Temperature Validation in a Nutshell*

1
Temperature Validation in a Nutshell

• Rapporteur
• Steven Pawson
• NASA GSFC

Its all about the kernels
2
Summary Comments

• Temperature validation seems to be in good shape,
  maybe because other retrievals depend on
  temperatures!
• Range of talks, covering
• Aura data validation
• Ancillary temperature datasets
• Features that need to be validated
• Commend presenters for sticking to time and
  thematic relevance

3
TES Validation

• TES-AIRS comparisons (A. Eldering)
• TES temperatures are less than 1K colder than
  AIRS below 400mb.
• TES is about 2K warmer than AIR above 400mb.
• RMS of less than 2K at all levels
• Similar mean differences with MLS
• Future L1B calibration will make TES temperature
  profiles less bias relative to AIRS in the upper
  troposphere
• TES versus sondes and aircraft data (B. Herman)
• Temperature comparisons have been carried out
  between TES L2 retrievals, WB-57, DC-8, and
  sondes.
• UT warm bias and LT cold bias in TES data are due
  to known calibration problems
• TES SSTs (Bob Herman)
• Promising avenue (compared against Hadley SST
  data)

4
TES-v-AIRS Summary
Mean profiles TES - AIRS
Bias in green TES-AIRS, rms differences in black
5
MLS Validation (M. Schwartz)

• MLS temperatures retrieved from several channels
• Core uses 118GHz (316-1hPa)
• 118240190GHz should be best in upper
  troposphere and mesosphere, but this combined
  product displays oscillations and bias!
• Comprehensive validation presented
• Residuals in radiance fitting
• Independent data types

6
MLS Summary
7
HIRDLS Comparisons with GEOS-4 Analyses
May 11
February 7
8
HIRDLS Validation

• Validation against GEOS-4 analyses (J. Gille)
• HIRDLS Temperatures agree with the geographical
  and vertical variations of GMAO temperatures from
  the tropopause to 50 km
• Temperature precision is what was described at
  CDR
• Profile comparisons show that HIRDLS results
  track high resolution features
• Data are being improved future results will
  improve on these
• Isolation of wave features (H. Lee)
• Geophysical signals of waves isolated - advantage
  of the near 1km vertical resolution
• Gravity waves apparently related to planetary
  wave-2 in SH
• Indications of agreement with GPS data
• Seasonal variations of wave activities

9
Waves (1) - HIRDLS

• 1.4e-4 s-1
• cgh Uh-70 m/s
• cgz 0.5 m/s

?z 3-4 km
?h 500 km
10
Waves (2) - Ticosonde

• Downward phase propagation of 4-5 day waves in
  temperature to tropopause level

Ticosonde-Aura/TCSP 2005 RS92 data
11
Ancillary Data(1) Sondes

• Ticosonde temperatures (R. Selkirk)
• Based in Costa Rica, in missions
• Six-hourly launches about 240 in summers of 2004
  and 2005
• About 20 of soundings indicate second coded
  tropopause lamination?
• Wave signals and diurnal cycle impacts
• Temperature soundings from ozone sondes (F.
  Schmidlin)
• Wallops, Ascension, Natal
• Addressed thermistor accuracy - impact on
  soundings
• Good agreement with MLSbetween 300 and 8hPa

12
Ancillary Data(2) Sounders

• MTP (M.J. Mahoney)
• Flown in missions
• Calibrated to sondes at flight level
• Accuracy better than 1K for 8km above and below
  flight altitude
• Scanning HIS (K. Vinson)
• Flown in missions
• Some departures from MTP - e.g., near tropopause

13
Discussion

• Need for ready data availability in AVDC and
  tools to help collocate validation data against
  Aura profiles
• Requirements for validation averaging kernels,
  representation error (points versus footprints)
• Other data types (lidar)
• Keep up the good work!