Modify the existing role-based permission system to also include filtering of access by processing level metadata. User X should be able to see data sets at processing level Y and above (where Y goes from 0 to 4 per NASA's scheme). Any action that allows full view of data sets should filter appropriately (e.g. there should be no way to get to the data page, download data, etc. that is at too low a processing level. Displays that show the existence of data that cannot be accessed (e.g. the list of search results or data listed by tag, etc.) should eventually at least differentiate the data sets that are accessible. (e.g. there may be cases, such as showing the provenance trail, where our users may think that seeing a thumbnail and knowing that the current data set was derived from something in the ACR - even if they can't access it) is the right behavior. The first part of the task is to deal with the filtering itself. Identifying all of the places where metadata may expose the existence of other data sets can be addressed as a second step.