Consistent MSU Radiance Dataset for Reanalysis
Cheng-Zhi Zou
NOAA/NESDIS/Center for Satellite Applications and research
CFSRR First Advisory Board Meeting, NOAA Science Center, November 7, 2007
NOAA MSU Satellites
4 channels to measure the
atmospheric temperature profiles
No cloud contamination
Each satellite has
a life cycle of a few years
Satellite Local Equator Crossing Time (LECT)
Each satellite overlaps vs time
With other satellites—good for bias
correction
LECT gradually changes
With time–
orbital drift phenomenon
MSU Scan Pattern and footprint sizes
Time series based on
pre-launch calibrated radiance data
Different bias correction
yield different trend results
(Christy and Spenser,
Mears and Wentz,
Vinnikov and Grody,
Zou et al.)
Reanalysis has its own
bias correction procedure,
how to ensure the
reanalysis trend to
represent the real
atmospheric trend?
Simultaneous Nadir Overpass
Method to find SNO matchups:
Use Cao’s (2004) method
to find the orbits that
have intersections
Use time and location
information in the 1B
Schematic viewing the overpasses
file to determine between two NOAA satellites
simultaneity between
two pixels
SNO Locations
MSU In-Orbit Calibration Process
Cold Space
T=2.73K
MSU Sensor
Warm Target
Temperature is
measured by PRT
Earth
Conceptual diagram of MSU observational procedure
Level 0 Calibration
(Cw, Rw)
(Ce, Re)
Radiance (R)
Onboard Warm Target
(Ce, RL)
Earth View
(Cc , 2.73K)
Cosmic Cold Target
Digital Counts (C)
SNO Radiance Error Model
Rk RL ,k Rk k Z k
R j RL , j R j j Z j
k j
Radiance Error Model for SNO Matchup
K and J :
R RL R k Z k j Z j
Consider colinearity between Zk and Zj
choose the coefficients that completely remove the
warm target temperature contamination
Tb comparison for SNO matchups
After pre-launch calibration After SNO calibration
Comparison Between Pre-launch
and SNO calibration
Time series
differences for
pre-launch
calibration Std=0.1K
Inter-satellite
differences
after SNO std=0.04K
calibration
Anomaly time series
Independent validation dataset for reanalysis
Website address:
http://www.orbit.nesdis.noaa.gov/smcd/emb/mscat/mscatmain.htm
Datasets available:
Level 2 radiance:
pre-launch (operationally) calibrated
SNO calibrated
Level 3 SNO calibrated gridded products: 2.502.50
pentad T2, T3, and T4, 1987-present
pentad anomaly T2, T3, and T4, 1987-present
monthly T2, T3, and T4, 1987-present
monthly anomaly T2, T3, and T4,1987-present
Trend patterns
T3 trend
T23, Mid-tropospheric temperature
trend, 1987-2006
T4, Lower
Stratospheric
temperature trend
1987-2006
Remaining issues
• Efforts needed to resolve short overlap problem
between NOAA 10 and 9 for accurate bias removal
• more calibration coefficients (higher order calibration
equation) maybe needed to solve Channel 3 for
NOAA 11 and 12
• Different frequencies between MSU and AMSU