Our Boston TV station shows the center of a low pressure area about 200 miles out to sea in the Atlantic. How do they know that? Are there numerous weather buoys at sea to measure such things. If so, how do they keep ships from colliding with them?
One way to identify a low-pressure area is from the cloud patterns in satellite photographs, particularly if it’s a cyclone.
Buoys are well marked on navigation charts. But I know of one situation where a 50 foot sailboat ran into one. (a neighbor) Although, I have no idea how they measure pressure at sea.
They might be able to infer it from sea surface height anomalies measured from satellites, although there are plenty of other things that cause changes in SSH (currents, water temperature, Ekman transport, etc). Basically, low pressure areas would “suck up” on the ocean surface (actually the surrounding high pressure areas “push down” on the surface, causing it to rise in the low pressure areas.). This contributes to storm surges associated with cyclonic storms.
Weather buoys. Speaking as someone who spends a lot of time looking at small objects over vast expanses of ocean I can say that weather buoys aren’t particularly common, at least not in the Indian Ocean off the west coast of Australia.
NOAA uses ships and buoys (both drifting and moored) as the prime information gatherers. Satellites can’t really measure air pressure, but there are some programs that try to analyze satellite data to do so. I’m not sure how advanced their work is, maybe they’re at the point where it’s reliable, but they do use the data from ships and buoys as calibration.
Of course, ships would make sense. They’re all over the place and they can be used in the same way that aircraft make weather reports at routine reporting points.
Here’s a link to NOAA’s Voluntary Observing Ship (VOS) Project.
I was going to say “aircraft”, but then I thought, can aircraft take reliable pressure readings given that they use pressure to work out their altitude (hence the use of “standard pressure” as an assumption, meaning actual altitude isn’t necessarily “barometer altitude”)?
Well, NOAA’s “hurricane hunter” aircraft are used to measure the pressure at the center of hurricanes, so they must have some way to do it.
GPS can give you a good position within 10m vertically. That’s a barometric equivalent of about 0.03". A pressure reading of 29.86 is just about the same as 29.83. In addition, the aircraft may have radar altimeters. I know radar altimeters are used for instrument approaches, but I don’t know if they work at heights of 30,000 ft.
Yes.
Aircraft weather information is used routinely more and more in weather models. Of course, calibration taking into account altitude and other factors must be done, but it needs to be done anyway for accurate altitude and airspeed information anyway. Recent years have seen better and better instrumentation for these purposes.
You’re right an aircraft can’t give an accurate QNH to anyone. For starters they don’t have an instrument that reads pressure directly*. You have to be at a known altitude then set the altimeter to that altitude to get an approximate QNH or you could set the altimeter to zero to get actual pressure. Even if parked on the ground at an aerodrome with a published elevation the derived QNH could be out by up to 2 hpa on a serviceable altimeter. Once airborne the only way it can be done is by matching barometric altitude to radar altitude after compensating for temperature error, but this can typically only be done below 2500’ which is not the sort altitude the type of aircraft that would have a radalt would be flying. Even then you still have the problem that you’re working from actual sea level or ground level and have no way of deriving mean sea level and the possible 2 hpa error may be greater.
I was thinking more of weather reports in a general sense. At nominated published reporting points aircraft give the temperature and wind to ATC along with their position report. They also include anything else that is significantly different from the forecast including cloud, turbulence, icing etc. These reports are used to update or verify forecasts as well as to provide warning of hazardous conditions to other aircraft. Obviously if you encounter un-forecast hazardous conditions you can make your report at any time, you don’t have to wait for a nominated reporting point.
*Although an altimeter does read pressure directly it is not displayed to the pilot directly.
that won’t work. 
to calculate SSH, one needs a weather map to remove the effects of air pressure. So by the time the data is available to users, the effect of barometric pressure have been removed.
Re the OP, worldwide weather models run every 3 hours. They may get some details wrong, but they are fine for mapping surface pressure throughout the world.
The “hurricane hunters” drop “sondes,” little instrument-packed gizmos with parachutes, that send readings back to the plane. According to this, the planes measure static air pressure 6 times per second and the dropsondes measure it twice per second. This is all fed into software to compute the conditions every ten seconds and average them every 30 for their High Density Observations (HDOB) Bulletin. They certainly have a good idea of the air pressure, but of course they are specialized planes and not used for tracking general weather conditions.