Patty Miller

1
WRF Data Ingest and Quality Control
Patty MillerSystems Development
DivisionForecast Systems LaboratoryWRF Model
Technical ReviewBoulder, ColoradoJune 19, 2001
2
WRF Data Ingest and Quality Control
Operational Implementation Development Team

• Data Handling and Archive Working Group (WG-12)

NCEP FSL AFWA IBM OU
Dennis Keyser- Lead Patty Miller James
Drake Keith North Keith Brewster
NCEP NCEP NCAR FSL FSL
Geoff DiMego Jeff Ator Francois Van den
Berghe Mike Barth Barry Schwartz
NCEP NCEP
John Derber Bill Collins
Attended 26 Jan 2001 Meeting
3
WRF Data Ingest and Quality Control
Initial Decision The first version of the WRF
data assimilation system will use NCEP-produced
prepared BUFR (PREPBUFR) files and BUFRLIB
software
Probable Default Reality Each organization will
use the data formats and processing software
specific to their facility
4
WRF Data Ingest and Quality Control

• Issues
• Portability
• Community Access
• Format Overlaps
• Ease of Use
• Versatility
• Solution
• A versatile easy-to-use data ingest system that
  supports multiple data formats and also provides
  data files to those who need them

5
Meteorological Assimilation Data Ingest System
(MADIS)
Goal To provide the observational
infrastructure necessary for data assimilation
research and development
Features

• Access to real-time and archived data sets
• FORTRAN callable Application Program Interface
  (API)
• Observational quality control

6
MADIS
Initial Data Sets

• Radiosonde
• Automated Aircraft
• NOAA Profiler Network (NPN) Wind Profiler
• Global Positioning System (GPS)
• Maritime
• Meteorological Aviation Reports (METARs)
• Surface Aviation Observations (SAOs)
• NOAA Profiler Network (NPN) Surface
• Local Data Acquisition and Dissemination (LDAD)
  Mesonet

7
MADIS
Quality Control Checks

• Surface
• Validity, temporal consistency, internal
  consistency, spatial consistency
• Radiosonde
• Validity, wind shear, super-adiabatic lapse rate,
  hydrostatic
• Automated Aircraft
• Validity, internal consistency, temporal
  consistency, position consistency
• Profiler
• Validity, bird contamination, time-height
  continuity checks

8
MADIS
API Advantages

• Invisible data formats
• Access to all observation types
• Multiple retrieval of similar observations
• Variable transformation
• Time windowing
• Domain specification
• Sounding generation
• QC information
• Utility routines
• Portability

9
MADIS
Status

• Alpha version released summer of 2000
• Beta version nearing completion
• Supports both FSL and AWIPS-supplied data sets
• Data feeds to NCAR RAP, NCEP, SDD/MD, and
  MesoWest
• API reviewed and tested by FSL data assimilators
  and NCEP central operations staff

10
MADIS
Plans

• Public release by FY2002
• Upgrade API to include NCEP data formats
• Include more data sets
• Work with Unidata and WRF data assimilators