Data Processing Levels
Data processing levels are defined as as attribute called processing_level. This attribute is defined in ACDD-1.3` as a global variable. FAAM recommend this be applied at the variable level unless all variables in a file or group share the same processing level, in which case it may be applied at the global or group level.
Processing Levels
Level 0-0
Unadulterated raw instrument data in the native instrument format. File/s may be compressed and/or combined with level 0-0 data from other instruments or time periods.
Level 0-1
Unprocessed instrument data at original resolution. Data may be flagged to designate data artefacts, calibration periods, pre- and post-flight tails, etc. Duplicates or exceptional empty data may be removed. Data may be converted to a different file format and metadata added. Automatic quality control may be applied.
Level 1A
Data variable derived from a single level 0 or 1A data product. An Automatic level of quality control has been applied to the process inputs and/or the derived output.
Level 1B
Data variable derived from multiple level 0 or 1A data products. An Automatic level of quality control has been applied to the process inputs and/or the derived output.
Level 2A
Level 1A data that has been irreversibly transformed through changes to the dimensions. This may include temporal changes through interpolation, averaging, or sub-sampling or spatial changes by aligning to a regularised geophysical coordinate (latitude, longitude and altitude). An Automatic level of quality control has been applied.
Level 2B
Level 1B data that has been irreversibly transformed through changes to the dimensions. This may include temporal changes through interpolation, averaging, or sub-sampling or spatial changes by aligning to a regularised geophysical coordinate (latitude, longitude and altitude). An Automatic level of quality control has been applied.
Level 3A
Level 1A or 2A data that has undergone a Standard level of quality control.
Level 3B
Level 1B or 2B data that has undergone a Standard level of quality control.
Level 4
Level 4 data is derived from Level 3A or 3B data spanning an extended period of time or from different platforms. For example, this may be derived from data from an entire campaign or collected over many flights or from multiple aircraft and/or ground sites. Level 4 data requires a Comprehensive level of quality control.
Quality Control levels
Automatic
Data quality is characterised based on instrument operating envelope, for example acceptable probe internal temperature; platform state, for example whether the aircraft is in the air or on the ground; or data values, for example negative data when this is not physically possible. This level of quality control is done automatically during processing based on a series of data- or instrument-specific criteria.
Standard
Data is checked based on the reasonableness of the physical quantity being measured at the time it is measured. Where possible, data are cross referenced with comparable or complementary physical variables from the same flight. For example, the static pressure may be checked against the aircraft measurement of the same property as well as being within some range of the standard atmosphere pressure given the altitude derived from the GIN. This level of quality control may be carried out automatically in the processing code or may require human intervention and manual flagging. At this level Automatic quality control is implied.
Comprehensive
An in-depth evaluation of data quality has been undertaken by an expert. This may include quantitative analysis of instrument state and behaviour over multiple flights. Manual QC flags are applied where data are found to be of degraded quality. At this level Standard quality control is implied.