In the early and middle parts of this century, weather forecasters really began understanding the distribution of weather and using it to advantage in weather forecasting. The foundation of this understanding was the concepts of air masses and fronts. Today, fronts remain as a fundamental concept, but it is rare to see any mention of air masses beyond introductory meteorology or climatology textbooks. Nevertheless, they are still used extensively in weather forecasting. Although they represent a simplification of the atmosphere, they are a very useful simplification.
The critical idea here is that weather conditions are not randomly distributed over the globe. Variations in weather elements, such as temperature, tend to be concentrated in narrow, almost two-dimensional bands, called fronts. Between these bands, weather elements change very gradually, and these broad, nearly uniform regions of the lower atmosphere are called air masses.
Consider the surface weather map shown above, from 19Z 24 Sep. 1996. Three types of fronts are shown, coded by color: a cold front (blue), a warm front (red), and an occluded front (purple). My apologies for the jagged computer-generated plotting: real fronts are much more smoothly curved. These fronts are fairly typical in that they represent zones of concentrated temperature gradients.
Scan the map, starting in the upper-left corner, in North Dakota. Temperatures (plotted to the upper left of each station) are in the upper fifties and low sixties. As you scan southeastward, into Missouri, the air becomes gradually warmer, approaching 70 F. But then in Mississippi, and most of the rest of the southeast, temperatures are in the mid eighties. So that's about a ten degree temperature difference from North Dakota to southern Missouri, followed by a fifteen degree temperature change across the cold front.
In central Texas, temperatures are different, but the same basic pattern is present. Temperatures are in the low seventies on the north side of the front, and in the low nineties on the south side. Along with the changes in temperature, there also tend to be changes in the pressure gradient and winds, caused by the density variations within the lowest 2-4 km of the atmosphere.
While in general temperatures vary rapidly across a front and slowly elsewhere, exceptions are frequently found. Look at West Virginia, with temperatures in the low sixties, while behind the cold front in Indiana, temperatures are in the upper sixties and low seventies. Behind the cold front, it's supposed to be colder, not warmer! What gives?
Assuming the frontal analysis is correct (which it is), the variations in large-scale (air mass) temperatures are frequently masked by the sorts of things that affect diurnal (daytime/nighttime) temperatures, such as clouds, rain, and wind. In this instance, it's cold in West Virginia because it's cloudy and raining there, while it's sunny in Indiana. At night, it will probably still be in the sixties in West Virginia, while Indiana will have cooled off to the upper forties.
Click on one of the two surface map sites below and look carefully at the current surface map. Think about the following questions: