In our every day's use of time we make a clear distinction between time (e.g. 5 o'clock) and date (e.g. 1.1.1999). By 'date' we mean the actual time measured in days passed since the date 1.1.0000. With 'time' we use a cyclic measurement of 24 hours ( = 86400 seconds = 794243386928000 oscillations of a cesium atom). This repeating measurement makes it easy to refer to a different time as for example when we make an appointment the same time tomorrow.
Since time does not have a start nor end point, we need to define some anchor point or multiple anchor points to which we can refer in our measurements. One was mentioned above, the date 1.1.0000, often referred to as the year of birth of Christ. But there are several others. In palaeontology the time span considered is so huge that the term 'present', however undefined it is, can be considered as a time anchor. Here the occurrence of certain species in sediments is also used as relative time anchors. The 'big bang' is sometimes considered as an anchor or starting point e.g. in geology, physics and astronomy. For different time scales different anchors and resolutions have been defined, starting from astronomical and geological time to paleontological and biological time until the finest time intervals considered in atomic oscillations and processes. Table 2.1 shows some of the common time anchors.
Anchor | Examples |
'Big bang' | assumed start of the universe |
01.01.0000 | Start of the Year of Birth of Christ |
01.01.1900 | Start time in the DOS/Windows operating system |
01.01.1904 | Start time in the Macintosh operating system |
02.12.1967 | my birthday |
01.01.1970 | Start time in the Unix operating system |
Solar year | |
Lunar eclipse | indigenous peoples |
Month | |
Day | starts at midnight (of time zone) |
Hour | |
Minute | |
Second | |
Present | Anchor in Palaeontology |
As it is the case with spatial data, data with temporal characteristics can have both exact measurements or boundaries or less precise specifications. At first, the range of temporal accuracy seems to be much larger going from sub-second measurements to an accuracy of several hundreds of million years in geology. But the corresponding values for spatial measurements also range from micrometers to maybe several thousand kilometers. Comparing the order of magnitude for time of about 1016 to the one for space of around 1010 there is only a difference of a factor of about 1:1 million. This should not cause a computational problem.
Originating in the different time scales used in the various disciplines several 'time systems' have been defined. In today's geographical research we often come across terms such as time zone, local time, UTC or standard time, solar time and more specific terms as for example GPS time. They can be considered as the equivalent in time for the different coordinate systems used for describing space. An event recorded in one system (e.g. time zone 2) may have to be translated twice if it will be used in another application: once for the (spatial) coordinate system, and once for the time system. The second translation is an aspect completely disregarded by today's GIS technology and will need attention in future temporal GIS research.