CF Unleashed Introduction to CfRadial Joe Van Andel

CF Unleashed: Introduction to Cf/Radial Joe Van. Andel National Center for Atmospheric Research vanandel@ucar. edu 2013/1/8 The National Center for Atmospheric Research is sponsored by the National Science 1 Foundation

CF -What is it? • Climate and Forecast Metadata Convention • Locate data in space–time and as a function of other independent variables, to facilitate processing and graphics • for – model-generated data – observational datasets • originally no support for radial data 2

Motivation • Support radar/lidar community • Provide better interoperability for data providers and tool creators • provide libraries and tools to – read/write datasets – convert from legacy formats – display radial data 3

What is Cf/Radial? • CF extensions for radial radar/lidar data • CF compliant net. CDF – net. CDF is familiar to many scientists – CF allows assimilating data into forecast models • Details: http: //www. ral. ucar. edu/projects/titan/docs/ radial_formats 4

What Kind of Instruments? • • Scanning Staring Vertically Pointing Mobile or Fixed 5

Stationary Scanning 6

Mobile/Scanning 7

Airborne Scanning Radar 8

Airborne Vertically Pointing 9

Fixed, Staring Profilers 10

Advantages • Readable on Linux, OS X, Windows • Supported by multiple languages: – C/C++ – Fortran – Python – IDL – Matlab 11

Advantages (2) • Byte order independent • Efficient storage: – supports variable gates/beam – native compression 12

Staggered 2 -D Storage 13

Transparent Compression • Net. CDF 4 uses HDF 5 as storage layer • Net. CDF 4/HDF 5 supports transparent compression • compressed files can be as small as 20% of the original size (data dependent) 14

Sample Surveillance Data 15

Range Height Radar Data 16

Lidar Data 17

Data Content Overview • data fields (moments) • each moment is a collection of “Rays” • each ray has range gates 18

Metadata • each ray has metadata: – range – elevation – latitude – longitude – altitude – time 19

Metadata(2) • Moving platforms: – heading – roll – pitch – rotation – tilt 20

Coordinate Conventions • Fixed location • Mobile ground-based (truck mounted radars) • Airborne radars/lidars 21

Current tools • Radx C++ library • Conversions: Read/Write • • • Cf. Radial DORADE (legacy binary radar format) Universal Format (UF) - very old binary format NEXRAD level 2 archive Foray 1 (EOL net. CDF) SIGMET (writing is not supported) 22

Future work (1) • Incorporating metadata from NODC net. CDF templates • Incorporating metadata from Net. Cdf Attribute Convention for Data Discovery 23

Future work (2) • convert SOLO (display/editing tool) to read/write Cf. Radial • Python libraries and tools • Matlab libraries • IDL libraries • Community archive of user contributed libraries and tools. 24

Who is using Cf. Radial? • • NCAR UNIDATA DOE/ARM NOAA/NSSL EEC (commercial radar vendor) Pro Sensing (commercial radar vendor) Various university users 25

Conclusion • Want to encourage more users to adopt Cf. Radial • We welcome your ideas on how to publicize this format NCAR is supported by the National Science Foundation. 26

Acknowledgements Co-Authors: Mike Dixon Wen-Chau Lee Bob Rilling Chris Burghart 27

Questions? ? ? ? 28