While graphical records of astronomical/meteorological events before telescopic observations are of particular interest, they have frequently undergone multiple copying and may have been modified from the original. Here, we analyze a graphical record of the cross-sign of 806 CE in the Anglo-Saxon Chronicle, which has been considered one of the earliest datable halo drawings in British records, whereas another cross-sign in 776 CE has been associated with the aurora. However, philological studies have revealed the later 806 event is derived from Continental annals. Here, records and drawings for the 806 event have been philologically traced back to mid-9th Century Continental manuscripts and the probable observational site identified as the area of Sens in northern France. The possible lunar halos at that time have been comprehensively examined by numerical ray tracing. Combined with calculations of twilight sky brightness, they identify a visibility window supporting monastic observation. Cruciform halos are shown to be fainter and rarer than brighter and more commonplace lunar halos. Physically credible cloud ice crystal variations can reproduce all the manuscript renditions. The manuscript records prove less than desirable detail but what is presented is fully consistent with a lunar halo interpretation. Finally, the possible societal impacts of such celestial events have been mentioned in the context of contemporary coins in Anglo-Saxon England and the Carolingian Empire. These analyses show that we need to trace their provenance back as far as possible, to best reconstruct the original event, even if graphical records are available for given astronomical/meteorological events.