Characterizing Tropospheric Dynamics Using MLS Geopotential Height

1
Characterizing Tropospheric Dynamics Using MLS
Geopotential Height
Evan Fishbein, Michael Schwartz, Nathaniel
Livesey and Richard Cofield Jet Propulsion
Laboratory California Institute of Technology
2
Motivation

• Can MLS geopotential height add information to
  characterize tropospheric dynamics.
• Weather forecast systems are excellent at
  representing large scale atmospheric dynamics.
• MLS geopotential height may be especially
  valuable for characterizing tropical cyclones.
• Develop over tropical ocean away from in situ
  measurements.
• High cloud amounts limit usefulness of microwave
  and infrared nadir sounders.
• Perform case study on Hurricane Katrina comparing
  MLS and NCEP GFS geopotential height.
• Are differences coherent and physically
  plausible?

3
Measurement Technique

• MLS measures pressure at tangent point from
  pressure broadening of thermal O2 emissions.
• Altitude of tangent point from dead reckoning
  of position and attitude of Aura.
• Complicated error characterization
• Component from interpretation of MLS radiances.
• Measurement requires Aura orbit/attitude
  knowledge and MLS antenna boresight alignment
  beyond design requirements.
• Error has a time varying orbital and seasonable
  component from thermal environment of Aura.

4
Amount of MLS Geopotential Height Data

• 3500 profiles per day.
• Histogram of repeat time of transects within 400
  km of Hurricanes.
• Repeat measurements 12 hours later 40 of time.
• Longer than 48 hours less than 13.
• Number of transects over life of storm (260
  events).
• 60 of storms have more than 4 transects.

5
Characterizating of Hurricane Dynamics

• Hurricanes form over warm ocean usually in remote
  areas
• Low pressure near surface carrying moist air into
  and upward into eye wall and rain band clouds
  generating potential vorticity
• High pressure in upper troposphere as airs
  detrains from eye wall
• MLS could improve characterization of upper level
  high pressure in mature cyclones and surface
  pressure depression during cyclogenesis (MLS/AMSU)

6
Hurricane Katrina Transects

• 6 MLS transects within 400 km
• 22 Aug 2005 064145 - Dsc
• 23 Aug 2005 182034 - Asc
• 25 Aug 2005 071126 - Dsc
• 26 Aug 2005 185153 - Asc
• 28 Aug 2005 074246 - Dsc
• 29 Aug 2005 192427 Asc
• 26 Aug 2005 1851 Transect

Caveat MLS pointing drift correction derived
from NCEP geopotential height
7
Overview of Environment from NCEP GFS Analysis on
22 Aug 2005

• 500 hPa Geopotential Height and Winds
• Designated tropical depression on 23 Aug at
  2100 UT
• Backward extrapolation of storm track
• Three weak lows, one near extrapolate storm track

8
Upper Level Development on 22 Aug 2005

• 100 hPa height is monotonically increasing
  towards northwest
• NCEP GFS 100 hPa Geopotential Height and Winds
  shows no development
• 100 hPa 500 hPa

9
Comparison of MLS and NCEP 100 hPa Height

• Two data sets are in agreement to within noise

Spikes dominates error in surface pressure
determination

• Need to understand if noise is limited by MLS
  tangent pressure sensitivity or pointing
  knowledge error.

10
Combined MLS and AMSU Surface Pressure

• Apply hypsometric equation to Aqua AMSU
  temperature layer using MLS 100 hPa height
  reference level

11
MLS and NCEP 100 hPa Mature Storm Geopotential
Height Comparison

• Coherent structure in MLS Height above storm

12
August 25 Surface Pressure Transect
Katrina

• Cuba

Gap is missing AMSU temperature
13
Upper Level Geopotential Height Mature Cyclone

• 100 hPa MLS NCEP height
• 28 Aug 2005 0743
• 1 day before landfall
• Amplitude of high are equivalent, but MLS decays
  faster toward equator
• Spike at 5 N is common near equator and could
  be dynamics associated with ITCZ

14
Summary

• MLS/AMSU crossed Katrina storm track six times
  between 22 29 August 2005.
• MLS and NCEP 100 hPa geopotential height track,
  but differences in structure.
• Differences between forecast NCEP surface
  pressure are small and have a spatially
  correlated component.
• Largest component of error budget is boresight
  pointing and satellite orbit uncertainty.
• MLS/AMSU surface pressures track forecasted
  pressures, but
• Core pressures not useful because of low AMSU
  temperature sensitivity in precipitating scenes.

15
Conclusions

• Goal was to determine if MLS geopotential height
  and MLS/AMSU surface pressure are scientifically
  useful YES.
• Differences between MLS and NCEP height are often
  coherent, but further validation is required.
• Further algorithm refinements would
• Improve error estimates.
• Improve pressure measurement with more optimal
  use of unsaturated radiances.
• Improve pointing and drift correction, remove
  dependence on NCEP gepotential height and include
  better pointing model.
• Working with FOT, near real-time geopotential
  height and surface pressure measurements could be
  generated for operational weather forecasting.