Thursday, January 23, 2014

A New Way to Rate Tornados - An Amatuer Proposal



Tornado Impact Rating (TIR, pronounced “tier”)


            The old F scale and current EF scale rates tornadoes based upon estimated wind speed alone. There is some controversy using these scales. The biggest concern right now appears to be using information from Doppler radar to rate tornadoes. The F scale and EF scale was meant to rate tornadoes based upon damage alone. If a massive tornado occurs out in the middle of a field and strikes nothing, the tornado is supposed to be rated EF-0, even is a Doppler on Wheels is sitting there taking wind speed measurements. Also, there is currently only 6 tiers of tornado rating; EF-0 through EF-5.

            A great deal of attention has been paid to tornadoes that fall in the EF-5 category, even though many EF-4 tornadoes were just as destructive, or maybe even more so. In 2011, several large tornado outbreaks occurred. Two massive tornadoes hit populated areas. The first was the Tuscaloosa tornado and the second was the Joplin tornado. Both tornadoes seemed just as impressive on film, but the Tuscaloosa tornado was rated EF-4 because the damage assessment team didn’t feel the tornado hit anything with “superior” construction. Had this tornado hit a structure with better construction, the rating may have been higher. Which tornado really had higher winds? According to the NWS, it was the Joplin tornado. We may never know the truth.

            The Tuscaloosa tornado was just as impressive as the Joplin tornado. It was on the ground much longer, in an environment that was being sapped for energy from other surrounding storms, and it had a much longer damage path. However, the Joplin tornado will be remembered for it’s tremendous amount of fatalities. When coming up with this new rating system, I tried to take into consideration that fatalities mean a lot more than the length and width of a tornado. One life lost is too many. I admire those who are out in the line of fire studying these tornadoes in order to help protect us in the future.
            I have written to a few people in the field of tornadoes. I haven’t heard back from any of them. I am only presenting this here as a system designed by an amateur. I hope it whets the appetite for a new system that rates tornadoes in a manner that looks at more than just estimated wind speed. I’m sure there are other factors that could have been considered here. I just took some of the things that I find impressive about tornadoes and arbitrarily came up with ways to rate them. I did it by adjusting the Wind Speed Rating. I hope this is but a seed and out of it I hope to see a new way to rate tornadoes that includes Doppler findings and any other methods that are yet unknown.
            When coming up with adjustments, I searched the tornado records and came up with consensus numbers. I gave arbitrary percentages for adjustments, mostly 10%. There is no upper limit so if a new record is set, the percent of adjustment can exceed 10%. It would be wise to stick to these numbers instead of adjusting for every new record as to keep apples to apples in comparison. Again, it would be nice if someone in the realm of science would pick up the ball and run with it. I’m sure they could do a much better job than me. 

Rating the Tornado:

            We start with the wind speed rating in MPH. Unlike the current system, it doesn’t matter how the wind speed is obtained. It can be either by the EF scale rating, or by Doppler on Wheels, or by any other method that may come in the future. We take the speed and move the decimal point two places to the left. This gives us our initial score, or Wind Speed Rating (WSR). It’s from this score that all adjustments will be made with the exception of Fatalities (F) adjustment. For instance, a tornado with a wind speed determined or estimated to be 95 MPH will be given an initial rating of .95 before adjustments. A tornado with wind speeds of 250 MPH will be given a rating of 2.50. There is no upper limit to this. 

I decided to rate a tornado based on the following factors; 

1. Wind Speed (WSR)
2. Width (W)
3. Path Length (PL)
4. Duration (D)
5. Time of Day (TOD)
6. Fatalities (F)           

          I know there are other factors that can play into determining how powerful a tornado is. My overall goal is to gauge a tornado’s overall impact. Of course, wind speed is the measure that we use today. We determine it based upon damage using the EF scale. When no damage has occurred, we currently assign it an EF-0 rating no matter how wide the tornado is, nor how long the path is, etc. We know why wind speed is important, but here is why I feel the other categories are important too. 

Width – While a wide tornado doesn’t necessarily indicate a strong tornado, it does indicate that a larger area has been affected by the circulatory winds. Since this adjustment is based upon the WSR, the adjustment will be lower with lower wind speeds. The greater the wind speed, the greater the adjustment. While not perfect, it is a start. 

Path Length – The average tornado has a path length of 1 mile. Some tornados have tremendous path lengths. It is generally understood that only powerful tornados can achieve long path lengths. Some tornados are nearly stationary. Depending on the duration of the tornado and the strength of the winds, a short path may not actually indicate worse impact for any one point. Again, it’s not perfect, but I am open to ways of improving this adjustment. 

Duration – The average tornado lasts for 15 minutes. The longer the tornado is on the ground, the greater the impact is upon people and property.  

Time of Day – I thought of adding time of year as well, but that would depend on where the tornado touched down. Dixie Alley has a lot of tornados when there’s still snow on the ground in the Midwest. So for that reason, I stuck to time of day only. The least frequent time for a tornado is between 4 and 5 a.m. For that reason, I chose 4:30 as the median time. The most frequent occurrence of a tornado occurring happens between 5 and 6 p.m. I chose 5:30 p.m. as the median time. That makes 11 hours between the median high point and median low occurrence point in the hours after median high point. There are 13 hours between the two points before median high occurrence. It takes special dynamics to fuel a storm at 4:30 a.m. without the help of the sun. For that reason, I choose to use this as a factor in determining strength and overall impact of the storm. Many people are sleeping at 4:30 in the morning and less apt to be tornado aware at that point.  

Fatalities – This is the hardest adjustment for me. No greater impact can a tornado have than taking someone’s life. Homes can be rebuilt. Property can be replaced. For that reason, I went with a 100% adjustment for extreme as opposed to a 10%. I also decided to figure the fatality adjustment by using the WSR with all other adjustments added. If you look at my ratings for the Joplin tornado vs. the Tuscaloosa tornado of 2011, you’ll see the fatality rating is what made the major difference in the score. The wind speeds were very similar. The Tuscaloosa tornado was on the ground longer and had a much longer path length. In those regards, the Tuscaloosa tornado scored much higher. However, the fatalities that occurred with the Joplin tornado meant that it had a much higher impact than the Tuscaloosa tornado. The EF scale alone cannot point those differences out to you while mine does.  
             I did not add injuries to my scale. I did this because injuries haven’t always been reported historically, and in many cases, injuries are approximated. I didn’t add a category for financial damage because the value of money changes all the time. I wanted to keep this apples to apples as much as possible. Again, it’s not perfect, but I’m hoping it’s the start of a discussion to scrap the current system for rating tornados and move to a system that rates a tornado on more than wind speed based upon damage alone.
            Also note, I got my information on tornado extremes and averages from several different sources, but mainly from the NWS sites themselves. I chose to put everything into feet and minutes. For forward speed, I divide the number of feet in the path length by the duration of the tornado. I have added some notes where the NWS gives different numbers for forward speed, but in many cases, forward speed is not included in NWS surveys. For that purpose, I choose to stick with my method of using the average.
Adjustments: 

Forward Speed – 10% for extreme. [6424-352(average)=6072/10=607.2]

To adjust, take forward speed in feet, subtract 352 and divide by 607.2. Move your decimal place two places to the left. Round to nearest hundredth. Multiply this by wind speed rating. This is your adjustment. Less than 352 ft/m = No Adjustment
           I would eventually like to see an adjustment for nearly stationary tornadoes as I believe a stationary tornado is capable of tremendous damage. This is one area where expert input and guidance would be most helpful.

Width – 10% for extreme [13728-250(average)=13478/10=1347.8]

To adjust, take width in feet, subtract 250 and divide by 1347.8. Move your decimal place two places to the left. Round to the nearest hundredth. Multiply this by the wind speed rating. This is your adjustment. Less than 250 ft.= No Adjustment

Path Length – 10% for extreme [1156320-5280(average)=1151040/10=115104]

To adjust, take width in feet, subtract 5280 and divide by 115104. Move your decimal point two places to the left. Round to the nearest hundredth. Multiply this by the wind speed rating. This is your adjustment. Less than 5280 ft. = No Adjustment

Duration - 10% for extreme [210-15(average)=195/10=19.5]

To adjust, take time in minutes, subtract 15 and divide by 19.5. Move your decimal point over two places to the left. Round to the nearest hundredth. Multiply this by the wind speed rating. This is your adjustment. Less than 15 minutes = No Adjustment

Time of Day adjustment – 10% for 4:30 a.m. target. Tornadoes are rarest from 4-5 a.m. They are most common from 5-6 p.m. I go from 5:30 p.m. to 4:30 a.m. That is 13 hours before and 11 hours after. To determine the adjustment, one needs to determine the median time that the tornado was on the ground. Take the times given for the start and end of the tornado and find the time exactly in the middle (+/- 1 minute). Use the following guidelines to adjust.

Take the median time of the tornado. Determine the number of minutes before or after 5:30 p.m.

If the median time is before 5:30 PM and after 4:30 AM, divide the minutes by 78. Move the decimal point two places to the left and round to the nearest hundredth. Multiply this by the wind speed rating. This is your adjustment for mean time before 5:30. 

If the median time is after 5:30 PM and before 4:30 AM, divide by 66. Move the decimal point to places to the left and round to the nearest hundredth. Multiply this by the wind speed rating. This is your adjustment for mean time after 5:30. 

Fatalities – 100% for extreme with adjustment added to final tally. [695-0=695/25=6.95]

To adjust, take fatalities divide by 6.95. Move your decimal point over two places to the left. Round to the nearest hundredth. Multiply by the wind speed sub-total with all other adjustments added. This is your adjustment to add to that number.

Some notes and facts

Average Tornado
Wind Speed: 110 mph
Forward Speed: 352 ft. per minute (average 4 mph)
Widest Width: 250 ft.
Path Length: 5280 ft.
Duration: 15 minutes
Time of day: 5-6 P.M.
Fatalities: Less than 1 

Extreme Tornado
Wind Speed: 301 mph 1999 Bridge Creek-Moore Tornado
Forward Speed:  6424 ft. per minute (73 Mph) Tri-State Tornado
Widest Width:  13728 ft. (2.6 Miles) May 31, 2013 El Reno, Oklahoma tornado
Path Length:  1156320 ft. (219 miles) Tri-State Tornado
Duration:  210 minutes (Tri-State Tornado)
Time of day:  None really 4-5 a.m. is the rarest timeframe.
Fatalities: 695 Tri-State Tornado 

Joplin
Wind Speed: 200+
Forward Speed: 3070.74 ft. per minute (average 34.89 mph)
Widest Width: 3960 ft.
Path Length: 116688 ft.
Duration: 38 minutes
Time of day: 5-6 P.M.
Fatalities: 158 

TIR
Wind Speed Rating (WSR) 2.00
Adjustments
Forward Speed (FS) .08
Widest Width (W) .06
Path Length (P) .02
Duration (D) .02
Time of Day (TOD) .00
Sub-total: 2.18
Fatalities (F) .50
___________________________________________________________
Total (TIR): 2.68
TIR: 2.68

Notes from NWS: The forward speed of the tornado through most of Joplin was about 20 to 25 miles per hour 

Tuscaloosa
Wind Speed: 190 mph
Forward Speed: 1681.21 ft. per minute (average 19.10 mph)
Widest Width: 7800 ft.
Path Length: 425990.4 ft.
Duration: 91 minutes
Time of day: 4-6 P.M.
Fatalities: 65 

TIR
Wind Speed Rating (WSR) 1.90
Adjustments
Forward Speed (FS) .04
Widest Width (W) .11
Path Length (P) .08
Duration (D) .08
Time of Day (TOD) .00
Sub-total: 2.21
Fatalities (F) .20
___________________________________________________________
Total (TIR): 2.41
TIR: 2.41

Notes from NWS: The storm traveled approximately 500 km in 8.5 hours, at an average forward speed of 59 km/hr (36 mph)  

Greensburg
Wind Speed: 205 mph
Forward Speed: 2339.45 ft. per minute
Widest Width: 8976 ft.
Path Length: 152064 ft.
Duration: 65 minutes
Time of day: 9:32 P.M. Median Time
Fatalities: 11 

TIR
Wind Speed Rating (WSR) 2.05
Adjustments
Forward Speed (FS) .06
Widest Width (W) .12
Path Length (P) .02
Duration (D) .06
Time of Day (TOD) .08
Sub-total: 2.39
Fatalities (F) .05
___________________________________________________________
Total (TIR): 2.44
TIR: 2.44
Notice how the Greensburg tornado is actually three hundredths of a point higher than the Tuscaloosa tornado. 



Bridge Creek-Moore 1999 
Wind Speed: 301 mph (Doppler)
Forward Speed: 2360.47 ft. per minute
Widest Width: 5280 ft.
Path Length: 200640 ft.
Duration: 85 minutes
Time of day: 7:05 P.M. Median Time
Fatalities: 36

TIR
Wind Speed Rating (WSR) 3.01
Adjustments
Forward Speed (FS) .09
Widest Width (W) .12
Path Length (P) .06
Duration (D) .12
Time of Day (TOD) .03
Sub-total: 3.43
Fatalities (F) .17
___________________________________________________________
Total (TIR): 3.60

























No comments: