Generic Herschel data product levels are defined as following:
Level-0 data products: Raw telemetry data as measured by the instrument, minimally manipulated and ingested into the mission data base/archive. Typically, readings are in binary units versus detector pixel number.
Level-1 data products: Detector readouts calibrated and converted to physical units, in principle instrument and observatory independent. It is expected that level-1 data processing can be performed without human intervention.
Level-2 data products: Further processed level-1 data to such a level that scientific analysis can be performed. For optimal results many of the processing steps involved to generate level-2 data may require human intervention, based both on instrument understanding as well as understanding of the scientific aims of the observation. These data products are at a publishable quality level and should be suitable for virtual observatory access.
Level-3 data products: These are the publishable science products where level-2 data products are used as input. These are products not only from the specific instrument, but are usually combined with theoretical models, other observations, laboratory data, catalogues, etc. Their formats should be virtual observatory compatible and these data products should be suitable for virtual observatory access.
All observations made after the first year of the routine phase will have a proprietary period of 6 months, after which the observation will become public. The proprietary time applies to each observation individually, counted from the day when the data are made available to the initial data owner. However, a scheme will be put in place whereby the Herschel Project Scientist and the HOTAC Chair in consultation can grant additional proprietary time to certain large programmes, in order to prevent the release of improperly or inhomogeneously calibrated or processed data. Note that data resulting from routine calibration observations will generally enter the public domain immediately after they are processed unless duplicating a science observation.