In any study of climate variability and change, one must pay careful attention to possible artifacts in the records. For instrumental records, these include changes in instrumentation (e.g., rain gauge types), observational practices, station
location, or the surroundings of the instrumental site, or even errors in transcribed data. Proxy records may suffer from errors in dating or interpretation. Even when climate signals are real, it may be difficult to ascribe them to unique causes owing to the complexity of the climate system, a
system which is characterized by myriad interactions between its various components on a suite of spatial and temporal scales (Figure 13.1).