Radar Reflectivity Quality Control and 3 D Mosaic

Radar Reflectivity Quality Control and 3 -D Mosaic (and ideas for eventual use of data in WRF GSI) 4 January 2005

Code Functionalities Input of Level-2 data (single radar compatible) n Regrid raw radar volume scan to 1° x 1 km spherical grid n Produce “hybrid scan” of base reflectivity to address beam-blockage issues n Perform QC n Produce 1 km cartesian grid from QC’d data n

Reflectivity QC Techniques n n Ø Terrain-based u Assuming normal propagation u Using high-resolution DEM data u Calculate occultation based on radar power density u Create a “hybrid scan” lookup table 3 D radar fields structure-based u Vertical reflectivity gradients u Horizontal reflectivity structure PURPOSE: To remove non-precipitation echoes

Terrain Based QC l All reflectivity bins below a terrain-based hybrid scan (O’Bannon, 1997*) are removed. The hybrid scan bins are the lowest ones which meet the following criteria: * Bottom of the bin must clear the ground by at least 50 m (500 ft for ROC). * Less than 50% power blockage at any shorter range bin in the same radial. * O'Bannon, 1997: Using a 'terrain-based’ hybrid scan to improve WSR-88 D precipitation estimates. Preprints, The 28 th International Conference on Radar Meteorology, September 7 -12, 1997, 506 -507

New Hybrid Scans Using High-resolution Terrain Power density distribution & beam blockage Beam geometry vs. terrain

Structure-based QC: Possible Nonprecipitation echoes Echo Ground/Sea Clutter Sun strobe Chaff Clear air returns Birds and insects Electronic interferences Non-precip clouds

Reflectivity QC Variables Z SIGN Median Texture Vert. Grad. SPIN Reflectivity Structure: KFWS, 4/20/1995, 04: 53 Z

Some Notes… u SPIN is not very useful since it shows very similar characteristics for clutter and around convective storms u SIGN and Texture are very similar u Median is similar to raw reflectivity u Vertical gradient is very useful, especially at close ranges

QC Decisions A reflectivity datum is removed if P_nonpcp > P_thresh

Reflectivity QC Before KFWS 1995 -04 -20. 0453 Z After

Reflectivity QC Before KFWS 1995 -04 -20. 1850 Z After

Reflectivity QC Before KTLX 2002 -10 -18. 1246 Z After

Future Use of Reflectivity Data in GSI

n Separate “cloud analysis” system u n n n Use Z-R/Z-M relationships to diagnose hydrometeor concentrations Nudge vertical momentum via 3 -D divergence constraint u Variational approach to w component Nudge (I know, bad word) to maintain rh/latent heat release to support w-component Similar to FSL’s Hot Start and CAPS ADAS cloud analysis

n Eventually: define error covariance matrices (? ) for use of reflectivity in a “true” 3 -D sense (not a diabatic or nudging technique, per se) n FIRST THINGS FIRST: MOSAIC/QC package running operationally at NCEP