Enhanced Fujita Scale 6 23 04

Slide1: 

ENHANCED FUJITA SCALE (EF Scale) 11A briefing presented to NATIONAL WEATHER SERVICE Silver Spring, Maryland June 28, 2004 Wind Science and Engineering Center Texas Tech University

Theodore T. Fujita: 

Theodore T. Fujita

Limitations of Fujita Scale: 

Limitations of Fujita Scale Difficult to apply consistently Not enough damage indicators Does not account for construction quality No definitive correlation between damage and wind speed

Fujita Scale Enhancement Project: 

Fujita Scale Enhancement Project WISE agreed to initiate project Funding was available through NIST Co-PI’s each have more than 30 years experience with tornado damage

Co-PI’s: 

Co-PI’s James R. McDonald, Ph.D., PE. Tornado damage documentation experience Tornado hazard assessment Tornado damage mitigation Kishor C. Mehta, Ph.D., P.E. Director of WISE (Retired) Chair ASCE 7 Wind Load Task Committee Internationally recognized researcher

WISE Strategy: 

WISE Strategy Choose a steering Committee Involve many users Develop a plan Obtain a consensus

Steering Committee: 

Steering Committee

Steering Committee Objectives: 

Steering Committee Objectives Organize a forum of users Identify key issues Recommend a new or modified Fujita Scale Develop strategies to obtain a consensus

Fujita Scale Forum: 

Fujita Scale Forum March 7-8, 2001 Grapevine, Texas 20 of 26 invited participants attended Developed strategies for an enhanced Fujita Scale

Strategies: 

Strategies Define additional damage indicators Correlate appearance of damage with wind speed Preserve historical tornado data base Obtain input from users

Damage Indicators (DI’s): 

Damage Indicators (DI’s) WISE team proposed 28 DI’s Buildings, structures and trees DI’s described in detail Additional DI’s can be added in future

Degrees of Damage (DOD’s): 

Degrees of Damage (DOD’s) Each DI has several degrees of damage DOD’s range from no damage to total destruction DOD’s are arranged in order of increasing damage They are a function of wind speed

Correlation of Damage and Wind Speed: 

Correlation of Damage and Wind Speed Need expected, upper and lower bound wind speeds for each DOD Expected wind speed based on “normal” conditions Upper and lower bound wind speeds represent possible deviation from the “normal” situation

Approach: 

Approach Deterministic Monte Carlo Expert elicitation

Expert Elicitation: 

Expert Elicitation Used successfully for estimating seismic physical parameters Senior Seismic Hazard Assessment Committee (SSHAC – 1997) Experts make best estimates of expected, upper and lower bound wind speeds Follow a well-defined protocol The end result is the best possible estimate of the desired parameter

SSHAC Elicitation Process: 

SSHAC Elicitation Process Describe DI’s and DOD’s Identify and engage a panel of experts Discuss issues with experts; provide data Train experts in elicitation process Conduct individual elicitations and group interactions

SSHAC Elicitation Process: 

SSHAC Elicitation Process Analyze and aggregate elicitations and resolve issues Refine wind speed estimates with several iterations Document and communicate process and final results Obtain additional peer review of process and results

Elicitation Experts: 

Elicitation Experts

Elicitation Procedure: 

Elicitation Procedure Wind speeds are 3-second gusts at 10 m in flat open terrain Experts met for one and one-half days Conducted 3 rounds of elicitation

Results of Elicitation: 

Results of Elicitation Name and description of DI DOD’s and estimated wind speeds Order DOD’s by increasing wind speeds Plot DOD’s versus wind speed Provide photo examples of DOD’s

One-and Two-Family Residences (FR12): 

One-and Two-Family Residences (FR12) Typical Construction: Asphalt shingles, tile, slate or metal roof covering Flat, gable, hip, mansard or mono-sloped roof or combination thereof Plywood/OSB or wood plank roof deck Prefabricated wood trusses or wood joists and rafter construction Brick veneer, wood panels, stucco, EIFS, vinyl or metal siding Wood or metal stud walls, concrete blocks or insulating concrete panels Attached single or double garage

One-and Two-Family Residences (FR12): 

One-and Two-Family Residences (FR12)

One-and Two-Family Residences (FR12): 

One-and Two-Family Residences (FR12)

One-and Two-Family Residences (FR12): 

One-and Two-Family Residences (FR12) FR12: DOD4: Uplift of roof deck and loss of roof covering (>20%); garage door collapses outward

One-and Two-Family Residences (FR12): 

One-and Two-Family Residences (FR12) FR12: DOD6: Large sections of roof removed; most walls remain standing

One-and Two-Family Residences (FR12): 

One-and Two-Family Residences (FR12) FR12: DOD7: Top floor (First floor in this case) exterior walls collapsed

One-and Two-Family Residences (FR12): 

One-and Two-Family Residences (FR12) FR12: DOD10: Total destruction of entire building

Correlation of Fujita Scale and EF Scale: 

Correlation of Fujita Scale and EF Scale Used a second group of experts They assigned Fujita Scale categories to each DOD Ratings were converted to 3-second gust median wind speeds Obtained average of Fujita Scale wind speeds

Correlation of Fujita Scale and EF Scale: 

Correlation of Fujita Scale and EF Scale Performed a regression analysis to obtain correlation between average Fujita Scale and expected EF Scale wind speeds Regression equation:

Correlation of Fujita Scale and EF Scale Wind Speeds: 

Correlation of Fujita Scale and EF Scale Wind Speeds

Derived EF Scale Wind Speed Ranges: 

Derived EF Scale Wind Speed Ranges

Recommended EF Scale Wind Speed Ranges: 

Recommended EF Scale Wind Speed Ranges

EF5 Wind Speed Range: 

EF5 Wind Speed Range We recommend no upper bound on this category Physical upper bound tornado wind speed not known Will avoid folks assuming worst case scenario for EF5 category

Rating an Individual Building: 

Rating an Individual Building Find DI that matches the building type and construction Observe the damage and match to one of the DOD’s Determine if wind speed to cause observed damage is higher, lower or equal to the expected value within the wind speed range

Rating an Individual Building: 

Rating an Individual Building The assigned EF Scale rating is the one whose range of wind speed contains the estimated wind speed to cause the DOD. Additional DI’s should be considered in assigning and EF Scale to a tornado event

Rating a Tornado Event: 

Rating a Tornado Event Conduct an aerial survey to identify potential DI’s and to define extent of damage path Identify 2 or more DI’s that seem to indicate the highest wind speed in the path Locate these DI’s within the damage path Follow steps for individual buildings or structures and document results

Rating a Tornado Event: 

Rating a Tornado Event Considering several DI’s, estimate maximum tornado wind speed Assign EF Scale category based on the maximum estimated wind speed Record basis for EF Scale rating Record other pertinent data relating to the tornado event

Presentations and Workshops: 

Presentations and Workshops Fujita Symposium, January 2000 National Severe Storms Workshop, March 2001 U.S. National Conference on Wind Engineering, June 2001 AMS National Conference, January 2002 21st Conference on Severe Local Storms, August 2002 11th International Conference on Wind Engineering, June 2003 22nd Conference on Sever Local Storms, October 2004 (Paper accepted)

WISE Website: 

WISE Website www.wind.ttu.edu

Conclusion: 

Conclusion We have followed the strategies of steering committee and forum Provided additional damage indicators Established correlation between damage and wind speed Determined correlation between Fujita and EF Scales Presented our work in a number of venues