A temperature anomaly is the departure from the average temperature, positive or negative, over a certain period (day, week, month or year). In standard usage, the normal average temperature would be calculated over a period of at least 30 years over an homogeneous geographic region. For example, if the reference value is 15 °C, and the measured temperature is 17 °C, then the temperature anomaly is +2 °C (i.e., 17 °C −15 °C).
Temperatures are obtained from surface and offshore weather stations, or inferred from meteorological satellite data. Anomalies can be calculated for surface and upper-air atmospheric temperatures or sea surface temperatures.
Description
Records of global average surface temperature are usually presented as anomalies rather than as absolute temperatures. A temperature anomaly is measured against a reference value or long-term average.[1][2] Temperature anomalies are useful for deriving average surface temperatures because they tend to be highly correlated over large distances (of the order of 1000 km).[3] In other words, anomalies are representative of temperature changes over large areas and distances. By comparison, absolute temperatures vary markedly over even short distances.
Anomalies are not sufficient to characterize the exceptionality of the temperature values. To take into account the spatial and temporal climatological situation, it is also necessary to calculate the standard deviation from the normal, called "standardized anomaly". Thus a variation of +2 °C can be more significant over a region with normally very stable temperatures than another of +3 °C from a region with strong temporal variations.[4]
Forecasting
Numerical weather prediction provides the temperature forecast for the next few days or weeks. This can be used to calculate anomalies during these forecast periods. There are two types of forecasts, deterministic and probabilistic, which will give different results.
Deterministic data are values obtained by running the forecast model with initial conditions determined by the initial conditions from data assimilation. Probabilistic data comes from predicting sets where the model (or different models) is run several times with a slight variations in the initial conditions each time.[5]
Deterministic anomalies have a standard deviation which depends only on the bias of the forecast. The deviation and the probabilistic anomalies, being calculated from several model solutions, are themselves probabilities that they will occur.
