AMDAR Forecast Applications

1
AMDAR Forecast Applications

• JMA Meeting
• March 23, 2009
• Carl Weiss
• NWS Aviation Services Branch

2
Before We Begin

• You may have heard aircraft data referred to
  as ACARS, MDCRS, or TAMDAR
• ACARS (Aircraft Communications, Addressing, and
  Reporting System) is the name of a datalink
  service provided by Aeronautical Radio, Inc. that
  sends data between aircraft and ground stations
• MDCRS (Meteorological Data, Collection and
  Reporting System) is the weather portion of the
  ACARS data stream provided by seven major U.S.
  air carriers
• TAMDAR (Tropospheric Airborne Meteorological DAta
  Reporting) data are provided by a private
  company, AirDat, who uses regional air carriers

However, Aircraft Meteorological DAta Relay
(AMDAR) is the preferred term by the WMO and NWS.
3
AMDAR Data Have Many Applications

• Aviation
• Low level wind shear
• Ceilings and visibilities
• Icing and turbulence
• Winter Storms
• Precipitation type
• Lake effect snow
• Thunderstorms
• Convective initiation
• Calculation of stability indices
• Fire Weather
• Mixing heights
• Haines indices
• Relative humidity forecasts
• Marine Forecasts
• Small craft and Gale Warnings
• Hazardous Materials Support
• AMDAR can be used to support HAZMAT teams

4
Aviation Applications

• AMDAR soundings in vicinity of airports provide
  meteorologists a real time view of the atmosphere
• This can result in more accurate forecasts of low
  clouds, fog, low level wind shear and other
  aviation weather elements
• AMDAR flight level data provide important
  information in flight corridors
• TAMDAR aircraft report icing and turbulence data

5
Low-Level Wind Shear

• AMDAR data can be useful in determining the
  presence of low-level wind shear (LLWS)
• Aircraft ascending or descending in the vicinity
  of an airport are in an ideal location for
  depicting low-level wind shear

6
Low-Level Wind Shear

• An example of this was noted by the forecasters
  at the Green Bay, WI WFO on the evening of
  October 29, 2005
• The TAF that night indicated LLWS was forecast to
  begin after 0600 UTC. TAMDAR soundings from
  around 0120 UTC showed LLWS already was present

7
Low-Level Wind Shear
8
Low-Level Wind Shear

• The forecaster was able to update the TAF and
  begin the LLWS more than 3 hours earlier than did
  the prior forecast
• This was mentioned in the Area Forecast
  Discussion that was issued around
  0245 UTC

9
Ceilings and Visibilities

• AMDAR provides very useful information for
  forecasting low ceilings and fog that can greatly
  impact airports
• Water vapor information is a very important
  element, but soundings without it can still be
  useful

10
Ceilings and Visibilities

• The Detroit, MI WFO found TAMDAR data useful in
  forecasting a dense fog event on the evening of
  February 4, 2005
• Soundings indicated light winds in the boundary
  layer, moisture near the surface and dry air
  above
• Normally, these are suitable conditions for the
  formation of low clouds or fog

11
Ceilings and Visibilities
12
Ceilings and Visibilities

• Forecasters at Detroit amended their TAF for the
  09 to 12 UTC period, reducing visibilities to ½
  mile. The METAR observations below show that
  visibilities dropped even lower.
• Kdtw 0532z 00000kt 2sm br clr
• Kdtw 0739z 17003kt 1 3/4sm br r04/1000v3500
• Kdtw 0936z 17004kt 1/4sm fg r04/0500v0600
• Kdtw 1154z 16004kt 1/4sm fg r04/2800v0600

13
Icing and Turbulence

• The U.S. Air Force Weather Agency (AFWA) used
  TAMDAR data to verify an icing forecast from
  their MM5 model.

14
Icing and Turbulence
15
Icing and Turbulence
Is icing occurring over Michigan?
Icing PIREPS for all levels 1600-1800 UTC
16
Icing and Turbulence
Yes! icing near TVC and PLN
Yes! Icing near TVC and PLN
Automated aircraft reports of icing and
turbulence will help AFWA and the AWC in the
forecast and verification of these aviation
hazards.
17
Winter Storms

• Precipitation type forecasts can be made with
  greater accuracy with the use of AMDAR data
• Frequent soundings show the height of the
  freezing level, and the presence of elevated warm
  layers - knowledge of such features is critical
  for accurate precipitation type forecasts

18
Winter Storms (Precipitation Type Forecasts)

• On December 15, 2005 forecasters at the Buffalo,
  NY WFO used AMDAR soundings to identify a
  substantial warm layer between 2,000 and 5,000
  feet above the ground
• Because there was such a deep layer of warm air,
  the forecast was updated to reduce snow
  accumulations and increase the amount of sleet
  and freezing rain

19
Winter Storms (Precipitation Type Forecasts)
20
Winter Storms (Precipitation Type Forecasts)
21
Convective Forecasts

• AMDAR data have been used extensively in support
  of convective forecasts
• Temperature and moisture data provide the needed
  information to determine the stability of the
  atmosphere
• Wind data can be used to construct hodographs

22
A Convective Non-Event
A linear mesoscale convective
system was located over eastern
Minnesota and western Wisconsin at
1500 UTC on June 7, 2005.
23
A Convective Non-Event
It was thought initially that this
elevated convection would become surface based
as the system moved east during the
day A Severe Thunderstorm Watch was
issued for much of Wisconsin at 1530 UTC.
24
A Convective Non-Event

• Shortly after the issuance of the watch,
  forecasters at the Green Bay, WI WFO noticed that
  TAMDAR soundings from the watch area appeared
  much too stable for surface based convection
• A sounding from the Central Wisconsin airport at
  1512 UTC showed a strong capping inversion, that
  was unlikely to break
• Another sounding at 1923 UTC showed the
  atmosphere was still much too stable for
  convection -forecasters then lowered the chance
  for thunderstorms, and the watch was later dropped

25
A Convective Non-Event
1512 UTC Sounding
1923 UTC Sounding
26
A Convective Non-Event

• The thunderstorms that prompted the severe
  thunderstorm watch never became surface based as
  they moved east
• Forecasters had been misled by model soundings
  that predicted the cap would erode sufficiently
  for surface based storms to form
• Misleading forecast guidance became a less
  frequent issue as more forecasters used AMDAR data

27
A Convective Non-Event
RUC Forecast (black) was much less stable than
TAMDAR sounding (purple)
28
Fire Weather

• Accurate fire weather forecasts require an
  understanding of atmospheric stability, wind and
  moisture
• The large spatial and temporal gaps in
  radiosonde data make this difficult
• locations distant from radiosondes are at
  particular disadvantage
• AMDAR data can fill many of these gaps and are
  now used increasingly in fire weather forecasts

29
Fire Weather

• Forecasters found AMDAR data useful in expanding
  a Red Flag Warning in effect for Northern and
  Central Wisconsin on the afternoon of July 15,
  2006.
• Very dry air could be seen on TAMDAR soundings
  earlier in the day when the Red Flag Warning was
  issued. Later soundings indicated sufficient dry
  air in other parts of the forecast area to expand
  the warning
• Temperature gt75F, RH lt25, winds of 25 mph

30
Fire Weather

• AREA FORECAST DISCUSSION...UPDATEDNWS GREEN BAY
  WI200 PM CDT SAT JUL 15 2006
• .UPDATED...ADDED AREAL COVERAGE FOR THE RED FLAG
  HEADLINES. RH/S ALREADY IN THE 20 PCTS EARLY
  THIS AFTERNOON AND WINDS NEAR CRITERIA OVER NW
  WI. TAMDAR SOUNDING THIS AM FROM RHI SHOWS A VERY
  DRY AIR MASS TO MIX DOWN WITH NEAR CRITIERIA
• WINDS.

31
Fire Weather
32
Marine Applications

• Surface winds and resulting waves are influenced
  greatly by both winds aloft and stability
• Unstable air helps mix down strong winds above
  the boundary layer, while stable conditions
  hinder mixing
• AMDAR data are ideal for determining both winds
  aloft and atmospheric stability

33
Marine Applications

• AMDAR data have been used in the Great Lakes
  region and along the coastlines of the United
  States for marine forecasts
• The following is an example from October 3, 2004.
  The Chicago, IL WFO forecasters used AMDAR data
  to determine that the atmosphere was sufficiently
  unstable to mix down 35 knot winds to the surface
• Based on this information, they wisely decided to
  continue the Gale Warning that had been issued
  earlier in the day

34
Marine Applications

• AREA FORECAST DISCUSSION
• NATIONAL WEATHER SERVICE CHICAGO IL
• 715 PM CDT SUN OCT 3 2004
• .MARINE...SHIP REPORTS OF GALES THIS EVENING FROM
  THE
• SOUTHWEST. SURFACE ANALYSIS AND USING THE ACARS
  SOUNDING FROM ORD AT 2230 UTC SHOW STRONG WINDS
  ALOFT AND BEHIND THE FRONT. WILL KEEP THE GALE
  WARNING FOR OVERNIGHT.
• WHW

35
Marine Applications
36
Hazardous Materials Support

• AMDAR data are used frequently by the National
  Transportation Safety Board in the investigation
  of aviation and marine accidents
• AMDAR data were used in support of the September
  11th disaster
• AMDAR data can be used in local models to predict
  the drift of hazardous materials

37
Hazardous Materials Support

• HYSPLIT - Hybrid Single Particle Lagrangian
  Integrated Trajectory model is an Air Resources
  Laboratory (ARL) model used for computing air
  parcel trajectories in dispersion and deposition
  simulations
• HYSPLIT dispersion model simulations were run
  centered over the Miami International Airport
  starting at 01 UTC on the evening of November 27,
  2008
• The model was first run using just the 12km NAM.
  The model was run again, using temperature,
  dewpoint, and wind information from commercial
  aircraft.
• The outputs from the HYSPLIT runs are quite
  different, especially in the first two hours, and
  likely due to the inaccurate model forecast of
  the low level winds by the NAM.

38
Hazardous Materials Support

Aircraft sounding from Miami International Airport
Sounding from an
aircraft equipped with a WVSS2 water vapor sensor
39
Hazardous Materials Support

• HYSPLIT Output

The figure on the left is the output of HYSPLIT
run with the 18Z NAM, while the one on the right
is the same HYSPLIT run with aircraft data.
There are significant differences, especially in
the first hour. The HYSPLIT model forecasts the
pollutant to drift to the northeast from the
start, while the HYSPLIT model with aircraft data
forecasts the pollutant to go in the opposite
direction for the first hour before drifting to
the northeast.
40
Hazardous Materials Support
After the first hour
The first hour of the HYSPLIT run with aircraft
data brings the pollutant in the opposite
direction, over the Fountainbleau and Sweetwater
areas.
The first hour of the HYSPLIT model brings the
pollutant over the Virginia Gardens and Miami
Springs areas.
41
Hazardous Materials Support

• After hour two

The second hour of HYSPLIT run with aircraft data
forecasts the pollutant almost entirely
southwest of the source point.
The second hour of the HYSPLIT model forecasts
the pollutant entirely northeast of the source
location.
42
Hazardous Materials Support
Results

• If this had been an actual event, using model
  data only in the HYSPLIT run could have led to
  the wrong area being warned
• The HYSPLIT run with only model data was probably
  incorrect because the NAM model did not
  accurately forecast the weak sea breeze that
  reached the Miami area during late afternoon
• It would appear that manually entered data from
  aircraft, VAD Wind Profiles or radiosondes can
  result in improvements to the HYSPLIT model
  forecasts

43
Questions?

• Carl Weiss
• National Weather Service, W/OS23
• SSMC2, Rm. 13326
• 1325 East-West Highway
• Silver Spring, MD 20910 USA
• [phone]x149
• carl.weiss_at_noaa.gov