sandbox/dek/09_adjust/README.md

In dankelley/argoFloats: Analysis of Oceanographic Argo Floats

Hypothesis 1: all "adjusted" data are NA for realtime data.

Hypothesis 2: some "adjusted" data are non-NA for delayed data.

The adjustment process is complicated. There are two cases, depending on whether a PARAMETER_DATA_MODE is present in the metadata stored in the netcdf file.

1. Case 1. If the PARAMETER_DATA_MODE field is present, then its columns are mapped to the parameter columns, and the following action is undertaken.

2. If PARAMETER_DATA_MODE is "A" (meaning "Adjusted") then (fill in ).

3. If PARAMETER_DATA_MODE is "D" (meaning "Delayed"), use the Adjusted values.

4. If PARAMETER_DATA_MODE is "R" (meaning "Realtime") then use the raw values, even if the file contains data with <PARAM>_ADJUSTED (e.g. "TEMP_ADJUSTED", which is the name used in netcdf files for the adjusted temperature) in their names.

5. Case 2. If no PARAMETER_DATA_MODE field is present, the DATA_MODE field is used. This does not apply to individual columns, so the action depends on the data within the columns.

6. If DATA_MODE is "D" then each data type is considered in turn, and if the <PARAM>_ADJUSTED values are all NA then the raw data are used, but if any of the <PARAM>_ADJUSTED values are not NA, then all the <PARAM>_ADJUSTED data are used.

7. If DATA_MODE is "R" then then use the raw values, even if the file contains data with <PARAM>_ADJUSTED in their names.

dankelley/argoFloats documentation built on April 18, 2024, 5:13 a.m.