Generating climate benchmark atmospheric soundings using GPS occultation data

1
Generating climate benchmark atmospheric
soundings using GPS occultation data

• Anthony J. Mannucci, Chi O. Ao, Thomas P. Yunck,
  Larry E. Young, George A. Hajj, Byron A. Iijima,
  Da Kuang, Thomas K. Meehan
• Jet Propulsion Laboratory, California Institute
  of Technology
• Stephen S. Leroy
• Division of Engineering and Applied Sciences,
  Harvard University

First Formosat-3/COSMIC Data Users Workshop
October 16-18, 2006 Boulder, CO
2
Overview of Todays Talk

• How GPS RO can become a climate benchmark
• Identify and analyze error sources
• Look at what we know from a different perspective
• Error analysis for realistic L2 tracking
• Future directions

3
Establishing a climate benchmark
2 a a point of reference from which
measurements may be made b something that
serves as a standard by which others may be
measured or judged

• Systematic errors assessed by comparison to
  independent observations
• Accuracy standard based on SI-traceability
• Error analysis
• There will not be an independent measurement of
  exactly the same quantity
• Documented absolute accuracy
• Methods for establishing accuracy must be
  documented and widely accepted
• Adequate archiving and transparency
• Appropriate management infrastructure
• Geophysical meaning for climate
• Measurement coverage and density

4
Approach Overview of Error Analysis
Ohring et al., BAMS 2005 Satellite Instrument
Calibration for Measuring Global Climate Change

• Extensive analysis conducted by Kursinski et al.,
  JGR 1997
• Linear error propagation
• Simulation

Minor impact
Limit range of validity
Error categories
Further work needed to establish SI-traceability
Additional discussion in Generating climate
benchmark atmospheric soundings using
GPS occultation data, Proceedings of the SPIE
Conference, San Diego CA, 2006, Mannucci et al.
5
Errors Limiting Measurement Interpretation For
Temperature
Upper altitude limit low SNR, upper altitude
initialization (temperature), ionosphere
Sweet spot for climate
Lower altitude limit water vapor
6
Systematic Errors That Might Impact Trend
Measurements

• Focus on troposphere trend requirement 0.04
  K/decade
• Systematic errors that vary over time
• Ionosphere
• 11-year solar cycle
• Orbit determination error
• Requirement 0.05 mm/sec or better
• Possible trends in OD strategy or results
• Need to establish SI-traceability of orbit errors
• Local multipath
• Varying satellite generations

7
Solar Cycle Variation
8
Modeled Electron Density Variations
9
Residual Ionosphere Error After Dual-Frequency
Correction
Method ray-trace signal through a model
ionosphere
From the solar maximum simulation of Kursinski et
al. JGR, 1997
10
An Instrumental Source of Ionospheric Error

• Extrapolation of L2 iono bending below 15 km

SAC-C July 1, 2006
Median 11.6 km
Number of Occultations
2
22 km Cutoff Height
12
11
Simulation Results Extrapolation Error
Temperature errors 0.1 K
12
GRACE Error Analysis August 2006

• Differences between measured and extrapolated L2

Solar Min
13
Role Of Ionospheric Error

• Simulations suggest gt 0.1 K systematic error over
  the solar cycle
• Processing methods can be modified to reduce this
  error
• Use data-driven global ionospheric models to
  compute bending explicitly
• Flag cases with large ionospheric irregularities
• Use new GPS signals (to maintain L2 lock)

14
Studying Ionospheric Irregularities
Quiet (80 of cases)
Disturbed

• From Pavelyev et al., GPS Solutions 2005

15
Comparing COSMIC Profiles
16
COSMIC Profile Comparisons
Inter-quartile Range Contains central 50 of
differences
COSMIC3 - COSMIC2
Median
Window 30 km 10 minutes June 4-16, 06 224 pairs
1 K
17
COSMIC Refractivity Comparison
Median
Inter-quartile Range
COSMIC3 - COSMIC2
Window 30 km 10 minutes June 4-16, 06 224 pairs
1
18
COSMIC3 - COSMIC1
Inter-quartile Range
COSMIC3 - COSMIC1
Median
Window 30 km 10 minutes June 4-16, 06 80 pairs
1 K
19
Interpretation

• Errors addressed with COSMIC comparisons
• Instrumental bias
• Non-common orbit error
• Multipath (most of it)
• Errors not addressed with COSMIC comparisons
• Ionosphere
• Temperature initialization

20
Questions For The Community

• Are we taking the steps necessary to enure that
  on-orbit SI-traceability can be verified a few
  decades from now?
• 0.04 K/decade requirement
• Are all necessary data being stored?
• Are instrument software being archived?
• What steps are needed to establish in-orbit
  absolute calibration (SI-traceability)?
• Ionosphere
• Orbits
• Multipath
• Cycle slips, quality control
• What is optimal processing strategy for climate
  records?

21
Summary and Conclusions

• Work remains to establish on-orbit
  SI-traceability for GPS radio occultation
  retrievals
• Existing analyses suggest the need for reducing
  residual ionospheric error
• Goal climate-oriented retrieval products with
  full traceability to absolute standards
• Recent results with COSMIC instruments are very
  encouraging