DENVER ARTCC EVALUATION OF PROFS
MESOSCALE WEATHER PRODUCTS
John W. Hinkelman
The Denver Air Traffic Control Center has had a Prototype Regional Observa-
tion and Forecast System (PROFS) display system since the Spring of 1982. At the
request of the Chief of the Denver Center, I will give an updated presentation of
the evaluation the Center has made of the PROFS products. A full PROFS work-
station capability was installed at the Denver Center in June 1984, and information
concerning this can also be obtained through PROFS.
A PROFS display capability was requested by the Denver Center through the
FAA Regional Office in 1981. The FAA was concerned with the safety and economic
impact of adverse weather in the Denver air traffic control (ATC) environment
(Figure 1). Stapleton Airport is a particularly difficult terminal from a weather
standpoint because it has a tendency during bad weather to adversely affect the
Denver Center’s operation, and this frequently affects transcontinental traffic as
well. Also Stapleton is a big hub operation for three airlines.
There are four significant and typical weather situations. One that occurs ev-
eryday involves wind shifts which force runway changes, and occur quite frequently
during heavy traffic hours (i.e., noon and the evening rush hour). Severe thunder-
storms, severe mountain wave turbulence, and upslope storms are the other three.
We have had three significant upslope cases since January 1, 1985, which severely
disrupted operations in the Denver Center area. The FAA real operational require-
ment is for timely detection and prediction (0-2 hours for safety, and 0-6 hours for
Figure 2 shows a hard copy printout of display data provided to the Denver
Center since the Spring of 1982 when the first system was installed. The circle is the
TRACON area, the isolines are weather radar reflectivity data (20,40,60 dBZ), with
Stapleton in the center. On this particular day, there was a level-6 thunderstorm
right over the outer marker. We are displaying flow lines, and the Limon Radar is
shown along with mesonet temperatures, humidity, and wind information.
During the 30 months of the initial display evaluation, the system operated 24
hours a day, seven days a week (Figure 3) . There were 150 significant cases reviewed.
Of those 150 cases, 38 were analyzed in detail, 25 of which were reported on to the
FAA. Of the 38 representative events analyzed, 14 were thunderstorms, 12 were wind
shear/wind shifts, 7 were upslope cases, and 5 were combined thunderstorm/upslope
cases. There was a positive impact on operations in 33 of the cases based on the
Adverse Weather Seriously Impacts:
• Both safety and cost operations in Denver ARTCC area
• Disrupts Denver Stapleton Airport operations
• Frequently affects transcontinental traffic flow
Four Significant Weather Situations:
0 Approach area wind shear/abrupt runway wind shifts
0 Severe thunderstorms
0 Upslope storms— widespread low visibility/precipitation
Timely (0-6 hour) Detection and Prediction Can:
0 Improve safety
0 Decrease delay costs
Figure 1. Operational requirement.
A Approach Fixes 40 dBz —
O NAVA1DS 60 dBz —
Typical PROFS mesoscale graphic product covering Denver
Stapleton area. Limon radar is in circle at bottom right.
• 30 Months— Daily use supporting Stapleton traffic operations
I Over 150 significant cases analyzed
I 38 serious events chosen for analysis strong operational
I Representative Events
t 14 Thunderstorms
• 12 Wind Shear/Wind Shifts
• 7 Upslopes
• 5 Combined Thunderstorm/Upslope
9 Positive Impact— 33 Cases
• 12 Thunderstorms
• 5 Terminal Wind Shears
• 6 Wind Shifts
• 6 Upslopes
• 4 Combined
Figure 3. Real-time weather impact.
I Conventional Radar Cell Tracking very valuable in predicting
arrival gate and terminal operations restrictions.
I Automated Surface Observations (mesonet) critical for low
ceiling and visibility onset/cessation.
• Automated Radar and Surface Data combines/overlays -extremely
valuable determining cell development and tracking.*
• Profiler Winds effective for flight path prediction,
forecasting upslope, thunderstorm cell tracks.
• Doppler Radar needed to better forecast airport wind shift
timing and locate low-level wind shear areas.
f ATC confidence level increases with timely accurate tailored
Figure 4. Product utility.
Case file sequence begins at 17:40:58 (1141L). The synoptic situation had
been analyzed early in the period and a strong frontal passage was inevitable.
Timing of the event remained Questionable due to sparse conventional data north
and east of the Denver Stapleton Airport in northeastern Colorado, western
Nebraska and southern Wyoming. The Denver ARTCC meteorologist began surveilling
the northern portions of the mesonet disday during early morning hours and
first detected a FROPA at about 1730Z. This information was used to provide a
forecast to Denver tower and ARTCC traffic management that the front should
reach Stapleton between 2030Z and 2100Z with resultant wind shift and possible
reduced visibility due to blowing dust. No precipitation was forecast for the
frontal passage due to low moisture availability. At 1935Z the CWSU meteorolo-
gist issued a center weather advisory (see below) based on 180° wind shear at
33 kn between LGM and LVE on the mesonet per FROPA at Denver at 2030Z.
Denver ARTCC and Denver tower personnel began to formulate traffic flow
restrictions and runway configuration plans immediately and traffic speed
reductions as well as expanded quota flow from CF^ after the 2000Z hour to
avoid sector saturation and excessive airborne delays.
The front passed Denver at 2027Z, within 3 minutes of forecast with
visibility reduction to 1 nm. Some airborne delays were experienced for a
period of time due to single runway operations but traffic flow was maintained
smoothly throughout and system perturbations were effectively Minimized.
The meteorologist knew it would be a strong frontal passage from synoptic
data. Where mesonet really helped was with timing, and also the amount of shear.
Apparently the front was first detected in NRN portions. of mesonet around
1730Z and an initial FROPA estimate for DEN was made.
At 1930Z, what really prompted the CWSU to issue .the CWA was the 197 52
mesonet - - notice the 180° shear at 33 kn between LGM and LVE.
Front passed Den at 2027Z, 3 minutes before fcstd In. CWA.
w2 cwa,,,1Q35z mon sep 19 1983.., an extremely strong frontal
BOUNDARY WITH WIND SHEAR TO 70-80 kn CURRENTLY MOVG THRU NRN PTNS
DEN TRACON AREA.,. WINDS WITH FRONTAL PASSAGE SHIFT TO N-NE AT 3Q-40
kn FRONT EXPCD BY DEN STAPLETON AT 2030z. ,. CAUTION FOR EXTRM LO
IV L TURBO AND BLWG DUST, 7.DV CWSU 191935Z
Description from Denver Stapleton, September 19,
(Transcribed from computer printout)
17 : 40:58
NU8 LXNON RADAR
(UP) 30 dBZ
HPL PROFILER DATA
Figure 6. PROFS Experimental Data Display, September 19, 1983,
19 : 17:46
NW8 LI RON RADAR
(UP) 38 dBZ
HPL PROFILER DATA
Figure 7. PROFS Experimental Data Display, September 19, 1983,
at 19: 17:46Z.
19 : 27:47
HUS LIKGII RADAR
<UP> 30 dBZ
WPL PROFILER DATA
r««t Oaaraas Knots
PROFS Experimental Data Display, September 19, 1983,
at 19 : 27 : 47Z .
20 : 19:0?
HUS L IKON RADAR
1 9-SEP- 1983 20:15:00
<UP> 30 dBZ
WPL PROFILER DATA
Faat Da«raas Knots
PROFS Experimental Data Display, September 19, 1983,
at 20: 19:07Z.
analyses performed by the CWSU and ATC people at the center.
From a product utility standpoint (Figure 4), conventional radar cell tracking
was extremely valuable for predicting arrival gate and terminal operations restric-
tions, particularly during rush hours. Automated surface observations (mesonet)
were critical for low ceiling and visibility onset and cessation. The automated radar
and surface data combinations were the most valuable for predicting cell devel-
opment and track information. Profiler winds were quite effective for flight path
prediction, input to the ATC 90/20 computer forecasting upslope conditions, and
thunderstorm cell tracking. The center would like to have Doppler radar informa-
tion for wind shear forecasting, etc. We do not plan to provide that capability for
several years. The ATC Weather Information confidence level has increased steadily
with the timely PROFS information.
Figure 5 is a case description for September 19, 1983, and Figures 6-9 are
actual hard copies of the information used in the control room. On each display it is
noted that the data are PROFS experimental data. What I would like to show you
in these figures is how accurately we were able to predict wind shifts which caused
runway changes. Figure 6 shows that at approximately 17:40 Z, which is about
11:40 a.m., a front was moving down from the north. You will note that although
the Limon radar was operative, no echoes were showing from the radar. Within one
hour and 15 minutes the frontal system moved down into and across the TRACON
area. At 19:35 Z, the frontal passage was forecast for Stapleton at 20:30 Z (Figure
7). Figure 8 shows that at 19:27, the system is continuing to move; and at 20:19,
Figure 9 shows that there has not been a shift at Stapleton, but there has been a
wind shift just to the north at Brighton. This was predicted roughly 55 minutes in
advance. The CWSU has been forecasting this type of wind shift consistently for
the past two years.
We are currently in the Denver Center Phase II Program (Figure 10), which
involves a full PROFS work station providing radar, mesonet, satellite and AFOS
products. We provide time series analyses of each of the mesonet stations, and are
depicting all the profiler stations’ data. We actually have four profiler stations in
operation now, and the network is being reconfigured with additional stations. We
are not providing Doppler radar coverage now, as mentioned earlier; however, we
may provide some output information from CP-2 this summer. We are providing
automated PIREP information to the center. In summary, we have been focusing
on display information for the CWSU work station, output products to the ATC
system, and developing functional specifications for the FAA CWSU work station
of the future.
Question from the floor: Mike Tomlinson, NWS
Full Meteorological Work Station
• Interactive displays
• Touch-screen menus
• High resolution color
t Overlaying, looping, zooming
Expanded Data Sets
• Doppler Radar coverage
• 5-station profiler network
• Automated PIREPs
• AFOS product overlay capability
• Output products to ATC system
t Work station display
• Functional specifications
Figure 10. Current Phase II Program.
Everything you have shown seems to concentrate almost exclusively in the
Denver Stapleton area, and the Denver ARTCC area is considerably larger than
that. Do you have similar types of capabilities to cover the rest of the area; and, if
not, what kind of impact is this concentration of that one terminal having on the
services provided for the remainder of the Denver ARTCC area?
Response: Jack Hinkelman
The Denver Center area may be unique because Stapleton is the primary throt-
tling feature in the Denver Center area. It is the big terminal. Of course, there
are Colorado Springs, Cheyenne, Pueblo, Grand Junction and other terminals; but
Stapleton is the sixth or seventh most congested airport in the world. It is about
the fifth in the U. S., and is a big problem area. They are normally operating at
approximately 80-90 aircraft per hour almost continuously from 7:00 a.m. to 8:00
p.m. They are always operating at maximum capacity. Whenever there is a weather
problem at Stapleton, both the en route system and the terminal system in that
area becomes very unstable, and may stay that way for about four to five hours
even after the weather dissipates. Therefore, we have concentrated on the Stapleton
area. Although I have not shown any cases, we have several where thunderstorm
track is predicted over the arrival and departure gates. Also, we are able to predict
thunderstorm tracks out over the en route area, particularly to the east. It is more
difficult out over the west. Most of the thunderstorms form in the mountains.
We have intentionally, at their request, concentrated on the Stapleton opera-
tion. It very frequently affects transcontinental operations. There is also profiler
data which covers the whole center area, and that has been put into the 90/20.
When the forecast winds appear to have been in error, we have been able to insert
real-time profiler data in the 90/20 and the ATC computers settle down. We do
not have radar data that covers the entire center area, so we can work only with
what we have.
Question from the floor: Doug Lundgren, AOPA
What do you see in a broader, national scale for the future of a PROFS-type
effort, particularly the mesonet? We can place sensors around a particular airport;
but how can we correlate this with more sensors nationwide?
Response: Jack Hinkelman
As you know, PROFS is an experimental prototype of the AWIPS-90 program,
which is the NWS program for implementation in the 1990’s. In fact, there are
mesonetworks around almost all major metropolitan areas in the country. EPA
and other groups all have mesonets. I think a good part of the state of Tennessee
is covered by automated surface observations, like the Tennessee Valley Authority
(TVA). There is almost a full mesoscale network covering the state of Tennessee.
There are four or five groups and none are reporting into a central computer. I
think Sandy MacDonald could verify this, but we believe there is extreme value
in mesonetworks. We don’t see any national program to implement mesonetworks
around major cities or airports, but it certainly would not be a bad idea. Maybe it
will come about.