Homogenizing Surface and Satellite Observations of Cloud. Aspects of Bias in Surface Data

摘要

A preliminary investigation into certain forms of bias known to affect surface observations of clouds was made using two separate months of global cloud reports. A search for evidence of nighttime detection bias gave variable results but over the analysis region of Scandanavia, where the evidence would be strongest, due to the ability to compare reports over short and long nights, there was good reason to believe that observers were consistently failing to detect cirrus (particularly) and altostratus at the local scale as well as over larger areas. If such bias prevalent over many regions of the world of value of surface-derived climatologies must be reduced. Finally, the global data were assembled into maps depicting frequency of occurrence, contingency probabilities and level amounts. The data highlighted such features as monthly variability compared to long-term statistics and the paucity of oceanic data, which are overall, unreliable as a result. The land data are more representative and may be of use when results from ISCCP are released. Keywords: Okta observing scale.