Basis for integration of conventional observations of cloud into global nephanalyses

A. Henderson-Sellers*, K. McGuffie

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

6 Citations (Scopus)


Water vapour is the largest and, radiatively, the most important trace gas in the Earth's atmosphere. Cloud amount and cloud optical depth feedbacks are, as yet, poorly understood and improvements in model parameterization schemes await an adequate observational data base. Satellite retrievals do not, and will not for some time, provide more than 'snapshot' (a few months to a few years) records. Conventional surface-based observations of cloudiness could complement the global coverage offered by satellite retrievals if a sound relationship between the two observational measurements could be found. Observations underline the importance of the vertical dimension of clouds which affects the observational geometries from satellites and the surface. A new basis for the relationship between the (vertical) earthview of cloud amount and the (whole dome) skycover of cloud amount has been sought. Over four and a half thousand all-sky camera photographs, representing a considerable range of seasonal and climatological conditions, have been analyzed to give rise to a database from which predictive relationships for earthview, E, and skycover, S, have been established. Cubic functions are the most soundly based both physically and empirically. We find: S=0.647+2.192 E-0.461 E2+0.037 E3 and E=-0.001+0.082 S+0.269 S2-0.019 S3 for the prediction of skycover from earthview and earthview from skycover, respectively. If earthview is required from skycover observations then E≃S could be used with little additional error. Hence, conventional surface observations of skycover could be compared directly with satellite-derived earthview values. More importantly, these results do not support the widespread assumption that conventional (surface) observations of cloud amount always exceed the earthview could retrieval. Furthermore, climate model predictions of total cloud amount may also be interpreted via these relationships. GCM-predicted layer cloud amounts can be synthesized into 'modelled E' values using the random overlap formula and hence it is possible to construct 'modelled S' values which are directly comparable with conventionally observed cloud climatologies. The 'baseline' observation of skycover by clouds therefore provides a valuable validation tool for both satellite programmes and climate models.

Original languageEnglish
Pages (from-to)1-25
Number of pages25
JournalJournal of Atmospheric Chemistry
Issue number1-2
Publication statusPublished - Jul 1990


  • climate models
  • Clouds
  • conventional observations
  • integration of global data
  • satellites


Dive into the research topics of 'Basis for integration of conventional observations of cloud into global nephanalyses'. Together they form a unique fingerprint.

Cite this