New Techniques for Filling Gaps in Instrument Records

Summary

The Atmospheric Radiation Measurement (ARM) Climate Research Facility’s Southern Great Plains (SGP) site in Lamont, Oklahoma, is home to one of the longest records of actively sensed cloud information anywhere in the world. Despite the best efforts of facility staff, however, instruments occasionally fail or are taken down for maintenance, resulting in holes within the observational record.  These gaps lead to uncertainty in monthly statistics of observed variables such as cloud fraction that are often used to evaluate model simulations or diagnose trends in the observations.

Researchers funded by the Atmospheric System Research (ASR) program used a statistical technique called self-organizing maps (SOM) to reduce uncertainties in the instrument record. The analysis took advantage of the fact that cloud occurrence is partly controlled by the large-scale environment and that the long time series of ARM measurements allows robust classification into meteorological regimes. Testing a number of SOM configurations, the analysis showed that uncertainty in the monthly total cloud fraction record can be reduced significantly and that the largest gain is provided by SOMs that have a large number of classes and separate data by month.  Using the new technique, uncertainty in monthly total cloud fraction was reduced in half from previous values.

This proof-of-concept work opens the door to a number of other opportunities. The methodology is adaptable to other ARM sites. Further, the results suggest that a combination of ARM observations and reanalyses can provide a better historical record of cloud occurrence prior to the existence of actively sensed observations. Finally, this work can move beyond cloud fraction and the techniques can be applied to other variables such as records of specific cloud types.

References

Kennedy, A. D., X. Dong, and B. Xi. 2015. “Cloud Fraction at the ARM SGP Site: Reducing Uncertainty with Self-Organizing Maps,” Theoretical and Applied Climatology, DOI: 10.1007/s00704-015-1384-3.