Mike Robinson

Report
Applying FAA Aviation
Weather Metrics Program
Research to Operational
Benefits Evaluations
Mike Robinson
AvMet Applications, Inc
FPAW - 2012
NBAA Orlando, FL
AvMet Applications, Inc.
1800 Alexander Bell Dr., Ste. 130
1
Reston, VA 20191
Challenges to Quantifying Operational Benefits
Attributed to Enhanced Weather Forecasts
• Normalizing for similar weather events
• Normalizing for similar operations
• Defining pertinent baselines; How much
of impact was “unavoidable”
• Attributing improved decisions to
(a) improved forecasts, (b) USE of
improved forecasts
• Providing objective, data-driven,
quantified benefits estimates
“Of course delays were down this July
compared to last…..there was 60% less
convection!”
“Of course delays are down,
ORD demand was down 20%!”
“What do you mean you have no
way to apply this new forecast?”
(Said during solid ME to Gulf line)
“Weren’t those delay improvements
Associated with that new procedure / TMI?
you say delays were saved during all 200
days of convection in ZMA Center….how
often were improved decisions derived from
this new forecast?
“We are in a tight fiscal environment….
I am not going to just take your word
that this new forecast increases
operational efficiency!”
2
FAA ANG-C6 Aviation Weather Metrics Program
Emphasis on:
• Weather impacts in context of operations (specifically, operational
decision-making)
• Weather event normalization and baselining
• Objective, relational data analysis (and simulation support)
3
Weather Impact Traffic Index (WITI)
Weather Weighted by Traffic
WITI is a weighted sum of three
components:
– En-route Component: hourly
frequency on major flows X amount
of convective Wx that these flows
cross
– Terminal Component
Non-linear (Queuing Delay) part
reflecting excess traffic demand vs.
capacity
Queuing Delay Modeled by Wx Index Software, PHL, Dec 14, 2006
150
120
Optimum capacity in good Wx
Dep+Arr demand
Dep+Arr capacity
Queuing Delay, Hrs
90
60
30
0
FO
G
FO
G
FO
G
FO
G
FO
G
RA
IN
FO
G
FO
G
RA
IN
RA
IN
RA
IN
RA
IN
NO
N
N E
LG O
T_ NE
RA
IN
FO
G
RA
IN
FO
G
FO
G
FO
G
FO
G
FO
G
FO
G
FO
G
–
Linear part: capacity degradation
due to terminal weather impact,
proportional to number of ops
Hourly Demand, Estimated Capacity, and Delay
–
Used by the FAA and NWS on a regular basis:
Actual queuing delays at
PHL on 12/14/06 were
significant (223 flights
delayed > 15 min) and
there were approx. 100
cancellations (so the
resulting queuing delay
was less than it could
have been if there were
no cancellations)
0600Z Dec 14 - 0600Z Dec 15, Hourly Wx Observations
• Macroscopic system performance measure in an objective manner – weekly reports
• Compare different seasons’ Wx/traffic impact with outcomes (e.g. delays)
4
4
Historical Scalable Record of WITI / WITI –
Forecast Accuracy (FA)
Region - Hourly
Airport – Hourly
(IAH)
NAS - Daily
5
Assessing Components of Unavoidable Weather
Impacts
LGA, Apr 3, 2009 Arrival Rates
Bad Weather
45
40
Reasonable risk mgmt
35
AAR
30
25
20
Network effects
Deficit
15
10
Actual arrivals
Arr rate based on METAR
Need to give preference to departures
Arr rate based on 4hr TAF
Natural “Inertia” (time to ramp up)
Scheduled arrivals
5
GDP rate
0
6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 0 1 2 3 4 5
Hour, Z
Impact caused by en-route airspace
Operationally Beneficial Delay
(slight arrival oversupply)
Accidents, outages, VIP flights, security, etc
Airlines’ business practices
(e.g.
related to cancellations)
Airspace design, procedures
Inaccurate forecast
Overly conservative TMI
TMI execution (in excess of risk mgmt)
En-route to Terminal to Final Transition
6
Estimating Unavoidable Arrival Delay
Unavoidable Delay Percentage Distribution
90
Mean: 44%
STD: 10%
80
Frequency
70
Model-Estimated Overall and Unavoidable Arrival Queuing Delay Factors,
Jan-Jul 2011, OEP34, Sorted from Highest to Lowest Overall Delay
60
50
1600
40
30
20
1400
10
Overall arrival queuing delay
Bad Wx at destination, Risk margins, etc
Network Effects
Operationally-Beneficial Delay
0
1200
Bin
1000
• Modeled estimated
unavoidable arrival
delay 40-55% of
overall delay
• Results similar to
independent estimates
of unavoidable delay
(e.g., REDAC report)
Daily
Delay
Total,
Hours
800
600
400
200
0
1
11
21
31
41
51
61
71
81
91
101 111 121 131 141 151 161 171 181 191 201 211
Day #
7
Similar Weather Impact Events (WX & Demand)
Example: ORD – Varying impacts throughout the day, both convective and non-convective
Wx may have
been the same
but there was
less traffic in the
morning
ORD, 20 August 2009 - Airport Weather Impact Matrix
WITI Values
8
Finding Similar Weather Impact Days
Similar Weather Event/Impact Evaluation Portal (SWEEP)
Identify, rank,
rank, and
and inspect
inspect similar
similar REGIONAL
NAS wx-impact
days days
•• Identify,
wx-impact
9
Dynamic Airspace Routing Tool
What is DART?
Weather-Aware Superfast-Time NAS/ATM Simulation Model
•
Full ETMS flight plans
•
Terminal Wx (convective and non-convective), TRACON and En-route convective Wx: actual
and forecasts; permeability
•
Airport RWY configurations and capacity (may be Wx-degraded) but no physical RWYs
•
Airspace (sector, Center) capacity (may be Wx-degraded)
•
TMIs (Playbook, GDP, GS, AFP, MIT)
- Can blend historically enforced TMIs and simulated TMIs
•
Reroutes, delays, Cnx, simulated airborne holding & diversions
•
User-definable rules, risk factors, equipage profiles, etc
•
Randomized Wx, airport/airspace capacity, Wx forecast, traffic
A “superfast-time” NAS simulation tool
•
Day-in-the-NAS (50,000+ flights and all the above detail) in 2-3 min
Abundant output on various aspects of NAS operations
•
Impact (cost) metrics include delay (airborne, ground, etc.) AND cancellations and diversions
•
Validation of DART includes all three primary impact metrics
10
Examining Alternative TMIs via DART Simulations
Optimized solution: Airway J29 open to relieve traffic
on VUZ playbook reroute; reduced MIT, less delay
Non-optimal solution: VUZ playbook reroute traffic uses
standard route; J29 closed; heavier MIT, longer delays
Only the traffic using NAS Playbook reroutes is shown; Color-coding by delay: 0-15, 15-20, 30-60, 60-120, >120 min arrival delay
11
Weather-ATM Analysis and Visualization
Environment (WAVE)
• “Weather-Centric” air traffic / ATM analysis tool and research
platform
• Utilizes 1-min ETMS traffic data (flight information, flight
plans, amendments, lat/lon/altitude positioning)
• Can ingest and display any gridded or polygon-based weather
product (diagnostic or forecast)
• Engine for extensive analysis, utilizing multiple data types;
output results in CSV format
• Generate standard output reports, targeted for specific
performance assessments accounting for weather / forecasts
12
WAVE
13

similar documents