Could Data Assimilation Be Useful for Cloud Verification · 2009-08-28 · DTC Verification Workshop - Aug. 26~28, 2009, Boulder 1 Could Data Assimilation Be Useful for Cloud Verification?

DTC Verification Workshop - Aug. 26~28, 2009, Boulder 1

Could Data Assimilation Be Usefulfor Cloud Verification?

Zhiquan Liu (liuz@ucar.edu)NCAR/MMM

With slides contributed from MMM and DATC colleagues

DTC Verification Workshop - Aug. 26~28, 2009, Boulder 2

• Data assimilation process• DA linkage to verification• Potential extension for cloud verification• Synergy with MET

Outline

DTC Verification Workshop - Aug. 26~28, 2009, Boulder 3

• Data assimilation process• DA linkage to verification• Potential extension for cloud verification• Synergy with MET

Outline

DTC Verification Workshop - Aug. 26~28, 2009, Boulder 4

Data Assimilation:correct “background” with observations

DTC Verification Workshop - Aug. 26~28, 2009, Boulder 5

Variational Data Assimilation:Incremental Formulation

J(!x) =1

2

T!x

"1B !x +

1

2[

TH!x " d] "1

R [H!x " d]

Find an optimal atmospheric state Xa by minimizing a cost function

!x = x - xb

(increment)

d = y ! H (xb)

(departure of obs minus background) in obs space (time, location, quantity)

Background in model space, often a short-term model forecast

Observation operator

Also called “innovation”, an important by-product from DA

DTC Verification Workshop - Aug. 26~28, 2009, Boulder 6

H - Observation operatorH maps variables from “model space” to “observation space” x y

– Interpolations from model grids to observation locations– Extrapolations using PBL schemes– Time integration using full NWP models– Transformations of model variables (u, v, T, q, ps, etc.) to “indirect” observations (e.g.

satellite radiance, radar radial winds, etc.)• Simple relations like PW, radial wind, refractivity, …• Radar reflectivity

• Radiance: Radiative transfer models (CRTM, RTTOV)

• Precipitation using simple or complex models• …

Z = Z(T , IWC, LWC, RWC,SWC)

DTC Verification Workshop - Aug. 26~28, 2009, Boulder 7

WRFDA Observations In-Situ:

- Surface (SYNOP, METAR, SHIP, BUOY).- Upper air (TEMP, PIBAL, AIREP, ACARS, TAMDAR).

• Remotely sensed retrievals:- Atmospheric Motion Vectors (geo/polar).- SATEM thickness.- Ground-based GPS Total Precipitable Water/Zenith Total Delay.- SSM/I oceanic surface wind speed and TPW.- Scatterometer oceanic surface winds.- Wind Profiler.- Radar radial velocities and reflectivities.- Satellite temperature/humidity/thickness profiles.- GPS refractivity (e.g. COSMIC).

Radiances (or Brightness Temperature):– HIRS from NOAA-16, NOAA-17, NOAA-18, METOP-2– AMSU-A from NOAA-15, NOAA-16, NOAA-18, EOS-Aqua, METOP-2– AMSU-B from NOAA-15, NOAA-16, NOAA-17– MHS from NOAA-18, METOP-2– AIRS from EOS-Aqua– SSMIS from DMSP-16– ……… many other instruments can be included.

DTC Verification Workshop - Aug. 26~28, 2009, Boulder 8

• Data assimilation process• DA linkage to verification• Potential extension for cloud verification• Synergy with MET

Outline

DTC Verification Workshop - Aug. 26~28, 2009, Boulder 9

MET Verification(point-stat)

Obs in prepbufr

pb2nc

Obs in netcdf

stat_analysispoint_stat

Get matched obs/fcstpairs and 71 continu-ous statistics value

Providing sum-mary statisticalinformation

Fcst in netcdf

WPP

Fcst in grib

pikc2ncl.f90

Pick up obs/fcstmatched pairsand BIAS, RMSE

NCLsctips

Figures for U,V, TMP, SPFH

Analysis/Forecast

X in NETCDF

Observations y

in prepbufr/ascii

WRF-VarInnovations

d = y - H(x)

Statistics &Visualization

WRFDA-based verification(vs. observations)

DTC Verification Workshop - Aug. 26~28, 2009, Boulder 10

Advantages of DA-based Verification• Can make use of build-in QC procedure in DA system

– Obs QC is necessary before verification: reject outliers

• Can make use of built-in observation operators– particularly those for non-conventional data: radar, satellite radiances

• Calculation of OMF is parallelized for most DA system.

• Provide also offline statistics and visualization package– WRF-Var: BIAS, ABIAS, RMSE; profile and time series

Disadvantage of DA-based Verification• Only basic statistics provided

– It is not specially designed for verification, it is more like a by-product of DA.

DTC Verification Workshop - Aug. 26~28, 2009, Boulder 11

Verification against NCEP prepbufr Obs.WRF-Var generated score MET generated score

U V

T Q

outlier

Obs outliers entered into the statistics

DTC Verification Workshop - Aug. 26~28, 2009, Boulder 12

• Data assimilation process• DA linkage to verification• Potential extension for cloud verification

• Satellite Radiances: cloud & precip.• Limited vertical information

• CloudSat: vertical cloud information• Synergy with MET

Outline

DTC Verification Workshop - Aug. 26~28, 2009, Boulder 13

Model cloud/rain and Synthetic SSMIS radiances affected by cloud/rain(use CRTM within WRF-Var)

Column-Integrated cloud water Column-Integrated rain water Radar Reflectivity

Simulated Ch2 Tbs Simulated Ch17 Tbs Scatter plots of observed v. calculated Tb

DTC Verification Workshop - Aug. 26~28, 2009, Boulder 14------ OMB before bias correction------ OMB after bias correction

DTC Verification Workshop - Aug. 26~28, 2009, Boulder 15

Tropical Storm Ernesto (2006/08/29)CloudSat path: 18:46:25 - 18:48:24

WRF 19h forecast (4km) from GFDL IC, valid at 2006-08-29-19Z.

Black line is CloudSat path

TPW

DTC Verification Workshop - Aug. 26~28, 2009, Boulder 16

Doppler Radar Reflectivity (diagnosed from RIP)

WRF 4km Forecast at 19h

CloudSat Radar Reflectivity

Could have a better way to compare model cloud with CloudSat radar reflectivity

Use Radar Simulator “QuickBeam” (Haynes et al., 2007)as part of obs operator in DA system

DTC Verification Workshop - Aug. 26~28, 2009, Boulder 17

• Data assimilation process• DA linkage to verification• Potential extension for cloud verification• Synergy with MET

• Make use of WRFDA’s capacity of calculating OMF,specially for Radar and Satellite radiances

• Combine with MET’s advanced statistics features(e.g., MODE)

• Use radiances from geostationary satellites• High time/spatial resolution

Outline

DTC Verification Workshop - Aug. 26~28, 2009, Boulder 18

ECMWF: validate model from geostationaryimages: Forecasted Tbs: WV channel

• 42-hour forecast vs. observed

RTTOV computed Radiances

DTC Verification Workshop - Aug. 26~28, 2009, Boulder 19

Answer is YES!