Applications of Aircraft Weather Data

1
Applications of Aircraft Weather Data

• American Meteorological Society Short Course
• San Antonio, Texas January 14, 2007

Applications in Field Forecast Offices
Rich Mamrosh NOAA NWS Green Bay, Wisconsin
2
Applications in Field Forecast Offices

• National Weather Service
• Warning and Forecast Offices
• Center Weather Service Units
• Storm Prediction Center

3
NWS Forecast Office Applications

• Data Comparison (Radiosonde, Radar, Profilers)
• Marine (Wind and Wave Forecasts)
• Fire Weather (Haines Index, Mixing Heights)
• High Wind Events (Santa Ana, Front Range,
  Synoptic)
• Aviation (Ceiling, Visibility, LLWS)
• Winter (Precipitation Type, Lake Effect Snow)
• Convective Storms (Watch and Warning Decisions)

4
Data Comparison
Wind Estimates from Doppler Radar
Wind Profilers
Commercial aircraft
GOES and POES Satellites
Radiosondes
Numerical models
5
Data Comparison
Radiosondes

• The NWS in La Crosse, Wisconsin used TAMDAR to
  verify what looked to be erroneous radiosonde
  data.
• The dewpoint from the Minneapolis, Minnesota
  radiosonde seemed way too low.
• TAMDAR data showed that the radiosonde
  information was indeed correct. This was very
  important in determining what precipitation type
  was likely later that day.

6
Data Comparison
Radiosondes
AREA FORECAST DISCUSSIONNWS LA CROSSE WIAFDARX
401 AM CST MON NOV 1 2004 .SHORT TERM... THE
NOVEMBER 1ST 00Z 850 MB DEW POINT OF -36C AT
CHANHASSEN LOOKED INITIALLY SUSPICIOUS WHEN
LOOKING AT TRAJECTORIES. HOWEVER...A MESABA
AIRCRAFT WITH THE NEW TAMDAR SENSOR AT 04Z HAD A
DEW POINT OF -32C IN THE CHANHASSEN AREA. THE
ACCURACY OF THIS DEW POINT IS RATHER CRITICAL
BECAUSE IT AFFECTS THE INITIAL ONSET OF
PRECIPITATION
7
Data Comparison
Model Forecasts
8
Data Comparison
Model Forecasts

• Senior duty meteorologists at NCEP often use
  AMDAR flight level data to determine the validity
  of radiosonde winds and temperatures.
  Meteorologists in WFOs and CWSUs compare AMDAR to
  model initial analyses and forecasts.
• The Lousiville, Kentucky NWS used AMDAR on May
  17, 2005 to validate their suspicions that the
  NAM model dewpoints in the low levels were
  erroneous.
• The boundary layer moisture was important in the
  fog forecast that night.

9
Data Comparison
Model Forecasts
AREA FORECAST DISCUSSIONNWS LOUISVILLE KYAFDLMK
940 PM EDT TUE MAY 17 2005 .UPDATE... HAVE
NOTED VERY HIGH DEW POINT ERRORS IN THE RECENT
RUNS OF THE ETA (NAM) MODEL. LOOKS AS THOUGH THE
NAM HAD WAY UNDER-FORECAST THE MAGNITUDE OF THE
DRY POCKET WITHIN THE DAYTIME BOUNDARY LAYER
STRETCHING FROM THE OHIO VALLEY INTO THE
MID-SOUTH THIS EVENING. AN ACARS DESCENT SOUNDING
INTO EVANSVILLE AROUND 1Z SHOWED 950MB DEWPOINTS
OF 4C...THE ETA HAD FORECAST 10 WITH THE GFS
MUCH CLOSER AT 5C. WITH THAT SAID...OUR DEW
POINT FORECAST WAS ON TARGET.
10
Marine Forecasting
11
Marine Forecasting

• A dense network of wind and temperature soundings
  are important to the production of accurate wind
  and waves forecasts.
• Temperature profiles are important to determine
  stability (which affects mixing down of stronger
  winds aloft).
• Wind profiles are needed to determine directional
  and vertical shear (which also affect transport
  of stronger winds aloft).

12
Marine Forecasting
A dense network of radiosondes is not available
near the oceans and Great Lakes.
13
Marine Forecasting

• AMDAR observations have provided meteorologists
  with valuable supplemental data near the oceans
  and Great Lakes.
• Forecasters have used AMDAR to issue, continue
  and discontinue Small Craft Advisories, Gale
  Warnings and Storm Warnings.

14
Marine Forecasting
Gale Warning

• The NWS in Grand Rapids, Michigan used AMDAR data
  in deciding to continue a Gale Warning for the
  nearshore waters of eastern Lake Michigan on
  November 6, 2005.
• TAMDAR soundings from Kalamazoo, Michigan showed
  that there still were winds at nearly 50 knots in
  the lowest 3,000 of the atmosphere.
• In addition, the air was unstable, and the winds
  had little directional shear both favorable for
  the transfer of strong winds to the surface.

15
Marine Forecasting
Gale Warning
GALE WARNING A warning of sustained surface
winds, or frequent gusts, in the range of 34
knots (39 mph) to 47 knots (54 mph) inclusive,
either predicted or occurring, and not directly
associated with a tropical cyclone.
Map of NWS Grand Rapids forecast area. Nearshore
marine area is in purple.
16
Marine Forecasting
Gale Warning
AMDAR data in Great Lakes region from 1400 to
1659 UTC on November 6, 2005
17
Marine Forecasting
Gale Warning
TAMDAR Sounding From Kalamazoo Showing West
winds At 46 knots About 2,200 AGL.
18
Marine Forecasting
Gale Warning
AREA FORECAST DISCUSSIONNWS GRAND RAPIDS
MIAFDGRR 1032 AM EST SUN NOV 6 2005 5 MINUTE
WIND DATA AT THE GLERL COAST SITES SHOWS
THE WINDS HAVE PEAKED AT MKG...LWA AND MICHIGAN
CITY IN THE PAST HOUR NEAR 45 KNOTS. THE TREND IS
NOW DOWN SLOWLY ON THE WIND SPEED. TAMDAR DATA
STILL SHOWS WINDS UNDER 5000 FT OF JUST UNDER 50
KNOTS OVER THE WESTERN CWA AT 10 AM. AS A RESULT
OF ALL OF THAT I PLAN ON LEAVING THE GALE WARNING
AND THE WIND ADVISORY AS IS INTO THE MID
AFTERNOON.
19
Marine Forecasting
Lake and Sea Breezes

• AMDAR can provide the information to determine
  the likelihood of lake and sea breeze formation.
• Lake and sea breezes are hindered by an offshore
  gradient wind, and enhanced by an onshore wind.
  Stability also plays a role.

20
Marine Forecasting
Lake and Sea Breezes

• Meteorologists at the NWS in Cleveland used AMDAR
  data on April 16, 2005 to forecast the formation
  and inland penetration of a lake breeze from Lake
  Erie.
• AMDAR soundings showed that there were
  significant south (offshore) winds that would
  slow the formation of the lake breeze, and limit
  its inland movement.
• In addition to its importance for mariners, this
  wind forecast also affects the large hub airport
  at Cleveland.

21
Marine Forecasting Lake and Sea Breezes
AMDAR sounding from 1244 UTC in the vicinity of
Cleveland
22
Marine Forecasting Lake and Sea Breeze
AREA FORECAST DISCUSSIONNWS CLEVELAND OHIO1159
AM EDT SAT APR 16 2005 .UPDATEDSOUTH WINDS
WERE INCREASING AND TEMPERATURES WERE RISING WITH
FULL SUNSHINE. HIGHS SHOULD RANGE FROM 65 TO 70
DEGREES. THE LAKE BREEZE, IF IT DOES DEVELOP,
SHOULD BE RIGHT ALONG THE LAKESHORE FROM NEAR
CLEVELAND THROUGH ERIE. AMDAR DATA NEAR CLEVELAND
WAS INDCATING THE WINDS FROM SOUTH AROUND 25
KNOTS AT 2,000 FEET.
23
Fire Weather
24
Fire Weather

• AMDAR data are becoming increasingly utilized by
  fire weather forecasters, as coverage expands
  across the U.S.
• Useful fire weather forecasts are dependent on
  accurate observations and/or model forecasts of
  mixing layer heights, Haines indices, and surface
  relative humidity.

25
Fire Weather
Smoke Dispersion

• Smoke from prescribed burns and wildfires can
  cause serious problems for those with respiratory
  problems. It also can also cause accidents by
  reducing visibilities on highways.
• The NWS produces a Smoke Dispersion index based
  mostly on model forecast soundings. These
  forecasts are used by fire weather agencies to
  determine whether prescribed burns should be
  allowed, and if wildfires might produce hazardous
  smoke conditions.

26
Fire Weather
Smoke Dispersion

• The Miami NWS used AMDAR to update smoke
  dispersion forecasts on May 2, 2005.
• ACARS soundings showed very light winds (3-10
  kts) in the mixed layer, which inhibits smoke
  dispersion.

27
Fire Weather
Smoke Dispersion
ACARS sounding From 1411 UTC On May 2, 2005 In
the vicinity of Miami, Florida Shows very light
Winds ( 28
Fire Weather
Smoke Dispersion
AREA FORECAST DISCUSSIONNWS MIAMI FL1030 AM EDT
MON MAY 2 2005 .FIRE...WILL AMEND GRIDS AND FWF.
WINDS IN MIXED LAYER ARE FAIRLY LIGHT, WHICH IS
CONFIRMED BY ACARS DATA. THIS WILL RESULT IN
ADJUSTMENTS TO TODAYS DISPERSION INDICES.
29
Fire Weather
AMDAR inspired systems

• The utility of AMDAR in fire weather forecasting
  fostered the development of locally driven
  efforts.
• An experiment called HI-RISE (Hazardous Incident
  - Rapid In-flight Support Effort) was conducted
  in the spring of 2005 by the following
  participants
• National Weather Service
• Texas Forest Service
• USDA Agricultural Station
• Aventech Research Inc.

30
Fire Weather
AMDAR inspired systems
HI-RISE involved the installation of
meteorological instruments on a small
fire surveillance aircraft, and downlinking the
data in real time to meteorologists.
AIMMS-20 instrumentation
Air Tractor AT-402B aircraft used in HI-RISE
Data flow in HI-RISE experiment
D
31
Fire Weather
AMDAR inspired systems
Non-commercial aircraft such as these can fly
REALLY close to fires and measure meteorological
parameters vital in forecasting fire behavior
Air Tractor AT-402B aircraft flying over Texas
Forest Service ground crew during HI-RISE
experiment
32
Fire Weather
AMDAR inspired systems
Example of aircraft sounding from HI-RISE fire
weather experiment. Notice highly detailed
temperature, moisture and wind data.
33
Fire Weather
AMDAR inspired systems

• This proof of concept experiment provided
  valuable real time data to fire weather
  forecasters at controlled burns in central Texas.
• All of the participants agreed that this concept
  should be developed further for fire weather and
  other applications, including HAZMAT support.

HI-RISE Team from NWS, USDA, Texas Forest Service
and AVENTEC
34
High Winds
35
High Winds

• Non - convective high wind events occur over the
  entire country.
• Meteorologists largely rely on model forecasts to
  issue high wind watches, and a combination of
  observations and model data to issue warnings.
• The lack of real time upper air data can make
  warning decisions difficult.

36
High Winds

• Downslope high wind events occur with some
  regularity in the Rocky Mountain region.
• Santa Anna winds in California and strong winds
  along the front range in Colorado are two well
  known phenomena that are forecast challenges.

37
High Winds
Schematic of Santa Ana winds in southern
California
38
High Winds
Santa Ana

• A Santa Ana wind event occurred the night of
  December 22-23, 2006 in southern California.
• Meteorologists at the NWS in San Diego monitored
  AMDAR soundings during the early morning hours,
  and decided to upgrade a wind advisory to a high
  wind warning based on AMDAR soundings.

39
High Winds
Santa Ana
AREA FORECAST DISCUSSIONNWS SAN DIEGO CA1120 PM
PST FRI DEC 22 2006 .DISCUSSION... UPDATE TO
UPGRADE THE WIND ADVISORY FOR THE INLAND EMPIRE
AND THE SANTA ANA MOUNTAINS AND FOOTHILLS TO A
HIGH WIND WARNING THROUGH EARLY AFTERNOON
SATURDAY. RECENT ACARS SOUNDINGS FROM ONTARIO
SHOW A VERY IMPRESSIVE SUBSIDENCE INVERSION NEAR
700 MB FAVORABLE FOR MOUNTAIN WAVE DEVELOPMENT AS
WELL AS AN OVERALL STRONG WIND FIELD BELOW 700 MB
WITH WINDS UP TO 50 KNOTS OR SO. WINDS IN THE
WINDIER LOCATIONS...BELOW THE CAJON PASS AND NEAR
FREMONT CANYON HAVE GUSTED TO AROUND 60 MPH IN
THE PAST HOUR.
40
High Winds
Santa Ana
AMDAR sounding at 0411 UTC December 23, 2006 from
Ontario, California showing mountain top
inversion and strong wind field.
41
High Winds
New York City

• A High Wind event driven by an intense cyclone
  occurred in the Northeast on November 13, 2003.
• The NWS in Upton (Long Island), New York used
  AMDAR data to monitor winds aloft in real time,
  and made warning decisions based on these data.
• The upstream data at the New York area airports
  was especially valuable.

42
High Winds
New York City
Surface weather map from November 13, 2003 shows
strong cyclone over southern Quebec
43
High Winds
New York City

• AREA FORECAST DISCUSSION
• NATIONAL WEATHER SERVICE UPTON NY
• 1000 AM EST THU NOV 13 2003 .
• DISCUSSION... .UPDATE...SEVERAL REPORTS OF DOWNED
  TREES/DAMAGE AND WINDS AS HIGH AS 52 KT AT
  AMBROSE LIGHT. HIGHEST LAND BASED WIND IS 48 KT
  AT LGA. EXAMINATION OF 12Z ACARS WINDS REPORTS
  SHOWS THAT 09Z RUC WINDS ARE TOO LOW AND THAT
  EVEN 06Z ETA WINDS ARE FEW KNOTS BELOW OBSERVED.
  AS COLD ADVECTION WORKS IN THIS
  AFTERNOON...HIGHER LEVEL WINDS WILL WORK DOWN TO
  SURFACE. USING A 90 RULE OF WINDS
  ALOFT...RESULTS IN PEAK SURFACE OF 60 KT OR 70
  MPH. WILL HEADLINE...WIND GUSTS TO NEAR HURRICANE
  FORCE.

44
High Winds
New York City
12UTC sounding From Brookhaven New York shows 60
knot winds Within 2,000 AGL.
45
High Winds
New York City
1243 UTC AMDAR sounding from JFK airport shows
51 knot winds around 2400 AGL.
46
High Winds
New York City
1433 UTC AMDAR sounding from Newark, NJ shows 63
knot winds around 3000 AGL and an almost
dry, adiabatic lapse rate.
47
High Winds
North Dakota
AREA FORECAST DISCUSSIONNWS GRAND FORKS ND340
AM CST SUN DEC 17 2006 .SHORT TERM
(SUN-TUE)...AMDAR SOUNDINGS AT 05Z FOR DVL/JMS
SHOWED THAT MORE COLD ADVECTION HAD OCCURRED AT
850 MB THAN MODELS ANTICIPATED WHILE LOW LEVEL
WINDS WERE STRONGER TOO. THE COLDER CONDITIONS
HAVE KEPT THE BOUNDARY LAYER MORE MIXED AND WINDS
ARE ABLE TO MIX THROUGH A BIT EASIER. JMS SHOWED
70 KNOTS AT 801 MB (JUST ABOVE THE INVERSION) AND
DVL SHOWED 50 AT 911 MB. THESE WERE 10 TO 25
KNOTS HIGHER THAN MODEL SOUNDINGS.
48
High Winds
North Dakota
TAMDAR sounding from 0509 UTC in the vicinity
of Jamestown, North Dakota on December 17, 2006
shows 70 knot winds at 801 mb.
49
Aviation Forecasting
50
Aviation Forecasting
Ceiling and Visibility

• Meteorologists at the NWS San Diego office are
  one of the biggest users of AMDAR data in the
  United States. The scarcity of upstream data,
  varied terrain, and a maritime influence make the
  availability of frequent upper air data very
  useful.
• The data are especially useful to monitor the
  depth of the marine layer, upon which the
  formation and inland penetration of low clouds
  and fog are dependant upon.

51
Aviation Forecasting
Ceiling and Visibility
52
Aviation Forecasting
Ceiling and Visibility
The San Diego Forecast Office is responsible for
the TAFs for Ontario (ONT), John Wayne (SNA), San
Diego (SNA), as well as Palm Springs, Thermal,
and Carlsbad (Not shown)
53
Aviation Forecasts

• Ceiling and Visibility

TAFONT
KONT
302327Z 010024 26015KT 5SM HZ SCT250
FM0400 25005KT 5SM HZ
SKC

SKC
FM0600 VRB03KT 5SM BR

SKC
FM1300 VRB03KT 2SM BR HZ
SKC
FM1900 26009KT 4SM HZ
FM2200 26015KT P6SM
SKC



The TAF for Ontario issued at 2325 UTC on April
30, 2001 mentioned that visibilities might fall
to 2 miles late at night, and that skies would be
clear. It was thought that the depth of the
inversion would be insufficient to climb the
range of hills, and reach the inland valley area.
54
Aviation Forecasts
Ceiling and Visibility
ACARS from Long Beach, California at 0232 UTC on
May 1, 2001 shows the base of a temperature
inversion around 1,000
55
Aviation Forecasts
Ceiling and Visibility
ACARS from San Diego, California at 0233 UTC on
May 1, 2001 shows the base of a temperature
inversion around 1,200.
56
Aviation Forecasts
Ceiling and Visibility
ACARS WVSS from Ontario, California at 0201 UTC
on May 1, 2001 shows the base of a temperature
inversion around 1,900.
57
Aviation Forecasts

• Ceiling and Visibility

TAFONT
KONT 010525Z 010606 VRB03KT 5SM BR SKC
FM1200 VRB03KT 2SM BR OVC008
TEMPO 1215 1SM BR OVC005
FM1600 VRB03KT 3SM BR HZ SKC
FM2000 26012KT 5SM HZ SKC
FM0500 VRB03KT 4SM BR OVC010
The meteorologist noticed that soundings from the
evening hours showed that the marine layer was
about 1,000 to 1,200 feet along the coast, and
1,900 feet MSL at Ontario. This suggested that
the damp, maritime air had already reached the
interior valley, and that low clouds and fog were
indeed possible. This information was used to
issue the 06UTC TAF, forecasting IFR ceilings of
5 hundred feet and visibilities around one mile
were expected overnight.
58
Aviation Forecasts
Low Level Wind Shear
AREA FORECAST DISCUSSION...UPDATED NWS FORT WORTH
TX 1055 AM CDT TUE MAY 4 2004 .AVIATION... 604
AM. VFR...LLWS THRU 14Z THEN NO CONCERNS. SW
WINDS AT 1-2 KFT HAVE SET UP A STRONG
INVERSION...AND A LLJ OF ARND 23035KT PER ACARS
SOUNDINGS/VWP. WITH LT 10 KTS AT THE SFC...I WILL
MENTION LLWS THRU 14Z. WITH QUICKLY WARMING
TEMPERATURES UNDER PLENTY OF SUN...ANY LLWS SHLD
BE GONE BY MID MORNING WITH GUSTY S/SW WINDS. SLY
WINDS 5-15 KTS TO PREVAIL TONIGHT. 30/SS
59
Winter Weather
60
Winter Weather

• Precipitation type remains one of the main
  challenges of winter weather forecasting.
• Accurate forecasts require a detailed knowledge
  of the temperature and humidity structure of the
  atmosphere.

61
Winter Weather
Northern Indiana

• Meteorologists at the NWS offices in Warsaw,
  Indiana used TAMDAR during the evening of
  December 25, 2006 to determine that cloud
  microphysics were not suitable for heavy
  snowfall.
• TAMDAR soundings into Fort Wayne showed that
  temperatures were too warm for significant ice
  crystal growth.
• They updated their forecast, to reduce (in half)
  the snow accumulation amounts.

62
Winter Weather
Northern Indiana
AREA FORECAST DISCUSSIONNWS NORTHERN INDIANA935
PM EST MON DEC 25 2006 .UPDATE... 2355 UTC
TAMDAR DESCENT SNDG INTO KFWA CONFIRMS TEMPS
THROUGH 700MB TO BE BARELY SUFFICIENT FOR
NUCLEATION WITH TEMPS AOA M7C. OUTSIDE OF
RELATIVELY WEAK AND POORLY DEFINED TRANSITORY
BANDING THAT BRINGS A MIX TO BRIEF BURST OF 2-3SM
-SN CHANGEOVER...THERE CONTINUES TO BE NO ACCUMS
GIVEN 2 INCH SOIL TEMP OF 38F AND MAX SUBCLOUD T
OF 2-4C AND WET GROUND MELTING SNOW AS SOON AS IT
FALLS. MOST OF CWA PLAGUED WITH SATURATION
BENEATH PRIME NUCLEATION HEIGHT AS SUCH HAVE
UPDATED AND BASICALLY HALVED ACCUMS OUTSIDE OF
FAR W/NW WITH MENTION OF TRACE MIXED EVENT
63
Winter Weather
Northern Indiana
TAMDAR sounding from 0027 UTC on December
25, 2006 shows a deep, moist layer that
is generally too warm for ice crystal Formation.
64
Convective Storms
Supercell thunderstorm near Joplin, Missouri
65
Convective Storms

• AMDAR data has been shown to be very effective in
  the issuance of convective watches and warnings.
• Real time data has been very valuable in
  determining the presence of capping inversions,
  wind shear, low level moisture, etc.

66
Convective Storms Warning Decisions

• The NWS in La Crosse, Wisconsin used TAMDAR in
  deciding NOT to issue a tornado warning despite
  a tornado signature on radar.
• This late season event in October, 2006 shows the
  utility of AMDAR in the proper depiction of the
  near-storm environment, which is vital to
  accurate convective warnings.

67
Convective Storms Warning Decisions
Surface warm front expected to be the focus for
surface based convection during the afternoon
hours. Given the high cape and high shear
environment the area was in a Moderate Risk for
severe weather with a heightened threat for
tornadoes.
68
2000 UTC SPC Day 1 Outlook

• Moderate Risk highlighted area of high shear,
  instability and moisture pooling along a frontal
  boundary.

69
Convective Storms Warning Decisions

• MCD indicated storms expected to develop in the
  outlook area with increasing CAPE just north of a
  warm front.
• Although a capping inversion was in place this
  was expected to weaken during the afternoon
  hours.
• Strong deep layer shear expected to promote
  supercells with possible tornadoes.

• Tornado Watch was issued at 2035 UTC.

70
Convective Storms Warning Decisions
Map of NWS La Crosse, Wisconsin forecast area
(shaded)
71
Convective Storms Warning Decisions
Large Supercell over the Minneapolis forecast
area.
Multiple tornado warnings were issued for the
storm as is moves into western Wisconsin.

• Warning decision needs to be made around 0100 UTC
  for ARX

72
Convective Storms Warning Decisions

• EAU sounding indicated very dry air at the
  surface and a strong CAP.

• MSP soundings shows a stronger CAP to the south
  but a strong and persistent cap in all directions.

• LSE soundings show a strengthening CAP and dry
  air at the surface.

• A subsequent MSP sounding showed a stronger CAP
  and deeper stable layer than the RUC.

73
Convective Storms Warning Decisions
Aircraft soundings gave forecasters confidence in
issuing Severe Thunderstorm Warnings and holding
off on Tornado Warnings. The Supercell
subsequently weakened as it moved through the La
Crosse forecast atea.
74
Summary

• The use of AMDAR has increased in WFOs, CWSUs,
  and NCEP centers in the past decade.
• Applications of AMDAR have been found for almost
  all weather phenomena.
• Need dense coverage of observations that can be
  provided by regional airlines.
• Need more water vapor soundings.

75
Acknowledgements

• Gary Austin - NWS Green Bay MIC
• Pete Browning - NWS Central Region SSD
• Dave Helms - NWS Office of Science and Technology