BODC Parameter Dictionary or A Tale of Topsy - PowerPoint PPT Presentation

1 / 22
About This Presentation
Title:

BODC Parameter Dictionary or A Tale of Topsy

Description:

Once upon a time (in 1987) the organisation MIAS had a parameter dictionary ... Many of these differed from the MIAS dictionary codes ... – PowerPoint PPT presentation

Number of Views:75
Avg rating:3.0/5.0
Slides: 23
Provided by: royl5
Category:

less

Transcript and Presenter's Notes

Title: BODC Parameter Dictionary or A Tale of Topsy


1
BODC Parameter DictionaryorA Tale of Topsy
  • Roy Lowry
  • BODC Project Data Management

2
The Prologue
  • Once upon a time (in 1987) the organisation MIAS
    had a parameter dictionary
  • It covered 129 parameters
  • 44 of these were wave statistics
  • 31 of these were meteorological parameters
  • The rest covered basic hydrography plus various
    flavours of sea level

3
The Prologue
  • The dictionary was implemented as part of a
    Codasyl database
  • It was maintained through monthly updates by the
    DBA
  • Change was driven by the need to handle different
    types of long time series data

4
The Prologue
  • The dictionary was based on 8-byte parameter code
    which comprised
  • A 4-byte parameter code
  • A 2-byte parameter subgroup
  • A 2-byte parameter discriminator

5
The GF3 Era
  • As part of the GF3 format development a parameter
    code table was developed
  • This was doomed to failure because the
    maintenance overheads (as a printed manual) were
    prohibitive
  • The code table was based on an 8-byte code, four
    of which described the parameter

6
GF3 Standardiation
  • GF3 provided international standard 4-byte
    parameter codes
  • Many of these differed from the MIAS dictionary
    codes
  • So the MIAS dictionary was updated to bring it
    into line with GF3

7
Unexpected Problems
  • However, the pre-GF3 codes were in widespread use
  • Updating the codes in the data holdings was an
    impossible task
  • Fifteen years on and we still have two different
    codes in our data holdings for sea temperature

8
MORAL
  • Dont mess around with an operational dictionary
    unless it is absolutely ESSENTIAL

9
Project Data Management
  • Early projects managed by BODC (North Sea Project
    and BOFS) broadened our horizons beyond time
    series and CTDs
  • However, they were supported by databases using
    simple non-parameterised data structures

10
Project Data Management
  • BOFS exposed the weaknesses of the simple data
    structures
  • Significant work went into database and
    associated documentation maintenance
  • OMEX threatened even more parameter diversity
  • A normalised approach was therefore adopted

11
A Dictionary is Born
  • A normalised database is based on parameter codes
  • Parameter codes need a parameter dictionary
  • No suitable dictionaries were available off the
    shelf
  • So, in 1994 I set about building one

12
Design Constraints
  • Two dictionaries in one organisation is madness
  • The dictionary had to be compatible with the MIAS
    dictionary, which meant
  • Parameter codes had to be 8 bytes (-)
  • Codes had to conform to the code/subgroup/discrimi
    nator model ()
  • Units linked to first four bytes ()

13
Design Principles
  • Systematic code naming is a secondary
    consideration (its nice to be consistent but)
  • Multiple units for a single parameter is avoided
    if at all possible (which brings us to..)

14
The Curse of the Kilogram
  • All water column parameters in the dictionary are
    defined in units of per litre and NOT per
    kilogram
  • Why?
  • Measurements are made per litre
  • T/S data are not always available
  • Consequently fixed factor conversions are used

15
How do We Get Away With it?
  • Wherever the T/S data are available an additional
    parameter is stored in the database that effects
    the per litre to per kilogram conversion
  • Consequently, we can always deliver per litre,
    but can also deliver per kilogram wherever this
    is possible

16
Enter Topsy
17
Putting it Another Way
18
How has it Been Done?
  • Dictionary maintenance is the responsibility of a
    single individual
  • The science is the important part, not the
    hieroglyphics
  • Great care is taken to ensure that the data are
    understood before they are coded

19
Dictionary Implementation
  • Implemented as four tables in a relational
    database
  • 8-byte code definitions (ZUSG)
  • 4-byte parameter name definitions (ZUPM)
  • Parameter category grouping (ZUCT)
  • Units definitions (ZUNT)

20
Dictionary Interrogation
  • Several tools available within BODC
  • ProFortran Search Tool
  • Oracle 2000 Form
  • Access Forms Interface
  • All tools support wildcard searches

21
Dictionary Availability
  • The four tables that make up the dictionary are
    available by FTP
  • ASCII (CSV) dumps of the dictionary tables
  • Data files are automatically updated every Sunday
    (contrast with GF3)
  • Updates are restricted to BODC areas of interest

22
Conclusion
  • A parameter dictionary has been built containing
    over 6000 codes
  • It is successfully underpins the most diverse
    oceanographic database in the world
  • It is freely available to anyone who wants to use
    it providing they dont expect BODC to modify it
    to their requirements
Write a Comment
User Comments (0)
About PowerShow.com