessly integrated into spatial and statistical applications.

The first chart below displays all tornado events from the years 1950 through 2006 for the month of October only. A brief examination shows an increasing frequency of events as the years progress. Historically this has been attributed to an increased “awareness” of severe weather potential, increased saturation of media coverage, and quite importantly the integration of the WSR-88D radar network utilized by the National Weather Service that replaced the aging WSR-74 and archaic WSR-54 systems.

U.S. October Tornado Frequencies (1950-2006) (click for full size)

According to this wikipedia entry “the first installation of a WSR-88D for use in everyday forecasts was in Sterling, Virginia on June 12, 1992. The last system was installed in North Webster, Indiana on August 30, 1997.” Glancing again at our October frequency chart above, one could glean a possible reason for the increase in events reported was due to the implementation of those systems. Similar to my previous discussions covering snowfall and tropical activity, our level of technologies for the detection and observation of any weather event are not nearly what is necessary for complete confidence in the analysis of historic tendencies. However, our country is certainly coming closer to a reasonable level of comfort, and is leading the way worldwide.

In keeping with the above concept, a quick synopsis of general statistics of events for the Month of October yield some interesting results. I decided to divide the reports above into two sections. First a window of October events from 1950-1987, and the second window covers the balance of 1988-2006.

October Events from 1950-1987
Total Events

AverageMedianMode
804221818
October Events from 1988-2006
Total Events

AverageMedianMode
1025545519

At first glance of the table above, confirming the former chart, one can immediately see the average of reported events during the second time frame are more than double those of the previous window. Does this confirm that the implementation of the WSR-88Ds dramatically increased our ability to detect, report, and confirm tornado events? Could it be that atmospheric processes are changing due to the influx of responses via global warming? One thing is certain, there has been an increase of reported events, not just for the month of October, but overall as well as I will discuss in later posts.

Historic October Tornado Event Locations for the Years 1950-2006 (click for full size)

An examination of the spatial probability of October tornadoes can at first glance be slightly misleading, which is often the cause for excellent debate. Note in the probability polygons above the broad deep spread of the red color along the coastal states representing the geographic areas of highest likelihood for October tornadoes to form. Do you notice a trend in the distribution? Can you immediately think of a cause?

Hurricane spawned tornadoes have often been debated as perhaps not being ‘genuine’ based upon the dynamics of their formation as compared to those formed through the variety of thunderstorm types. In addition, the ability to detect and then easily differentiate the events in ‘after the fact’ physical investigation has only recently become available as discussed above. Further discussion has also been held representing tornadoes that form further inland from the remnants of hurricanes propagating over land as an “if-then” type of debate. These types of discussions are important as the sciences of meteorology and climatology mature both over time, and through future advances in technology that will allow a deeper examination and understanding of weather related events.

Based upon the graphic above, where do you think tornadoes might form this month? Recall from one of my posts yesterday that two reports of ‘possible’ tornadoes were received in eastern North Carolina. Looking into my crystal ball, I also see the possibility of tornadoes in our future for next week. Be sure to stay tuned to your favorite weather outlet, stay informed, and stay safe!

cheers,

–patrick

HWphp 3.991 Released

We have released HWphp 3.991. This release provides a few bug fixes.

A few of the changes include (See the complete change list for the full listing):

  • Fixed bug in Template.php that was causing url_pands to do a double urlencode
  • Fixed a type that prevented “template_html_encode=#” from setting the global encode option.
  • Implemented fix when using html encoding with a formatting command
  • Change in the hw3.php to fix issue with some users of the use_sites functionality having issues.

If upgrading from a previous version of HW3php you only need to download the appropriate upgrade file listed under “upgrade distributions”. To determine the version of your current HW3 installation,  you can normally query your HW3 installation by appending “?version=1″ to your HW3 URL. I.E.  hw3.php?version=1

NOTE: If you are using HW3 with a debug mode password required then the above URL may not display the HW3version information. In such cases you may need to query your HW3 installation requesting debug output with your debug password. I.E. hw3.php?debug=1&debugpw=XXXXX (where XXXXX is your debug password)

Calendar
October 2008
MonTueWedThuFriSatSun
« Sep Nov »
 12345
6789101112
13141516171819
20212223242526
2728293031