When determining an average value, a month or year's worth of data (or longer) is retrieved for any given parameter and the arithmetic average is calculated by adding the values and dividing by the number of values involved. A multi-year average can also be calculated by adding up all the individual years' averages and calculating one average that covers a long time period. Then, later months and years can be described as being above or below the long-term (multi-year) average.
For precipitation, monthly and annual total rainfalls are averaged to create a long-term average amount of rain by month and year. These can also be compared to the multi-year average to show how much above or below expected conditions a particular time period is.
Figure A shows how average precipitation can change over different periods of time. From 1971-80, the annual average precipitation in North Carolina is about 52 inches a year, whereas in 1981-90 the average was about 48 inches. In 1991-2000 the ten-year average rose back to about 51 inches. This decade by decade variation could be caused by a number of factors, including changing ocean temperatures, regional droughts, and the passage of tropical storms through the state in certain years. The exact averages depend on the years that are chosen to use for the averaging period. These changes become important when normals are calculated.
A "normal" value is usually taken from the average of a parameter over a 30-year period. This value is updated and statistically adjusted every 10 years to account for changes in the record like stations that move from one location to another or change their type of instrument or time of observation. They are published by NOAA (National Oceanic and Atmospheric Administration). The next new set of normals for 1981-2010 will be published in 2011 or 2012 once all the adjustments have been done.
The "normal" temperature or precipitation for an area, for example, is not always the same as the long-term average, since they are calculated over different time periods. The graph above shows the annual precipitation for North Carolina along with the long-term average in purple and the current "normal" for 1971-2000 in red. Note that the current "normal" rainfall is above the long-term average rainfall for North Carolina since this time period was somewhat wetter than earlier years in the rainfall record.
The standard deviation is the average variation from a long-term average. Most actual values for any parameter (like temperature) lie around an average value plus or minus the standard deviation. For example, annual average rainfall amounts for Raleigh, NC according to the 1971-2000 climate normals are 43.1 inches with a standard deviation of +/- 6.4 inches. Sixty-two percent of all yearly rainfall amounts will fall between the average and standard deviation bounds for Raleigh. Ninety-five percent of all annual rainfalls will fall within two standard deviations of the average value based on statistical principals. Any climate parameter with a large standard deviation will have a lot of year-to-year variability which may make planning more difficult.