AOTW: Storm Prediction Center Moderate & High Risks

Posted by pmarshwx on 3 Feb 2011 | 11 comments

Although most of the country is still digging out and cleaning up from a historic winter storm, commonly referred to as the “Groundhog’s Day Storm”, I decided to go ahead and answer the Question of the Week regarding Storm Prediction Center (SPC) Moderate and High Risks.

Before I answer the questions, first let me explain how I arrived at the answers…

I took the polygon outlining the first Moderate [High] risk during a given year, regardless of time of issuance. This means I treated a 12 UTC issuance the same as a 1630 UTC issuance and 2000 UTC issuance. I took the risk polygon, placed it on a 4km grid (specifically grid number 240), and activate all grid points that fell inside the risk polygon. This left me with a grid of 1′s (inside outlook) and 0′s (outside outlook). I created a grid for each year and then summed all the grids together. This gave me a grid containing the number of times each grid point was within the first Moderate [High] risk of the year. I then divided each grid point by the number of years I was examining. This left me with the probability that a grid point would be contained in the first Moderate [High] risk of the year.

The resulting plots were extremely blocky because of the straight edges of the outlook polygons and the relatively small size of the resulting sample size (19 years for Moderate risks and 18 years* for High risks). In an attempt to better retrieve the underlying probability distribution, I employed a technique called Kernel Density Estimation. I used a gaussian kernel with a bandwidth of 120 kilometers (per Brooks et al. 1998), and extended the resulting distributions out to 5 standard deviations.

To determine when the first Moderate [High] risk is typically issued, I converted the day and month of the first Moderate [High] risk into the corresponding day of year (e.g., 1 July is the 182 day of the year). I then took the mean day of the year and converted it back to day and month. This means that the average first issuance of a Moderate [High] risk may not fall on a day that actually had a Moderate [High] risk issued.

For this analysis I only evaluated the years between 1990 and 2008. This is because in the early years of the National Severe Storms Forecast Center (NSSFC) / Storm Prediction Center categorical risks had a different meaning than they do now. Thus, I tried to limit the analysis to relatively current definitions.

So, here are this week’s answers:

In the graphics below, the colored contour intervals are every 1% and the thin black contours are every 5%.

• Climatologically, where is the most likely location to experience the first SPC Moderate Risk of a given year?
There are actually two areas when using the raw probabilities: southwest Arkansas and southern Alabama. However, when applying the kernel density estimator, the area in southern Alabama has a slightly higher probability. As for when this first Moderate risk is issued, the average first issuance is 28 January.
• What about High Risk?
There is actually more variability (both spatially and temporally) in the issuance of the first High risk than the first Moderate risk. This is most likely the result of the greater variability in time of year, and thus greater variability of the large scale pattern. High risks have been issued as early as 21 January (1999) and as late as 1 July (1997). [Technically it has been issued as late as never, because 2000 did not have a single High risk issued.] For early High risks, the favored location is in the southeast United States where the ingredients that favor severe thunderstorms is more often confined to the Gulf Coast areas. Later in the year, more northern locations are favored as the cyclone track, and ingredients for severe thunderstorms, lifts northward. With this said, the first High Risk is typically issued across portions of southeast Arkansas and northern Mississippi between 27-28 March.

QOTW: Storm Prediction Center Moderate & High Risks

Posted by pmarshwx on 24 Jan 2011 | 4 comments

This week’s “Question of the Week” comes from Trevor Gramling and Ryan Vaughan.

• Climatologically, where is the most likely location to experience the first Storm Prediction Center Moderate Risk of a given year?
• What about High Risk?

Since I do not know the answer ahead of time, this may end up being a two week “Question of the Week” as I write the code to create the answers…

This ought to be fun!

AOTW: More on January Tornadoes

Posted by pmarshwx on 21 Jan 2011 | 5 comments

The answers to this week’s “Question of the Week” might surprise some of you. But first, here are the graphs that answer these questions

Here is a chloropleth map of January tornadoes by state:

Here is a bar graph showing January tornado counts by state, ordered from most to least:

Keeping the states in descending order of most January tornadoes to fewest tornadoes, here is a bar graph depicting the number of injures by state:

And, once again, keeping the states in descending order of most January tornadoes to fewest tornadoes, here is a bar graph depicting the number of fatalities by state:

And even though this plot is a bit crowded, here is a bar chart that combines the three previous charts into one:

Could you figure out the answers? Well, if not, here they are:

1. Which state(s) had the most January tornadoes?
   • Florida (151)

2. Which state(s) had the most January tornado injuries?
   • Mississippi (580)

3. Which state(s) had the most January tornado fatalities?
   • Mississippi (42)

4. Which state(s) had the most injuries per tornado?
   • Deleware (7 per tornado)

5. Which state(s) had the most fatalities per tornado?
   • Oklahoma (0.62 per tornado)

And here are the raw numbers,


State, Tornadoes, Injuries, Fatalities, IPT*, FPT**
AL, 89, 294, 19, 3.30, 0.21
AR, 117, 218, 13, 1.86, 0.11
AZ, 6, 0, 0, 0.00, 0.00
CA, 39, 3, 0, 0.07, 0.00
DE, 1, 7, 0, 7.00, 0.00
FL, 151, 259, 5, 1.71, 0.03
GA, 91, 130, 5, 1.42, 0.05
HI, 6, 4, 0, 0.66, 0.00
IA, 13, 11, 1, 0.84, 0.07
IL, 28, 140, 1, 5.00, 0.03
IN, 17, 7, 3, 0.41, 0.17
KS, 3, 0, 0, 0.00, 0.00
KY, 24, 39, 4, 1.62, 0.16
LA, 123, 142, 10, 1.15, 0.08
MD, 3, 0, 0, 0.00, 0.00
MI, 1, 0, 0, 0.00, 0.00
MO, 77, 276, 8, 3.58, 0.10
MS, 127, 580, 42, 4.56, 0.33
NC, 24, 50, 1, 2.08, 0.04
NE, 6, 0, 0, 0.00, 0.00
NV, 1, 0, 0, 0.00, 0.00
OH, 6, 3, 0, 0.50, 0.00
OK, 16, 32, 10, 2.00, 0.62
OR, 1, 0, 0, 0.00, 0.00
PA, 6, 18, 0, 3.00, 0.00
SC, 26, 44, 0, 1.69, 0.00
TN, 48, 210, 14, 4.37, 0.29
TX, 139, 73, 2, 0.52, 0.01
UT, 1, 0, 0, 0.00, 0.00
VA, 13, 14, 1, 1.07, 0.07
WA, 3, 0, 0, 0.00, 0.00
WI, 3, 5, 0, 1.66, 0.00
WV, 2, 0, 0, 0.00, 0.00

*IPT = Injuries Per Tornado
**FPT = Fatalities Per Tornado


QOTW: More on January Tornadoes

Posted by pmarshwx on 17 Jan 2011 | 2 comments

The answer to last week’s “Question of the Week” sparked a lot of posts on Twitter and at least one question posed in the comments. These comments and questions got me thinking about more ways to dissect the tornado database. Thus, I thought for this week’s installment of QOTW, I would continue with the same theme.

Above is the image originally posted in the answer to last week’s question. It depicts the number of January tornadoes each year, broken down by F/EF-Scale rating. Although containing a lot of information, it fails to answer the question I received most often: “How many fatalities have resulted from January tornadoes?” There figure below displays the number of injuries and fatalities by year for 1950-2009. (Note: The y-axis is scaled via a square-root. The thin, smooth line is a smooth trend-line. Thanks to the fine folks on the “ggplot2” listserv for helping me debug an issue with this plot!)

After a small spike from the mid-1960s through the mid-1970s, injuries from January tornadoes decreased slightly and has held relatively steady around 20-25 per year. The exception to this was 1999, which holds the record for most number of January tornadoes, including the largest January tornado outbreak on record. Fatalities appear to follow a similar trend as injuries, albeit with much lower numbers. In total

• January Injuries: 2455 (40.9 per year)
• January Fatalities: 138 (2.3 per year)

We can break down January tornado casualties even more and examine them by F/EF-Scale ratings.

As one might expect, a general increase in casualties is found as F/EF-Scale rating increases. This leads me to this week’s questions.

Above is an image depicting the number of January tornadoes between 1950 and 2009 broken down by county. Using the above image as a guide, between 1950 and 2009:

1. Which state(s) had the most January tornadoes?
2. Which state(s) had the most January tornado injuries?
3. Which state(s) had the most January tornado fatalities?
4. Which state(s) had the most injuries per tornado?
5. Which state(s) had the most fatalities per tornado?

(Hint: A tornado that crosses a county boundary is counted in both counties. Thus, one cannot sum the number of tornadoes per county in a state to find the number of tornadoes per state.)

AOTW: January Tornadoes

Posted by pmarshwx on 15 Jan 2011 | 2 comments

I apologize that this week’s answer is a day late. I was traveling yesterday and unable to access the Internet. But without further adieu, here is this week’s answer.

1. Since 1950, how many (official) tornadoes have occurred in January?
   • There have been 1193 tornadoes during the month of January in the 60 years spanning 1950 to 2009. This equates to an average of 19.88 tornadoes per year in January. Broken down by rating:

     • F/EF-Unknown: 30
     • F/EF-0: 391
     • F/EF-1: 443
     • F/EF-2: 251
     • F/EF-3: 67
     • F/EF-4: 11
     • F/EF-5: 0

As you can see, 1999 was a very active year in terms of January tornadoes. There were several tornado outbreaks that January, including the largest January outbreak on record, which affected parts of Arkansas, Tennessee, Mississippi, and Louisiana.

AOTW: Temperature Records for KOKC, KLIT & KMEM

Posted by pmarshwx on 7 Jan 2011 | 2 comments

This answers to this week’s “Question(s) of the Week” are:

1. When was the last time Little Rock (KLIT) or Memphis (KMEM) recorded a daily minimum temperature below 0F?
   • The last time either site recorded a daily minimum temperature below 0F was 23 December 1989. KLIT recorded a minimum temperature of -1F and KMEM recorded a minimum temperature of -3F.
2. When was the last time they recorded a daily maximum temperature less than 10F?
   • Since 1950, KLIT has not recorded a day where the the temperature failed to rise above 10F. During this time period, KLIT has recorded only 1 day where the maximum temperature was 10F, which was 24 December 1989. KMEM, however, has recorded two days since 1950 where the temperature failed to reach 10F: 10 January 1962 (6F) and 22 December 1989 (8F).
     
3. Since 1950, has either site recorded a day where the temperature failed to warm to greater than 0F?
   • Since 1950, neither KLIT nor KMEM has failed to warm to greater than 0F.

Congratulations to Tim for his correct answers!

Below are the data for these two sites. If you are interested in the full climatological data (that I have available), let me know and I’ll post it somewhere.

Daily Minimums of 0F or less for KLIT:
1951-02-02: -5 (2.0)
1962-01-10: -1 (5.0)
1962-01-12: -4 (3.0)
1963-12-24: -1 (5.0)
1983-12-25: 0 (0.0)
1985-01-20: -2 (0.0)
1989-12-23: -1 (0.0)

[The format is Date: Minimum (Observed Snow Depth)]


Daily Maximums of 10F or less for KLIT:
1983-12-24: 10 (0.0)

[The format is Date: Maximum (Observed Snow Depth)]


Daily Minimums of 0F or less for KMEM:
1940-01-19: 0 (0.0)
1940-01-25: -2 (2.0)
1940-01-26: -2 (1.0)
1940-01-27: -1 (1.0)
1940-01-28: -1 (0.0)
1951-02-02: -11 (5.0)
1962-01-10: -1 (3.0)
1962-01-11: -1 (3.0)
1962-01-12: -4 (2.0)
1963-01-24: 0 (1.0)
1963-12-23: -8 (13.0)
1963-12-24: -13 (12.0)
1982-01-17: 0 (1.0)
1983-12-25: 0 (0.0)
1985-01-20: -4 (2.0)
1989-12-22: -4 (0.0)
1989-12-23: -3 (0.0)

[The format is Date: Minimum (Observed Snow Depth)]


Daily Maximums of 10F or less for KMEM:
1962-01-10: 6 (3.0)
1989-12-22: 8 (0.0)

[The format is Date: Maximum (Observed Snow Depth)]


QOTW: Temperature Records for KOKC, KLIT & KMEM

Posted by pmarshwx on 3 Jan 2011 | 1 comment

In meteorology we tend to deal in extremes and anomalies. After all, this is what causes the biggest impact on society. Days in which the temperature and precipitation are average, or nearly average don’t result in significant impacts to travel, construction, etc. It’s the extreme or anomalous events that get people’s attention and negatively impact society. With this said, consider the current forecast for the upcoming weekend into much of next week. Below is the Climate Prediction Center’s (CPC) 6-10 Day Temperature Outlook. Notice how the entire contiguous United States is expected to have below normal temperatures!

Both the ECMWF and GFS are signaling a major cold air outbreak of the likes we haven’t seen in years. Forecast temperatures from the GFS keep the central and southern plains below 0C (32F) for well over a week. Additionally, the GFS forecast from Sunday evening (00 UTC Monday) for the Martin Luther King, Jr holiday weekend has daily minimum temperatures of -20C (-4F) or below as far south as northern Texas and northern Louisiana. (Although, admittedly, the 12 UTC run was slightly warmer.) Also, as the next sequence of images shows, portions of Arkansas are not forecast to have daily maximum temperatures above -4F!

Surface (2-meter) temperatures valid Monday, 17 January 2011, at 6 AM CST

Surface (2-meter) temperatures valid Monday, 17 January 2011, at 6 PM CST

Surface (2-meter) temperatures valid Tuesday, 18 January 2011, at 6 AM CST

After seeing the 00 UTC GFS, I was curious about the frequency of such cold periods. I examined the daily temperature and precipitation records for KOKC dating back to 1 January 1948, to determine the number of times KOKC has recorded a minimum temperature of 0F or less. I then repeated this exercise to determine the times in which the daily maximum temperature was less than or equal to 10F. Since 1948, OKC has recorded a minimum temperature of 0F or less 31 times and a maximum temperature of 10F or less 7 times.

Broken down by decades:
1950s, 3, 0
1960s, 5, 0
1970s, 8, 1
1980s, 13, 5
1990s, 2, 1
2000s, 0, 0

[The format is: Decade, # of mins, # of maxes]
(Actual dates/values can be found at the end of this post)


As you can see, the 1990s and 2000s had fewer of each than did the 70s and 80s. This leads me to a point that Harold Brooks brought up earlier today. The National Weather Service (NWS) updates its 30-year averages (normals) every 10 years, in contrast to World Meteorological Organization (WMO) standards of updating them every 30 years. This means that 2011 is the first year that NWS, and specifically the CPC, should be using the “new normals”. Thus, if CPC is using the “new normals”, what impact does this have on the Temperature Outlook above? Simply looking at the extremes listed above, replacing the 1970s with the 2000s in any 30-year average calculation would result in a warmer average. Thus, it would be (theoretically) easier to have a Temperature Outlook where the entire CONUS has a 33% or greater chance of below normal temperatures this year than last year. (Note, the next 30-year average update for WMO will be in either 2020 or 2021).

A word of caution here, I do not know what the actual time range of the 30-year averages are in the CPC forecast above. However, assuming that a contiguous 30-year period is used in the averaging, and we’ve replaced at least one of the 1960s, 1970s, 1980s, with the 1990s or 2000s, then the argument I’ve laid forth is still valid, regardless of the actual years used.

So, readers, my questions to you:

1. When was the last time Little Rock (KLIT) or Memphis (KMEM) recorded a daily minimum temperature below 0F?
2. When was the last time they recorded a daily maximum temperature less than 10F?
3. Since 1950, has either site recorded a day where the temperature failed to warm to greater than 0F?

Leave your answers in the comments! The answers will be posted on Friday!

And as promised, below are the data for the OKC Minimum/Maximum statistics mentioned in the article.

Daily Minimums of 0F or less:
01 Feb 1951: -1 (1.0)
04 Jan 1959: -4 (2.0)
05 Jan 1959: -2 (1.0)
11 Jan 1962: -1 (0.0)
12 Jan 1963: -3 (0.0)
13 Jan 1963: -1 (0.0)
23 Jan 1963: -1 (0.0)
24 Jan 1963: -1 (0.0)
09 Jan 1973: 0 (5.0)
12 Jan 1973: -1 (5.0)
09 Jan 1977: -2 (3.0)
10 Jan 1977: -3 (3.0)
18 Feb 1978: -1 (3.0)
08 Jan 1979: 0 (2.0)
31 Jan 1979: -1 (2.0)
09 Feb 1979: -3 (5.0)
11 Feb 1981: 0 (0.0)
10 Jan 1982: 0 (0.0)
11 Jan 1982: 0 (0.0)
21 Dec 1983: -2 (1.0)
22 Dec 1983: -3 (1.0)
24 Dec 1983: 0 (1.0)
25 Dec 1983: -1 (0.0)
19 Jan 1984: -3 (2.0)
08 Jan 1988: -4 (12.0)
03 Feb 1989: 0 (0.0)
21 Dec 1989: 0 (0.0)
22 Dec 1989: -4 (0.0)
23 Dec 1989: -8 (0.0)
03 Feb 1996: 0 (0.0)
04 Feb 1996: -3 (0.0)

[The format is Date: Minimum (Observed Snow Depth)]


Daily Maximums of 10F or less:
09 Jan 1977: 9 (3.0)
19 Dec 1983: 9 (0.0)
22 Dec 1983: 9 (1.0)
23 Dec 1983: 10 (1.0)
24 Dec 1983: 3 (1.0)
22 Dec 1989: 4 (0.0)
22 Dec 1990: 6 (2.0)

[The format is Date: Maximum (Observed Snow Depth)]