The winds aloft at about 500 mb (18,000 ft) play a key role in the winds we ultimately feel on the surface. The relationship is discussed in our textbook Modern Marine Weather. In short, the direction of the winds aloft tell us which way the surface Lows will approach us, and the speed of the winds aloft are a measure of the speed of the surface Lows (about 30 to 50% of the winds aloft speeds) as well as the severity of the surface winds we can expect. The seminal explanation of the practical relation between winds aloft and surface winds is by Joe Sienkiewicz and Lee Chesneau.
Short of having an actual map of the winds aloft (which is easy to come by if we have wireless connections), we can gage both speed and direction from visual observations. Again, I refer to our text for details, but one way to spot direction is to note that the winds make waves in high cirrus clouds, much as wind makes waves in water. These waves show up as fine ripples in cirrocumulus clouds, called mackerel sky. The direction of the winds are, as is the case with water waves, perpendicular to the waves, inline with the motion of the waves.
The speed of the winds aloft can be judged from other cirrus cloud patterns, such as prominent mare's tails. The very existence of a prominent and persistent wave pattern is some indication of at least a well defined wind pattern, but a more positive sign is the topic at hand. Namely, if you can actually see the high cirrus clouds moving, then the winds must be fast. We mention this in several books, but this is not a very common observation, although we do indeed often see cirrocumulus in neat wave patterns.... as if waves in sand—as opposed to the much broader and more common patterns seen in altocumulus waves.
Note that most internet discussion and many books and some glossaries assign "mackerel sky" to altocumulus patterns, and there is no particular problem with that, but those clouds (waves in altocumulus) are not forecasters of strong winds, and not what was meant in the old saying "Mackerel skies and mare's tails make tall ships set low sails." This refers to cirrocumulus patterns, indicators of the winds aloft. Altocumulus patterns are determined mostly by boundary layer winds (~850 mb) that reflect what is going on now, not what is coming over the horizon.
Actually, let me take that back. There is a problem with the misidentification of the clouds that go with that famous forecasting jingle. Namely it dilutes our ability to do useful onboard forecasting based on what we see alone without weather maps or official forecasts at hand. In short, to use modern jargon, such discussion is fake news that can lead us astray. In any event, once you see that error in place, it is fair warning to be careful about other things stated in that reference. Sometimes we need to make the difficult distinction between high-end altocumulus and cirrrocumulus. See notes at the end about cirrocumulus v. altocumulus cloud waves. The importance of cirrus clouds for forecasting was known in the late 1800's.
The difficulty with spotting the cloud motion at sea is we need a reference to judge the motion by. Here on land, and in the sample below we have a building to use.
To follow up on this observation, which was made in Seattle at 1700 PDT, Sept 25, 2020, (00z on Sept 26), we can go to the FTPmail folder of weather maps. Strangely enough, the OPC, which makes these maps, do not post the most recent 500-mb map on their own web page. They include a 24h and 48h forecast instead, but we can get the map valid at 00z 9/26 from the link above. At that link, go to Pacific / Upper Air Charts / 500 mb / Most Recent, which is shown below.... you will of course not see this when you do it as this will no longer be the most recent!
We see that the winds aloft we were viewing to the S at the moment could be as large as some 85 kts (exact values not clear as they are not labeled overhead and I have not tried to compute how far south these were at 40º high in the sky). The model winds on this map, however, are indeed high winds aloft speed (average may be closer to 60), although they can be much higher (>100 kts), but on the average they are lower.
If we look at the surface analysis at this time, we do not see much activity as shown below.
By this time, the winds aloft have backed down to the SW from the westerlies we were observing, a direction you can discern in the surface forecast from the orientation of the isobars in the warm sector of the accompanying frontal wave—and they are predicted to get even faster.