Yesterday’s letter produced one example of vivid graphic evidence in support of my basic claim re the relationship between precipitable water (PW) and surface air temperatures. What I presented was a limited amount of raw data from purely visual sources, which was enough to display the tight correspondence between major differences in temperatures and logarithmic differences in PW. I am personally convinced that this particular correspondence holds true for all qualified locations outside of Earth’s tropical belt at all times, but I do need to add one more point that should have been mentioned yesterday—this same correspondence does not work over large bodies of open water, where surface air temperatures are always wedded to surface water temperatures. Surface water temperatures respond to all types of energy inputs in ways that differ significantly from responses made by the much more rigid and opaque structure of either land or broad stretches of sea-ice cover. Water readily accepts energy inputs but the method of acceptance differs because of its transparency to solar wavelengths. The processing of energy following capture also differs, mainly as a result of heat transfers made via ongoing movement of an assortment of internal currents.
The PW phenomenon that was on display yesterday was real, and in want of further explanation, which I purposely avoided talking about. How did those great differences originate? That should be a prominent subject of study even when the differences are not as large as those in this dramatic example. Lesser examples should be similarly effective, in accord with logarithmic proportions, and we’ll get into that, but first I want to say more about why the tropical belt is excluded from this discussion, with a few possible exceptions. To begin with, apart from certain large tropical land masses like parts of North Africa, the abundance of PW in the atmosphere is almost always relatively high, usually not far below the saturation level that produces rainfall and becomes breached with regularity. There are not many opportunities to make changes large enough to obtain an unequivocal temperature effect. In addition, while the evaporation rate from tropical water surfaces is relatively strong the great majority of vapor produced never leaves the zone or travels very far afield. The best way to illustrate this fact is accomplished by studying the 5-day animation of Total Precipitable Water at this website: http://tropic.ssec.wisc.edu/real-time/mtpw2/product.php. There is a way to copy and transmit a single snapshot from the site for reference purposes, as you see here:
From a weather standpoint the tropical belt moves up down as the seasons change. Currently it can roughly be defined as the area between 25N and 30S, outside of which there are obviously large differences in both the volume of atmospheric PW and its manner of behavior. Ordinary volumes are quickly reduced and then keep dropping lower and lower as global surfaces recede toward either pole. This type of map is full of distortion, but I believe the actual volumes are still correctly translated, undiluted, by proper spreading technique. (Motion speed, on the other hand, cannot be visually translated as effectively.)
The most noteworthy feature of this map, especially when observing it in animation, is the unique way that PW is distributed in both hemispheres outside of the tropical belt. The streams you see behave in a way that is exactly the reverse of streams and rivers of water on the surface. Here, “upstream” is the concentrated part that has the most massive volume and is found only in a few selected locations. As the contents flow downstream the volumes keep diminishing, such that only a small amount of the original volume remains at the end of the stream, where it disperses. Also see how the point of origin of any one stream tends to move a little every day, or may just call it quits, and how so-called streambeds never stay in one place but keep shifting in all sorts of ways. The original high concentrations follow suit with much variation of their own. Some stay fairly intact over a relatively long distance while others are quickly split up. All things considered, with so many deviations, you can see from the 5-day animation that practically every single spot of mid to upper latitude surface location is likely to experience some amount of overhead flow derived from one of these streams each and every day. On some days the flow will be robust and on others small, occasionally practically nothing, and generally randomized. The range of possible effects for any one location on any one day appear to be open-ended, while resulting long-term averages tend to form everywhere in a common and regular sort of way.
Carl