An Assessment of Operational GOES Sounder Products

1
An Assessment of Operational GOES Sounder
Products
By James Hocker
Undergraduate, University of Oklahoma Norman, OK
2
A special thanks to Bill for giving me the
opportunity to volunteer at the office this
summer. Also thanks to Greg for overseeing my
project and providing me with an aggressive
activity plan.
My short stay at the office has taught me that

• Experience teaches much more than the books

• Food is the heartbeat of this office (this is
  especially true for several forecasters)

• Sooners and Aggies can actually get along pretty
  well together!!

3
Outline of Topics

• Brief overview of the GOES Sounders

• Products available

• Configuration in AWIPS

• Case Study July 23, 2003 North Texas localized
  severe weather event

• 2-week statistical study

• Results of other studies

• Summary

• Resources available

4
Overview of the GOES sounders

• The GOES-8/9 sounders are a new generation of
  geostationary sounders that replaced the VAS
  (Visible and Infrared Spin Scan Radiometer
  Atmospheric Sounders) in 1994.1

• The sounders

• provide hourly observations over large regions

• fill in RAOB location gaps 2000 to 3000
  retrievals made each hour at 50 km spacing

• illustrate short term changes in moisture and
  temperature profiles useful in severe weather
  nowcasting

• depict wind profiles by analyzing thermal
  gradients and moisture transport.2

5
Overview of the GOES sounders

• The sounders utilize 19 spectral bands (18 IR, 1
  vis) that measure varying earth-emitted
  radiances. Temperature and moisture profiles can
  be inferred from this data.1

Figure 1. GOES spectral band centers (and
widths) with respect to the earth-emitted
spectrum. The bands (lower portion of figure)
are displayed at their respective levels of the
atmosphere. Temperatures are inferred from CO2
data and moisture is inferred from H2O data.2
6
Overview of the GOES sounders

• In order to produce moisture and temperature
  profiles, the GOES sounding process includes

• determining if the field of view (FOV) is clear
  or cloudy

• making a first guess using the AVN 6-18 hour
  forecast (e.g. 12Z run used for Day 1 18Z through
  Day 2 06Z) in conjunction with surface
  observations

• averaging radiances within each clear 5 x 5 FOV,
  assigning a retrieval location, and applying a
  bias adjustment

• comparing measured radiances to calculated
  radiances (model) if deviations are within
  threshold, products are created (otherwise no
  products are created)3

The soundings are highly dependent on the model
first guess!!
7
Overview of the GOES sounders

• Overall, the GOES soundings are different from
  RAOB in that

• GOES soundings depict temperature and moisture
  profiles as layer averages, not level specific
  values

• GOES soundings smooth out temperature and
  moisture features, but capture the mean vertical
  profile very well

• GOES soundings are 50 km averages in the
  horizontal as opposed to single point RAOB
  soundings3

The only way to verify GOES soundings is
through RAOB comparisons, however, this method is
not ideal
8
Products Available

• The GOES sounder products are separated into 3
  groups in AWIPS

• derived product imagery (DPI)

• atmospheric sounding products

• high density wind products

• Lifted Index (LI)

• Total Precipitable Water (PW)

• Cloud Top Height

• Skin Temperature

• These images present data in a color-defined
  image without contours. IR satellite data is
  overlaid.

9
Products Available

• Atmospheric sounding products in AWIPS consist
  of clear sky temperature and moisture profiles

• More sounding products are available from NESDIS
  online. Some of the products include vertical
  sounding profiles, horizontal sounding fields,
  cross sections, and 3-hour time difference fields.

Figure 4, 5, and 6 (from left to right). GOES
sounder skew-T (Ft. Worth), cross-section
(relative humidity between Dallas, TX and San
Diego, CA), and horizontal sounding TPW water
images, respectively.4
10
Products Available

• AWIPS does not currently contain GOES high
  density wind products

• They will become available after the AWIPS OB2
  upgrade at the end of September 2003

• Wind speed and directions are inferred from
  satellite image trends

Figure 8. AWIPS high density wind menu after OB2
upgrade in late September.5
Figure 7. GOES high density water vapor wind
product.4
11
Configuration in AWIPS

• The major GOES Satellite Sounder products are
  available under the satellite tab in D2D

• DPI are accessible under the GOES Sounder
  Derived Product Imagery tab in the satellite
  menu. These products are available at any scale
  except WFO. The product is compromised slightly
  at this level due to its 50 km x 50 km
  resolution.

• DPI are available roughly 30-35 minutes after
  the hour

Figure 9. Future AWIPS satellite menu.5
12
Configuration in AWIPS

• Viewing GOES temperature and moisture profiles
  requires a few more steps

• select GOES Sounding Availability from the
  satellite menu

• select the editable points icon from the AWIPS
  toolbar and place a point at the site of interest

• open the volume browser and choose GoesBufr as
  the source, sounding as the field, and the
  specific point as the plane

To see an application of these different
products, lets consider a severe weather case
study
13
Case Study July 22, 2003 N. Texas

• On the evening of July 22, 2003, localized
  severe thunderstorms occurred across North Texas.
  Only slight chances of precipitation were
  forecast in the northeast, and SPC day 1
  forecasts did not include much of North Texas in
  a slight risk for severe thunderstorms until
  after 00Z.

Figure 10 (left) and Figure 11 (center). SPC Day
1 outlooks on 07/22/03 at 20Z and 00Z,
respectively.6 Figure 12 (right). Storm reports
from the Storm Prediction Center for 7/22/03.
There were a total of 12 reports in N. Texas (7
hail and 5 wind).7
Did the GOES sounder products provide useful data
that could have assisted forecasters in
anticipating this event?
14
Case Study July 22, 2003 N. Texas
Figure 13. 12Z surface station analysis with
frontal boundary and radar overlay. Isobars
(blue lines) are also given.8

• The 12Z morning surface analysis identified a
  stationary front extending across Arkansas and
  central Oklahoma. Models forecasted the boundary
  to slowly progress southward into North Texas
  later in the day.

15
Case Study July 22, 2003 N. Texas

• Interesting features from the 12Z KFWD sounding
  from that morning included

Figure 14. 12Z KFWD sounding
16
Case Study July 22, 2003 N. Texas

• In addition, the morning sounding from KFWD
  showed

• an overall moist sounding with a column PW value
  of 1.42 inches

• a fairly unstable atmosphere with an LI of -4

• With the frontal boundary forecast to approach
  the area by mid to late afternoon, the main
  forecast problems were the level of instability
  in place later in the day as well as the depth of
  moisture in the mid and lower levels

Lets see how the GOES sounder LI DPI and GOES
temperature/moisture profiles depicted the
environmental changes that occurred throughout
the day
17
Case Study July 22, 2003 N. Texas
-1

• The GOES 18Z LI image with the 6 hr. ETA
  forecast LIs overlaid (12Z run) show good
  overall agreement

-2
0

• GOES depicts slightly more unstable regions
  across Texas and Oklahoma than the ETA forecast.
  Notice the axis of higher instabilities in
  Oklahoma with respect to the surrounding values.

-3
-4
-5
-7
-6
Lets progress the GOES LI data throughout the
afternoon
Figure 15. The 18Z GOES LI DPI (image) with
added analysis (black contours). The 12Z ETA 6
hour forecast LIs (green contours) are overlaid.
The GOES scale is shown below.
18
Case Study July 22, 2003 N. Texas
Trend continues
Storms initiate
19
Case Study July 22, 2003 N. Texas
Figure 20. 21Z GOES LI image with 21Z LAPS LI
contours overlaid. The two are in very good
agreement.

• LAPS LI data helps to support this
  destabilization trend as well. At 21Z, LAPS data
  showed an LI maximum of 9 along the Red River.
  Storms formed soon after in Grayson county in the
  area of maximum instability.

20
Case Study July 22, 2003 N. Texas

• GOES soundings also gave good insight into the
  change in moisture that was occurring as the day
  progressed. At 19Z, the Grayson County sounding
  showed

Figure 21. 19Z GOES sounding from Grayson County

21
Case Study July 22, 2003 N. Texas

• This can be compared (verified) with the 00Z
  KFWD sounding

TPW has increased from 1.42 inches to 1.84 inches
since 12Z
For example the 750 mb dew point has increased
from -2ºC to 10ºC since 12Z
Figure 22. 00Z KFWD sounding
22
Case Study July 22, 2003 N. Texas

• Overlaying the KFWD 00Z sounding with the
  nearest GOES sounding from 00Z, we see that the
  GOES sounding does a good job of depicting the
  increase in lower level moisture.

RAOB LI is 4.6 and the GOES LI is 6.3
TPW values from each sounding are equal (1.84
inches)
Figure 23. 00Z KFWD sounding (red) with nearby
00Z GOES sounding (green) overlaid
23
Case Study July 22, 2003 N. Texas

• Although not perfect, the GOES soundings around
  the DFW area and northward depicted

• an erosion of the cap

• a significant increase in lower and middle layer
  moisture

• increasing instability (max LIs between 6 and
  8)

• increasing PW values (approaching 2 inches)

• Such favorable comparison of the GOES sounding
  near KFWD with the 00Z RAOB helps to validate the
  trends that were being observed throughout the
  day

24
Case Study July 22, 2003 N. Texas

• Using the GOES data in conjunction with LAPS
  data such as moisture convergence and CAPE as
  well as 0-6km shear, thunderstorms appeared
  evident in the following few hours on July 22,
  2003

Figure 26. SPC 22Z 0-6 km shear.9
Figure 24. LAPS 21Z moisture convergence image
with contour overlay.
Figure 25. LAPS 21Z CAPE image with contour
overlay
25
2-week statistical study

• From July 21, 2003 through August 1 (no weekend
  data), 2003, I compared the moisture and
  temperature profiles from the 12Z KFWD sounding
  to the nearest GOES sounding profile13

• Although only 10 sounding comparisons were made,
  the results are interesting

Figure 27 (left). KFWD 12Z sounding temperature
data compared to nearby GOES retrieval. Note the
warm bias (blue curve) and the RMS error (red
curve) near 1 throughout. Figure 28 (right).
KFWD 12Z sounding dewpoint data compared to
nearby GOES retrieval. There is a warm (moist)
bias (blue curve) and the RMS errors (red curve)
are considerably higher.
26
2-week statistical study

• The temperature and dewpoint verification show
  that the GOES soundings near KFWD had a warm
  (moist) bias throughout the majority of the
  atmosphere

• Overall, the temperature profile had an RMS
  error of 1.08 ºC and the dewpoint profile had an
  error of 5.66 ºC

• The more significant error in the moisture data
  can be attributed to the model forecast first
  guess

• Although the moisture profile had a fairly high
  bias and RMS error, total column values (PW) were
  quite close to RAOB PW values

• Comparing these results with an 11-month study
  is also quite interesting

27
2-week statistical study
Figure 29 (left) and Figure 30 (right). RAOB
temperature and dewpoint comparisons with nearby
GOES sounding data, respectively (bias is left
curve, RMS is right curve).2

• The temperature profiles are similar with a warm
  bias and an RMS error which is slightly higher
  than my study. The moisture profiles, however,
  are quite different with a cooler (drier) bias,
  most likely due to the season and higher volume
  of stations observed in the study.

28
2-week statistical study

• Statistical studies have found that

• moisture profiles have a much higher error due
  to model first guess

• satellite-derived integrated parameters compare
  better with their sonde-derived counterparts than
  do level-specific data10

• there is a demand for the next generation of
  higher resolution spectral sounders

While we look forward to the day of
high-resolution geostationary sounders, the
radiometers currently available are adding
information for both numerical weather prediction
and forecasting applications.14
29
Results of other studies

• The brief studies I have performed represent
  only a small fraction of the studies done on GOES
  sounders

• A four month study done by Rao and Fuelberg in
  1995 found that

• at 620 mb the NGM first guess temperature was
  improved 49 of the time, degraded 11 of the
  time, and not changed (within /- 0.25) 40 of
  the time

• at 620 mb the NGM first guess dew point was
  improved 51 of the time, degraded 34 of the
  time, and not changed 15 of the time

• surface to 300 mb NGM first guess PW values
  were improved 48 of the time, degraded 31 of
  the time, and not changed 21 of the time10

30
Results of other studies

• A 12-month study by Dostalek and Schmit in 1999
  found that

• DPI ETA first guess biases were reduced in
  relatively moist atmospheres

• there was an increased accuracy in measuring
  time tendencies of TPW11

• An 11-month study during 1996-1997 by Menzel et.
  al. found that

• GOES LI values were found to be 0.6ºC more
  unstable on average than LIs found from
  collocated RAOB sites

• GOES retrievals reduced first guess RMS errors
  at all levels and the bias was reduced by up to
  0.3 K between 960 and 650 mb2

31
Results of other studies

• Testing the ETA model with GOES PW data in 1996

• reduced the 24-hour precipitation false alarm
  rates by 10-20

• improved the equitable threat score from 0.2 to
  0.3

• resulted in operational use of PW retrievals in
  Oct. 19972

• The NWS performed a field assessment of the GOES
  sounder products in 1999 (with 37 Forecast
  offices participating) and found

• Participants indicated the use of GOES Sounder
  products heightened their situational awareness
  to potential watch/warning scenarios.

• the GOES sounder products led to improved
  forecast products in over 79 of all active
  weather situations12

32
Summary
GOES sounding Pros

• Hourly updated data

• Helps to pick up on trends fills in RAOB
  space/time gaps

• Column data (e.g. PW) found to be quite good

• GOES data improves model data more often than not

GOES sounding Cons

• Sounder profiles only available in clear
  conditions

• Soundings are highly dependent on model data

• Moisture data is inconsistent (more volatile
  than temps.)

• Sounder resolution (horiz. and vert.) needs
  improvement

33
Summary

• GOES sounder data has proven to be a useful
  operational tool through various operational and
  statistical studies

• Although not perfect (however there is no way to
  fully measure its accuracy), GOES data helps to
  visualize temporal trends and give a good
  volumetric depiction of the atmosphere at many
  more locations than RAOB data offer

• Its most effective uses at this particular
  office would be in

• pre-convective environments, especially when cap
  strength or moisture trends are the forecast
  problems of the day

• spring season Gulf of Mexico moisture return
  scenarios

• events when the magnitude of upper level jet
  streaks and wind shear are significant forecast
  problems

34

• Online Case Studies

• http//orbit-net.nesdis.noaa.gov/goes/cases/cs1/c
  asestudy1.html

• http//orbit35i.nesdis.noaa.gov/goes/cases/jarrel
  l

• http//orbit35i.nesdis.noaa.gov/goes/cases/fltorn
  /html/fltorn.html

• http//www.crh.noaa.gov/fsd/eros.html

• Online GOES Data links

• http//orbit35i.nesdis.noaa.gov/goes/

• http//orbit35i.nesdis.noaa.gov/goes/soundings/sk
  ewt/html/skewtus.html

• http//cimss.ssec.wisc.edu/goes/realtime/

• Training Software

• Comet Forecasters Multimedia Library, Satellite
  Meteorology Using the GOES Sounder (6 case
  studies)

35
Questions, comments?
36
References
1Menzel, W. P. and J. F. W. Purdom, 1994
Introducing GOES-I The First of a New Generation
of Geostationary Operational Environmental
Satellites. Bulletin of the American
Meteorological Society, 75, 757-781. 2Menzel, W.
P., F. C. Holt, T. J. Schmit, R. M. Aune, A. J.
Schreiner, G. S. Wade, and D. G. Gray, 1998
Application of GOES-8/9 Soundings to Weather
Forecasting and Nowcasting. Bulletin of the
American Meteorological Society, 79,
2059-2077. 3Satellite Meteorology Using the GOES
Sounder, 1998. COMET Forecasters Multimedia
Library. 4http//orbit35i.nesdis.noaa.gov/goes 5ht
tp//onestop.noaa3.awips.noaa.gov/awipsdoc/aumob2/
2_1_6_9.htm 6http//www.spc.noaa.gov/outlooks 7htt
p//www.spc.noaa.gov/climo/reports/030722_rpts.htm
l 8http//weather.unisys.com/archive/sfc_map/0307/
03072212.gif 9http//www.spc.noaa.gov/exper/mesoan
alysis 10Rao, P. A. and H. E. Fuelberg, 1998 An
Evaluation of GOES-8 Retrievals. Journal of
Applied Meteorology, 37, 1577-1587. 11Dostalek,
J. F. and T. J. Schmit, 2001 Total Precipitable
Water Measurements from GOES Sounder Derived
Product Imagery. Weather and Forecasting, 16,
573-587. 122000 1999 NWS Field Assessment of
GOES Sounder Products, Final Report
37
13http//weather.uwyo.edu/upperair/naconf.html
and htttp//orbit-net.nesdis.noaa.gov/goes/soundi
ngs/skewt/html/skewtus.html 14Schmit, T. J, W. F.
Feltz, W. P. Menzel, J. Jung, A. P. Noel, J. N.
Heil, J. P. Nelson III, and G. S. Wade, 2002
Validation and Use of GOES Sounder Moisture
Information. Weather and Forecasting, 17,
139-154. Bosart, Lance F., 2003 Whither the
Weather Analysis and Forecasting Process?
Weather and Forecasting, 18, 520-529. Feltz, W.
F., W. L. Smith, H. B. Howell, R. O. Knuteson, H.
Woolf, and H. E. Revercomb, 2003
Near- Continuous Profiling of Temperature,
Moisture, and Atmospheric Stability Using the
Atmospheric Emitted Radiance Interferometer
(AERI). Journal of Applied Meteorology, 42,
584-597. Hayden, Christopher M., G. S. Wade, and
T. J. Schmit, 1996 Derived Product Imagery from
GOES-8. Journal of Applied Meteorology, 35,
153-162. Moller, A. R., Severe Local Storms
Forecasting. Meteorological Monographs, 28,
433-480. Rao, P. A. and H. E. Fuelberg, 1997
Diagnosing Convective Instability from GOES-8
Radiances. Journal of Applied Meteorology, 36,
350-364.