About

The monthly temperature forecasts (outlooks) are made using the Empirical Climate Prediction System (ECliPS). It is built on the Climatic Expert System for the North Atlantic (CESNA), developed by Rodionov and Martin (1996, 1999). Due to existence of so-called "teleconnections," or linkages between climate anomalies over great distances, an accurate forecast in one area is practically impossible without taking into account climatic processes in other areas of the world, as well as external factors, such as solar activity. Therefore, while the forecasts here are focused on the contiguous United States, this is a true systems approach. In this respect, ECliPS can be considered as a knowledge management system designed to organize and process data and information about the entire climate system. Currently, ECliPS's knowledge base contains more than 2500 empirical rules, and this number is growing.

The forecasts are presented as probabilities of 1) above or below normal temperatures (departures from the 1981-2000 base period) and 2) temperature changes from the same month of the previous year (year-over-year forecast). The latter is particularly important in the areas where temperature experiences a significant trend. For example, a warming trend in the Southeast increases the probability of the above normal category, while changes from year to year remain practically unaffected by the trend. 

These are two-category forecasts, and the categories on the maps are marked as follows:

A – above normal,
B – below normal,
W – warmer than last year,
C – colder than last year.


The areas between A and B, may be marked either as N (near normal) or EC (equal chances). The difference between N and EC is that the former conveys more confidence that the mean monthly temperatures will be close to the averages for the base period, while the latter indicates that the forecast is simply uncertain regarding the temperature category and says nothing about the magnitude of the anomaly. Similarly, the areas between W and C may be marked either as NC (no change) or EC (equal chances that the temperature will go up or down). The contour colors indicate the most likely temperature category, either above-normal/increase (red) or below-normal/decrease (blue) with the lines giving the probability (p >50%) of the more likely category. In each point on the map the probability for the opposite category is 100% - p. The higher probability of the temperature category means a higher confidence in the forecast of this category. It may also mean a higher magnitude of temperature anomaly (or temperature change), although this is not always true.

The forecasts are prepared by the 15th of each month with the lead time from 0.5 month to 10.5 months (a total of 11 forecasts). By the end of each month, the forecast for the next month may be adjusted (zero lead time), if necessary.

Sergei Rodionov, Ph.D.
Climate Logic, LLC