Status of AIRS Data at Met Office A. Collard , R. Saunders, J. - - PowerPoint PPT Presentation

AIRS Net Meeting, 24th February 2003

Status of AIRS Data at Met Office

n Data monitoring n Cloud detection n Bias tuning n Data assimilation plans

• A. Collard , R. Saunders,
• J. Cameron, Y. Takeuchi, P. Rayer and
• B. Harris

AIRS Net Meeting, 26 Feb 03

Data monitoring

n BUFR radiance data are compared with 6hr forecast

fields from the Met Office global model

n Results from these comparisons are placed on our

monitoring pages (password protected)

n Plots updated at 8Z each day n Maps of O-B for all 324 channels for clear fovs n Selected spectra and O-B spectra n The plots show time series of past 3 months

– % rejections from 1DVar q/c checks – Observed-Background mean and st. dev.

AIRS Net Meeting, 26 Feb 03

Monitoring web page

Available to the AIRS team in mid-December via password protected page

• n Met Office site.

AIRS Net Meeting, 26 Feb 03

Time series of observations

CO2 CO2 H2O O3

AMSU

Global Observations: 4/5th Dec 2002 21Z-3Z Rejects caused by Channel 2357

AIRS Net Meeting, 26 Feb 03

Tartan Plots: O-B clear mean bias

Global Observations H2O O3 CO2 CO2 AMSU 1&2

AIRS Net Meeting, 26 Feb 03

20-70N 20-70S 20N-20S

O-B st. dev plots

Stratosphere Water vapour

AIRS Net Meeting, 26 Feb 03

Day & Night spectra

AIRS Net Meeting, 26 Feb 03

Ocean skin to bulk temperature model validation with ATSR

n Night time satellite/buoy

matchups

n Spatial match to within

1/6 degree/temporal match within 1 hour

n Fairall model for skin

effect

n Can “convert” satellite

skin SST to bulk using model!

June 1995 – June 1997, n=1243

ATSR-2 – buoy skin effect Fairall model skin effect

Horrocks et al., JGR, in press

AIRS Net Meeting, 26 Feb 03

Bias tuning for AIRS

Predictors being investigated

n Scan angle n Model Tskin n Model Thickness 850-300 hPa n Model Thickness 200-50 hPa n Simulated brightness temperature n Model total column water n Model Lapse rate convolved with Jacobian

AIRS Net Meeting, 26 Feb 03

Example of HIRS-12 bias tuning

Uncorrected biases Corrected biases

AIRS Net Meeting, 26 Feb 03

Cloud detection

Cloud detection methodology

AIRS Net Meeting, 26 Feb 03

PCA cloud detection

112bbbobsLn{(|,)} ~(){()()}()Constcloud.JP-D=-++D-T

• bsb()=D-yxyy

Attempt to determine the probability of having cloud in the field of view given the observed radiances and the NWP background profile

) ( ) (

1

y y J

T

D D =

• S

) ( ) (

1 1

y U UX y T D D =

• 11()/NNTjjiiiijyUX===DÂÂ

is the ith component of

y D

AIRS Net Meeting, 26 Feb 03

Cloud detection: Results from simulated dataset

PCA01 PCA2 PCA01 PCA3

Red is clear Green is liquid water cloud Blue is ice cloud

AIRS Net Meeting, 26 Feb 03

Cloud detection: Validation

PCA scheme VAR scheme Blue Jc<2 Green Jc>2 Red Jc>20 Blue Jc<.94 Green Jc>.94 Red Jc>20

AIRS Net Meeting, 26 Feb 03

Assimilation trials

n Once initial bias tuning has been finalised will start

assimilation trials in March (as planned).

n Period will be October-November 2002, several poorly

forecast and well forecast storms during this period.

n Control will be current operations which includes:

– ATOVS (HIRS/AMSU-A/AMSU-B) – SeaWinds – SSM/I winds

n Plan to report on results of trials by August 03 n Met Office relocation will delay possible operational use of

AIRS until 2004

AIRS Net Meeting, 26 Feb 03

Summary

n AIRS continues to perform well and O-B stats

are stable and as expected. See web pages.

n Cloud detection (clear FOV) schemes using

VAR+PCA investigated. VAR scheme used for initial trials.

n Radiance bias correction is being

implemented

n We hope to begin assimilation trials in March.