Stratospheric Measurements: Microwave Sounders I. Current Methods – MSU4/AMSU9 Diurnal Adjustment Merging II. Problems and Limitations III. Other AMSU Channels
Dec 30, 2015
Stratospheric Measurements:Microwave Sounders
I. Current Methods – MSU4/AMSU9Diurnal Adjustment
Merging
II. Problems and Limitations
III. Other AMSU Channels
Weighting Functions
Drifts in Measurement TimeDescending node is 12 hours earlier
Harmonic Sensitivity of Diurnal Correction
Since measurements are separated by ~12 hours, thefirst harmonic of the diurnal cycle is mostly cancelled if we average the ascending and descending nodes together.Instead, the second harmonic dominates.
Calculating the MSU Channel 4 Diurnal Cycles from CCM3
• 5 Years of hourly output from the Community Climate Model 3 (CCM3), run in “reanalysis” mode (1980-1984).
• Use a radiative transfer model to calculate a simulated MSU Channel 4 brightness temperature time series for each point on earth.
• Average together to compute mean monthly diurnal cycles for grid point.
MSU Channel 4 Diurnal Cycles from CCM3Second harmonic appears to be important and larger in the tropics
Are these reasonable???????
Impact of the CCM3 Diurnal Adjustment on MSU4/AMSU9
CCM3-based adjustment is fairly small, but this doesn’t mean it’s correct!
“Global” average, 50S to 50N
Validating the Diurnal Cycle/Adjustment
I. Radiosondes – Solar heating problems well documented. Probably can’t be trusted to see the small diurnal signal.
II. Ascending minus Descending satellite comparisons mostly sensitive to first harmonic – useful as a sanity check in the troposphere but less so for the stratosphere.
III. Multiple satellites at different measurement times (I.e. NOAA-15, NOAA-16, NOAA-17 and AQUA and now, I hope, Metop-A) but need independent relative calibration to < 0.1K. Maybe the “simultaneous nadir overpass” (SNO) calibration scheme can help. (Cheng-Zhi Zou et al)
IV. Or ……?????
MSU4/AMSU9 Merging
Antenna Temperature from 11 SatellitesEach dot is a 5-day average, 50S to 50N
MSU4/AMSU9 Merging
Intersatellite differences show offsets, trends, and wiggles.
Error Model and Pentad Difference Equations
For each pentad where 1 or more satellites have a good observation, we can from a difference equation for each satellite pair
, , , ,MEAS MEAS TARGET TARGETi j i j i i j jT T A A T T T
Typically > 1300 valid pentad pairs/equations
Solve equations to minimize T2
’s probably related to non-linearity in the radiometer
, ,0MEAS TARGETi i i i iT T A T
Error model includes intersatellite offsets, plus a dependence on target temperature anomaly (Spencer and Christy)
MSU4/AMSU9 MergingMSU-only differences are fit well by the empirical error modelAMSU-MSU differences include additional variation partly due (presumably) to small differences in the weighting function.
AMSU appears to drift relative to MSU!
MSU and AMSU Weighting Functions
Black: MSU4Red: AMSU9 nadirBlue: AMSU9 near limb
We use a near-limb set of AMSU views to help match MSU4. (views 4 - 7 and 24 - 27)
Not much impact on variability of MSU minus AMSU.
MSU minus AMSU Difference
Red: UnadjustedBlue: Mean seasonal cycle removed empirically
Trend of ~0.1K decade still present
Map of MSU 4/AMSU 9 Trends
Comparison to UAH Results
Reasons for the Difference not yet known, but probably a combination of
1. Differences in the Diurnal Adjustment
2. Differences in Merging Parameters, esp. Target Factors.
3. UAH is currently revamping their diurnal adjustment and treatment of AMSU
Extrapolating MSU4 upward
MSU4, near-nadir, Trend = -0.36K/Dec.
MSU4, near-limb, Trend = -0.44K/Dec.
MSU4,limb-nadir, Trend = -0.52K/decadeFOV_Weights = (0.5,0.5,0.0,0.0,-0.33,-0.33,-0.33,0.0,0.0,0.5,0.5)
SSU 15X
Other AMSU Channels
AMSU 9
AMSU 10
AMSU 11
AMSU 12
AMSU 13
SSU 15x
SSU 25
SSU 26SSU 27
SSU channels can reconstructed from AMSU data.
Sanity check AND diurnal adjustment.
Upper channels are on very narrow lines – small drifts in LO frequency could cause big problems – need to check stability
AMSU Channel 12 Measurement Bands
Frequency (GHz)
30,10,3 hPa
Abs
orbt
ivit
y (k
g-1)
Quick Conclusions
• MSU 4 probably accurate to ~0.1K/decade
• MSU 4 sees no evidence of the upward trend in SSU 15X after 1998
• All Stratospheric probably channels need to be corrected for diurnal effect. (Maybe AMSU can help here.)
• AMSU can probably be used for a sanity check for SSU channels after mid-1998. But there is a good chance that AMSU (too?) is drifting.
MSU OperationThe MSU is a cross-track scanning microwave sounder operating at 4 frequencies near the oxygen absorption line at 57 GHz.
Channel 1 Channel 2 Channel 3 Channel 4
Frequency (GHz)
50.30 53.74 54.96 57.05
Altitude of peak weight
Surface 4-7 km 10-12 km 17-19 km