Nimbus Satellite Data Rescue Project for Sea Ice Extent: Data Processing

Early Nimbus satellites collected both visible and infrared observations of the Earth at high resolution. Nimbus I operated in September, 1964. Nimbus II operated from April to November 1966 and Nimbus III operated from May 1969 to November 1969. We will discuss our procedures to recover this data into a modern digital archive useful for scientific analysis. The Advanced Videocon Camera System data was transmitted as an analog signal proportional to the brightness detected by a video camera. This was archived on black and white film. At NSIDC we are scanning and digitizing the film images using equipment derived from the motion picture industry. The High Resolution Infrared Radiance data was originally recorded in 36 bit words on 7 track digital tapes. The HRIR data were recently recovered from the tapes and TAP (tape file format from 1966) files were placed in EOSDIS archives for online access. The most interesting parts of the recovery project were the additional processing required to rectify and navigate the raw digital files. One of the artifacts we needed to identify and remove were fiducial marks representing latitude and longitude lines added to the film for users in the 1960's. The IR data recording inserted an artificial random jitter in the alignment of individual scan lines. We will describe our procedures to navigate, remap, detect noise and remove artifacts in the data. Beyond cleaning up the HRIR swath data or the AVCS picture data, we are remapping the data into standard grids for comparisons in time. A first run of all the Nimbus 2 HRIR data into EASE grids in NetCDF format has been completed. This turned up interesting problems of overlaps and missing data issues. Some of these processes require extensive computer resources and we have established methods for using the 'Elastic Compute Cloud' facility at Amazon.com to run the many processes in parallel. In addition we have set up procedures at the University of Colorado to monitor the ongoing scanning and simple quality control of more than 200,000 pictures. Preliminary results from September 1964, 1966 and 1969 data analysis will be discussed in this presentation. Our scientific use of the data will focus on recovering the sea ice extent around the poles. We especially welcome new users interested in the meteorology from 50N to 50S in the 1960's. Lessons and examples of the scanning and quality control procedures will be highlighted in the presentation. Illustrations will include mapped and reformatted data. When the project is finished a public archive from September 1964, April to November 1966 and May to December 1969 will be available for general use.