Assimilation of Dual-Polarimetric Radar and GPM Observations with GSI in

regional WRFXuanli Li1, John Mecikalski1,

Bradley Zavodsky2, and Jayanthi Srikishen3

1Department of Atmospheric Sciences, University of Alabama in Huntsville2NASA Marshall Space Flight Center

3Universities Space Research Association

21-23 May 2014

12th JCSDA Science Workshop on Satellite Data Assimilation, College Park, MD 1

Outline

Background and objectives

Previous work: Assimilation of ground-based dual-pol radar data with GSI Comparison of dual-pol radar data assimilation using GSI vs. WRFVAR

Work plan and current progress: assimilation of GPM ground validation data with GSI

Background and Goals

• ROSES-13 A.33 project to assimilate GPM DPR reflectivity and GMI products, collaborated with NASA SPoRT Center.

• NWS WSR-88D network has been updated to include dual-polarimetric capability. The dual-pol radar can provide more information on cloud and precipitation particles. Assimilation of the dual-pol radar data is a relatively new area.

• GPM has been launched and DPR and GMI data will be available soon. Broader coverage than the ground-based radar system, better measurement for snow storm events.

• Project goal is to develop methodology to implement GPM

DPR and GMI data with GSI into regional WRF model, and investigate the potential of using GPM observation in convective scale NWP for operational environment.

3

Dual-Polarimetric Radar

Horizontal and vertical signals: more info about the type, shape, and size of the hydrometeors – more accurate estimates of precipitation and cloud particles.

Variables:

ZH: Horizontal reflectivity

VR: Radial velocity

ZDR: Differential reflectivity ZDR = 10

log10(ZH/ZV)

ρHV: Correlation coefficient, the coefficient between the horizontal and vertical power returns. ΦDP: Differential phase, the measured phase shift between horizontal and vertical pulses SW: Spectrum width, measures the consistence of the phase shifts

4

5

Dual-Pol Radar Data Assimilation with GSI

WRF model ARW v3.5

GSI v3.2

Assimilation procedure:

• Reflectivity is used by the Global Systems Division

(GSD) cloud analysis to improve precipitation analysis

• ZDR information is added in calculation of rain amount.

GSD Cloud Analysis for rain:

• Kessler (1969):

With ZDR, using Ulbrich and Atlas (1984):

br aq arg)(

94.141028.1 DRHr ZZq

WSR-88D Dual-Pol Radar Observation

VR

ZDR ZH

6

2 km height

0631 UTC 2 September 2013

Case Study: 2 September 2013

7

Reflectivity at 0600 UTC 2 September 2013

8

Model starts at 0000 UTC, no convection in model simulation at 0600 UTC

Data Assimilation Experiments

9

Experiment Radar Data Assimilation Variables

VR 0600 and 0900 UTC2 September 2014 Vr

ZH 0600 and 0900 UTC 2 September 2014 Vr and ZH

ZHZDR 0600 and 0900 UTC 2 September 2014

Vr, ZH and ZDR

Reflectivity 0900 UTC 2 September 2013

ZH

ZHZDR

10

VR

NEXRAD

Zdr data shows impact on the initial reflectivity and hydrometeor fields

Impact found in low level temperature field

Temperature 0900 UTC 2 September 2013

VR ZH

ZHZDR

11

Stronger dry region found in low level moisture field

Moisture 0900 UTC 2 September 2013

VR ZH

ZHZDR

12

Forecast Validation 1200 UTC 2 September 2013

NEXRADZHZDR

VR

13

ZH

Zdr data assimilation shows impact on the convective scale model forecast

14

GSI: GSD cloud analysis system

Indirect incorporation of dual-pol radar data

Convert to cloud type, precipitation amount

WRFVAR: direct assimilation of dual-pol radar data

moist control variables: water vapor, rain water, and

cloud water mixing ratio

Assimilation: using Ulbrich and Atlas (1984):

cycled assimilation at 0600 and 0900 UTC

94.141028.1 DRHr ZZq

Dual-Pol Radar Data Assimilation: GSI vs. WRFVAR

15

CTRL GSI

WRFVAR

Analysis field at 0600 UTC for reflectivity: more significant increment in WRFVAR than GSI

Dual-Pol Radar Data Assimilation: GSI vs. WRFVAR

GSI WRFVAR

16

Analysis field at 0600 UTC for low level temperature: more significant temperature change in WRFVAR than GSI

Dual-Pol Radar Data Assimilation: GSI vs. WRFVAR

GSI WRFVAR

17

Analysis field at 0600 UTC for low level moisture: higher value of moisture in GSI field than WRFVAR

Dual-Pol Radar Data Assimilation: GSI vs. WRFVAR

NEXRAD

GSI

18

WRFVAR

1200 UTC 2 September 20136 h forecast: similar location and storm pattern

Dual-Pol Radar Data Assimilation: GSI vs. WRFVAR

NEXRAD

GSI

19

WRFVAR

1500 UTC 2 September 20136 h forecast:different pattern Storm dissipate quicker than WRFVAR

Dual-Pol Radar Data Assimilation: GSI vs. WRFVAR

Work Plan

Generate a dual-frequency radar and microwave radiometer observation dataset, analyze GPM data

-- Ground validation data now available

Develop a methodology for assimilation of GPM DPR and GMI observations with GSI

-- Test first with ground validation data

Assimilation of real GPM data

Current Work Case study: 2012-02-24 snowstorm observed by GPM

ground validation GCPEx field campaign.

WRF model control run NEXRAD

1700 UTC 24 February 2012