| Home > Research > Ifsdocs > OBSERVATIONS > |
IFS documentation Front Page
Chapter 2. Observations: Types, variables and error statistics
Chapter 3. CMA creation (MAKEMA)
Chapter 4. The FEEBACK task
Chapter 5. The TOOLS task
Chapter 6. Central-memory array (CMA) structure/format
Chapter 7. BUFR feedback data structure/format
Chapter 8. SIMULATED-observations data structure/format
Chapter 9. NAMELISTS
Chapter 10. Processing of scatterometer data
REFERENCES
Previous Section
6.1 Computer representation and CMA file structure
All of the information in the CMA file is FORTRAN 64 bits real, in IEEE formats. Some of the word positions contains integer information, but they are stored as their real equivalents. Characters are stored as real equivalents of their Holleriths values. When it is mentioned later on in this chapter that some of the information in the CMA is of the type integer or character, it is important to keep in mind that this means that the information is logically of the type integer or character. Physically all of the information is 64 bits real.
The file structure is a simple bit stream. It is comprised of the following basic record units:
| • data-description record 1 (DDR 1) |
| • data-description record 2 (DDR 2) |
| • dbservation reports |
The extraction of data from the CMA file, and the storing of data in the CMA file, require a machine that supports 64 bits precision for real and at least 32 bits precision for integer. If these requirements are fullfilled the CMA file is machine independent.
Next Section
Previous Section
 |
16.04.2002 |
 |
© ECMWF |
