System Performance and Convective Weather...ATL BOS BWI CLE CLT CVG DCA DEN DFW DTW EWR FLL IAD IAH...

F065-B06-007© 2006 The MITRE Corporation. All rights reserved.

System Performance and Convective Weather

Kenneth Lamon, Ph.D.15 March 2006

©2006. All rights reserved. This is the copyright work of The MITRE Corporation and was produced for the U.S. Government under Contract Number DTFA01-01-C-00001 and is subject to Federal Aviation Administration Acquisition Management System Clause 3.5-13, Rights in Data-General, Alt. III and Alt. IV (Oct., 1996). No other use other than that granted to the U.S. Government, or to those acting on behalf of the U.S. Government, under that Clause is authorized without the express written permission of The MITRE Corporation. For further information, please contact The MITRE Corporation, Contracts Office, 7515 Colshire Drive, McLean, VA 22102, (703) 983-6000.

F065-B06-007© 2006 The MITRE Corporation. All rights reserved.2

June, July and August are the Peak of the Convective Weather Season

Source: NOAA, FSL; measures the average area of thunderstorms over the US at 19Z, 21Z, and 23Z , weekly average

0.0%

0.2%

0.4%

0.6%

0.8%

1.0%

1.2%

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

Perc

ent C

onve

ctiv

e C

over

age

(Wee

kly

Ave

rage

, Lev

el 3

) 20022004

2003

2001


Arr

ival

and

Dep

artu

re D

elay

s (>

1 H

our)

-

1,000

2,000

3,000

4,000

5,000

6,000

16,500 17,500 18,500 19,500 20,500 21,500 22,500Scheduled Arrivals

200020032004

Delays versus Scheduled ArrivalsDelays = 500 + a*Weather*(Schedule - 16,500)


Min

utes

of A

rriv

al D

elay

-

100,000

200,000

300,000

400,000

500,000

600,000

16,500 17,500 18,500 19,500 20,500 21,500 22,500Scheduled Arrivals

200020032004

Minutes of Delay versus Scheduled ArrivalsMinutes = 100,000 + b*Weather*(Schedule - 16,500)


An Airport-specific Convective Weather Metric

• Take an airport’s maximum storm intensity each hour

• Multiply by the number of scheduled arrivals that hour

• Sum over all hours of the day• Sum over 45 airports• (Divide by total number of scheduled arrivals)


View the PPT movie here


R2 = 0.51

-

100

200

300

400

500

600

700

800

900

1,000

0 200 400 600 800 1000 1200 1400 1600 1800Weather score

One

-Hou

r Del

ays

at A

TL

ATL Delays Versus Weather Score


0

2000

4000

6000

8000

10000

12000

ABQ

ATL

BNA

BOS

BWI

CLE CLT

CVG DC

AD

END

FWD

TWEW

RFL

LH

OU

IAD

IAH

IND

JFK

LAS

LGA

MC

IM

CO

MD

WM

EM MIA

MSP

MSY

OR

DPB

IPD

XPH

LPH

XPI

TR

DU

SEA

SLC

STL

TEB

TPA

200406200407200408

YYYYMM

Monthly Convective Weather Score by Major Airport


Number of Hours Having Thunderstorm Intensity>3

BROWN indicates locations where the frequency of thunderstorms increased from 2004 to 2005

34

30

14

35

15

25

3736

41

18

21

64

42

19

61

3619 37

29

91

32

31

75

70

24

58

3028

38

30

32

97

June 1 - August 31, 2004

72

28

34

46

3215

77

26

66

107

28

26

104

44

341631

52

41

58

97

25

28

24

2919

35

37

31

22

28

59

June 1 - August 31, 2005


SEA

PDX

TPATEBSTL

SLC

RDU

PIT

PHX

PHL

PBI

ORD

MSY

MSPMIA

MEM

MDW MCO

MCI

LGA

LAS

JFK

IND

IAH

IAD

HOU

FLL

EWR

DFW

DEN

DCA

CVG

CLT

CLE

BWI

BOS

BNA

ATL

ABQ

-

2,000

4,000

6,000

8,000

10,000

12,000

14,000

16,000

- 5,000 10,000 15,000 20,000

Weather Score

Dela

ys (

>1 H

ou

r)

DTW

Delay and Convective Weather Score:

June - August 2004


33

3

33

3 3

3

2

2

2

2

22

2

2

2

2

2

2

22

2

11

1

1

1

1

1

1

1

1

1

1

1

1

1

11 1

1

1

1

1

G1 G2 G3ABQ DEN BWIATL HOU CLEBNA LAX CVGBOS MCI DTWCLT MCO INDDCA MIA MDWDFW MSY PITEWR OAK RDUFLL PBIIAD PDXIAH SANJFK SEALAS SFOLGA SJCMEM TPAMSPORDPHLPHXSLCSTLTEB

R-Squared Improves (a little) if Airports are Grouped


Delays Versus Weather Score:45 Airports

R2 = 0.67

R2 = 0.62

-

1,000

2,000

3,000

4,000

5,000

6,000

- 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000

Weather*Schedule

Del

ays

(>1

Hou

r)

2003

2004


Regression Results 2003 and 2004

Delays = 507 + 1.8*Weather*(Schedule - 15,429)

+ 298*Year2004R-squared = 0.65

Minutes = 106,000 + 174*Weather*(Schedule - 16,870)

+ 24,454*Year2004R-squared = 0.67


Airborne Holds Versus Weather Score2003 and 2004

R2 = 0.59

R2 = 0.53

-

100

200

300

400

500

600

0 1000 2000 3000 4000 5000 6000 7000 8000 9000Weather*Schedule

Airb

orne

Hol

ds (>

15 M

inut

es)

2003

2004

45 Airports


Diversions Versus Weather Score2003 and 2004

R2 = 0.50R2 = 0.48

-

50

100

150

200

250


Div

ersi

ons

20032004

45 Airports


R2 = 0.63

R2 = 0.52

-

100

200

300

400

500

600

700

800


Can

celle

d D

epar

ture

s

2003

2004

Cancellations Versus Weather Score2003 and 2004

45 Airports


2005

2003

0

1000

2000

3000

4000

5000

6000

7000

8000

0 2000 4000 6000 8000 10000Weather*Schedule

One

-Hou

r Sch

edul

e D

elay

s

200320042005

2004

Delays Versus Weather Score2003, 2004, and 2005

45 Airports


R2 = 0.46

R2 = 0.51

0

100

200

300

400

500

600

0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000Weather*Schedule

Airb

orne

hol

ds (>

15 m

in)

2005

2004

Airborne Holding Up in Summer 2005

Year

There were 28 additional holds perday in summer 2005 compared to 2004

45 Airports


Airborne Holding Up in Summer 2005

*June-August

Year

0

10

20

30

40

50

60

70

ATL

BOS

BWI

CLE CLT

CVG DCA

DEN

DFWDTW

EWR

FLL

IAD

IAH

JFK

LAS

LAX

LGA

MCO

MDW

MEM MIA

MSP

ORD

PHL

PHX

PIT

SAN

SEA

SFO

SLC

STL

TPA

Num

ber o

f Dai

ly A

irbor

ne H

olds

(>15

min

)

200320042005

Holding was up sharply at ATL in summer 2005


Schedule Delays Lasting More Than One Hour Up 16% Compared to 2000

221,928

171,629

120,721

149,168

219,108

256,904

-

50,000

100,000

150,000

200,000

250,000

300,000

2000 2001 2002 2003 2004 2005

Arrivals and Departures More than One Hour Late Relative to Schedule

Source: ASPM, 45 Airports


Delays Due to GDPs Up in 2004 and 2005

Source: OPSNET, includes delays attributed to en route centers

0

10

20

30

40

50

60

70

80

90

2000 2001 2002 2003 2004 2005

thou

sand

s

GDPGSDeparture

OPSNET Delays by Type, 45 Airports, June-August


0

20

40

60

80

100

120

2000 2001 2002 2003 2004 2005

thou

sand

s ATL, ORD, PHL, EWR, LGA,IAD, IAH, JFK37 other airports

Proportion of OPSNET Delays at the 8 Most Delayed Airports Has Grown

Includes delays attributed to ARTCCs. Delays up 7% overall compared to 2000

Up 23% compared to 2000

Down 39% comparedto 2000

OPSNET Delays, 45 Airports, June-August


0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

2000 2001 2002 2003 2004 2005

mill

ions

ATL, ORD, PHL, EWR, LGA,IAD, IAH, JFK37 other airports

Operations, on Average, Up 9% at the 8 Most Delayed Hubs, and Down 9% Elsewhere

Down 9% comparedto 2000

Up 9% compared to 2000

Operations down 5% overall at OPSNET 45 airports compared to 2000

Operations June-August (OPSNET)


ATL Simulated Minutes of Arrival Delay Using Arrival Capacity from 8/7/05

-

10,000

20,000

30,000

40,000

50,000

60,000

1150

1200

1250

1300

1350

1400

1450

Scheduled Arrivals

Arr

ival

Del

ay M

inut

es (S

imul

ated

)

Aug 11 - 15, 2003

Aug 10, 2003

Aug 9, 2003

Schedule Dates2003

Schedule Dates2004

Aug 9 - 13, 2004

Aug 8, 2004

Aug 7, 2004

Aug 8 - 12, 2005

Aug 7, 2005

Schedule Dates2005


Summary

• Introduced a convective weather metric defined as the product of an airport’s maximum hourly thunderstorm intensity and the number of flights scheduled to land that hour

• Metric has several advantages– Can be aggregated and disaggregated– Small, computationally simple dataset– Yields many insights about NAS performance


Other Findings

• Factoring out differences in weather and the total number of scheduled ops, hour-long delays increased substantially from 2003 to 2004 and from 2004 to 2005– Increases caused by greater concentration of traffic at

already delayed airports• Airborne holding and diversions were unchanged

from 2003 to 2004, however, holding increased from 2004 to 2005


And Lastly

• Similar approaches for measuring system performance based on delays, cancellations, diversions, holding, etc. ought to yield results similar to those presented here.


Appendix


-18-10-4+4+10+18

Difference(Hours)

Source: National Convective Weather Detection

Number of Hours Having Thunderstorm Intensity>3 (June - August 2005 Minus June - August 2004)


OPSNET Delays by Type, 45 Airports, June-August

OPSNET Delays Up 7% Compared to 2000

0

20

40

2000 2001 2002 2003 2004 2005

145,919 138,312

94,087 92,543

127,111 136,715

60

80

100

120

140

160

thou

sand

s

GS/GDPArrivalDeparture

Source: OPSNET, includes delays attributed to en route centers


Average Number of Seats per Departure has Decreased

-8%

-19%

-15%-16%

-1%

-17%-10%

0

20

40

60

80

100

120

140

160

ATL EWR IAD IAH LGA ORD PHL

Ave

rage

Sea

ts p

er D

epar

ture

Jul-00Jul-05

Average number of seats per scheduled departuredown in July 2005 compared to July 2000

Source: OAG


Change in Operations by Major Airport

SFO-17%

-15%LAX

IAH

SEA-23%

-2%

+19%

-16%DFW

ORD+9%

PHX

STL-38%

BOS

MEM-1%

-15%MSP+2%

CVG+6%

DTW-5%

MDW-2%

IAD+16%

ATL+5%

CLE-22%

Summer 2005 vs Summer 2000

SLC DEN+23%+11%

MIA-26%

LAS+29%

DCA-18%

LGA

PHL+12%

+8%

JFK+1%

EWR-1%

PDX-19%

SAN+14%

BWI-3%

FLL+14%TPA

-4%

MCO-3%

Source: OPSNET, Overall change = -2%, 35 Airports


Change in Operations by Major Airport

SFO+1%

-1%LAX

IAH

SEA-5%

+4%

+12%

-10%DFW

ORD-1%

PHX

STL+4%

BOS

MEM+9%

-1%MSP0%

CVG-1%

DTW+1%

MDW-14%

IAD+10%

ATL-2%

CLE-1%

Summer 2005 vs Summer 2004

SLC DEN+13%+2%

MIA-4%

LAS+8%

DCA+3%

LGA

PHL+15%

+1%

JFK+8%

EWR0%

PDX-4%

SAN+9%

BWI0%

FLL+7%TPA

+12%

MCO+9%

Source: OPSNET, Overall change = -2%, 35 Airports


14,3

19

7,08

1

5,33

2

4,24

3

3,76

1

2,58

4

1,75

4

1,43

8

856

492

349

295

279

179

102

94 76 71 67 55 51 48 35 29 17

-

2,000

4,000

6,000

8,000

10,000

12,000

14,000

16,000

ATL

OR

D

EWR

LGA

PHL

BOS

IAD

SFO

JFK

BWI

LAS

IAH

DC

A

MD

W

MSP TE

B

FLL

DTW CVG CLT

SEA

PHX

HO

U

LAX

DFW

Delays from GDPs: June - August 2005

In summer 2005 ATL accounted for 33% of all delays from GDPsThe average duration of this type of delay was 74 minutes

Source: OPSNET


PHL Simulated Minutes of Arrival Delay Using Arrival Capacity from 7/13/05

-

5,000

10,000

15,000

20,000

25,000

30,000

35,000

40,000

45,000

50,000

55,000

480

530

580

630

680

730

Scheduled Arrivals

Arr

ival

Del

ay M

inut

es (S

imul

ated

)

Jul 14 - 18, 2003

Jul 12, 2004

Schedule Dates2003

Jul 12 - 16, 2004

Jul 11, 2004

Jul 10, 2004

Schedule Dates2004

Jul 11 - 15, 2005

Jul 10, 2005

May 30, 2005

Schedule Dates2005


PHL: Actual Minutes of Arrival Delay, June - August

-

5,000

10,000

15,000

20,000

25,000

30,000

35,000

40,000

45,000

480 530 580 630 680 730Scheduled Arrivals

Min

utes

of A

rriv

al D

elay

(Act

ual)

200520042003

Actual Delay2003

Actual Delay2004

Actual Delay2005

Source: ASPM, Each point represents one day, one outlier day excluded from 2003 and 2004


0

5

10

15

20

25

30

ORD ATL DFW LAX PHX LAS MSP DTW IAH CVG PHL

tho

usa

nd

s

0

5

10

15

20

25

30

IAD MIA BOS EWR STL LGA SEA SFO JFK MDW TEB

tho

usa

nd

s

2000 2001 2004 2005

Ordered by busiest airport, includes delays attributed to en route centers

OPSNET Delays for 22 Most Delayed Airports: June - August


ATL: Actual Minutes of Arrival Delay, June - August

-

10,000

20,000

30,000

40,000

50,000

60,000

70,000

80,000

90,000

1150 1200 1250 1300 1350 1400 1450Scheduled Arrivals

Min

utes

of A

rriv

al D

elay

(Act

ual)

2005

2004

2003

Source: ASPM, Each point represents one day, two outlier days excluded from 2003 and 2004

Actual Delay2003

Actual Delay2004

Actual Delay2005


To Improve Airport Model

• Adding a variable representing amount of system-wide delay raises r-squared to 0.66 for ATL

• For ORD two-variable model has r-squared of 0.71


Three-Letter Airport Abbreviations

ATL - Atlanta International MDW - Chicago Midway

BOS - Boston/Logan InternationalMEM - Memphis International

BWI - Baltimore-Washington InternationalMIA - Miami International

CLE - Cleveland-Hopkins InternationalMSP - Minneapolis St Paul International

CVG - Covington/Cincinnati International

MSY - New Orleans International

DCA - Reagan National

OAK - Metropolitan Oakland International

DEN - Denver International

ORD - Chicago O'Hare International

DFW - Dallas-Ft Worth International

PBI - Palm Beach International

DTW - Detroit Metro Wayne Co

PDX - Portland International

EWR - Newark International

PHL - Philadelphia International

IAD - Washington Dulles International

PHX - Phoenix Sky Harbor International

IAH - Houston/G Bush Intercontinental

PIT - Pittsburgh International

IND - Indianapolis International

RDU - Raleigh-Durham International

JFK - John F Kennedy International

SEA - Seattle Tacoma International

LAS - Las Vegas/Mc Carran International

SFO - San Francisco International

LAX - Los Angeles International

SJC - San Jose International

LGA - La Guardia

SLC - Salt Lake City International

-

STL - Lambert-St Louis International

TEB - Teterboro

ABQ - Albuquerque International Sunport

BNA - Nashville International

CLT Charlotte/Douglas Internationa

MCI Kansas City International

MCO - Orlando International

FLL - Fort Lauderdale/Hollywood InternationalHOU - William P. Hobby Airport

SAN - San Diego International

TPA - Tampa International


Glossary

ASPM Aviation System Performance MetricsFSL Forecast Systems LaboratoryGDP Ground Delay ProgramNAS National Airspace SystemNOAA National Oceanic & Atmospheric AdministrationOAG Official Airline GuideOPSNET Operations NetworkUS United StatesZ Zulu Time

System Performance and Convective Weather...ATL BOS BWI CLE CLT CVG DCA DEN DFW DTW EWR FLL IAD IAH...

Documents

Transcript of System Performance and Convective Weather...ATL BOS BWI CLE CLT CVG DCA DEN DFW DTW EWR FLL IAD IAH...