science guidance is needed to assess errors in long-term time series and correct or compensate for physical changes in long time series.

  1. CDR stewardship is needed throughout the entire process. A critical requirement is that the science overview and involvement is end-to-end and that oversight will be needed continuously for the duration of the period to ensure that the CDR is generated and maintained.

  2. Reprocessing and reanalysis is an ongoing activity. The development of a CDR is often an evolutionary process and repeated reprocessing of the entire input dataset is often necessary. The CDR algorithm will change over time as measurement hardware and software and understanding of the measurement process evolve.

  3. Raw data must be archived. The reprocessing requirement for satellite data means that the raw satellite data must be archived for future reprocessing.

  4. Multiple, independent processing can provide important insights into the CDR. Multiple efforts provide information on the precision of a CDR, and biases based on one processing method may not be obvious without comparison with other independent methods.

  5. Data management and oversight are critical components of a successful program. CDR datasets should be managed in a way that allows easy, meaningful access by users of varying technical sophistication. Web access could include browse images, read software, subsetting tools, and online plotting

  6. Validation and overlap of successive satellite missions is critical for developing consistent CDRs over time. Satellite measurements are by their very nature “remote” and thus in situ observations are needed to validate remotely sensed data and monitor sensor degradation, while overlap is needed to reduce satellite biases

  7. Orbital and instrument decay need correction for consistent long-term CDRs. Another source of change that requires in situ validation measurements is caused by orbital and instrument degradation decay. This can change the atmospheric path length for each instrument, alter the Sun-Earth sampling geometry, or introduce aliasing of the diurnal cycle.

The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement