In Flint, 1633, Chapter 14, Jesse tells Jim, “We need someone to organize a weather service. . . .” In Huff and Goodlett, “High Road to Venice” (GG19), set in 1634, Merton Smith of TransEuropean Airlines calls up the weather service and he has weather information from Rome and Bolzano (and it appears that at one point there were stations in Saxony and Brandenburg). By 1635, there is a weather observation network in part of Russia.
In parts 1 and 2, I discussed the meteorological instruments that can be made based on knowledge in Grantville. Now we’ll look at how those instruments can be integrated into an observation network in the new time line (NTL) created by the Ring of Fire (RoF) event in May, 1631.
This article will from time to time describe meteorological developments in the old time line (OTL). The historical perspective can help us understand how difficult it can be to win acceptance of new ideas and some of the obstacles our characters may encounter in trying to institute a weather observation and forecasting program.
Meteorological Knowledge in Grantville
None of the up-timers majored in meteorology, worked as meteorologists before RoF, or are identified in the grid as having amateur meteorology as a hobby.
Hence, formal educational exposure to meteorology is likely to be limited to a single course, taken as an elective. Meteorology, like astronomy, is a popular science distribution elective for non-science majors but such an elective is likely to be light on mathematics (see, e.g., Ahrens, Meteorology Today). Hopefully, one of more of the up-timer physicists took the more rigorous version for those majoring or minoring in meteorology. It is also possible that one or more of them, or of the chemical engineers, took fluid mechanics.
There are several up-timers with piloting experience in Grantville, and it is important to neither exaggerate nor discount their knowledge of meteorology. The logical starting point is, what would they have needed to know in order to receive a pilot’s license? In general, their study would cover the broad divisions of the atmosphere; the trends of pressure, temperature and density in the standard atmosphere, the relationship of pressure to altitude; the proper adjustment of (and causes of possible errors in) a barometric altimeter; Buy-Ballots law; cloud classification and (qualitatively) formation; measures of sky opacity; causes of turbulence; weather associated with lows, highs, and fronts; local wind and fog phenomena; and how to read a METAR weather bulletin, station plot, or synoptic chart. (PSA; Air Force Manual 11-203, Weather for Aircrews). But they don’t learn how to construct meteorological instruments, how to analyze observations and make a synoptic chart, or how to forecast (except in the most limited sense, as in recognizing the significance of towering cumulus or nimbostratus, or of a fall in pressure).
Naturally, in the course of flying, they will have an intimate opportunity to observe weather in action. An individual pilot with experience may of course have a very good feel for what the weather is going to do in their normal area of operation in the next few hours. But it is not knowledge that we can count on, and it is not knowledge that is easily transferable to others (or to distant places). The experience with aviation weather in the USA is only partially applicable to Europe (admittedly, ex-military pilots may have flown there).
The up-time pilots also received briefings from meteorologists, and there may have been some acquisition of some qualitative guidance (e.g., reference to omega blocking patterns) by osmosis. Again, relevancy depends in part on whether they flew in America or Europe.
For ship pilots it’s only a little different. Look at the weather chapters (34-37) in Bowditch 2002. Besides what we see for aircraft pilots, these talk about the Beaufort scale, about the safer and less safe sectors in a hurricane, and also about adjustments to weather observations. But they do not discuss forecasting, except in the limited sense that they mention the signs of an approaching hurricane (which I used in 1636: Seas of Fortune).
Looking at the pre-RoF holdings of Mannington libraries, the high school has Blanchard, From Raindrops to Volcanoes: Adventures with Sea Surface Meteorology; Dickinson, Exploring the Sky Day by Day; Neiburger, Understanding Our Atmospheric Environment; Chandler, The Air Around Us; Taylor, Weather and Climate; Reiter, Jet Streams; Berger, Can It Rain Cats and Dogs; Lane, The Elements Rage; and Sutton, Nature on the Rampage. The public library has Allaby, A Chronology of Weather; Burroughs, Weather; and a few juvenile books on extreme weather.
Private library holdings are likely to be slim pickings; but it would not be surprising for a farmer to own a lay book on weather forecasting. My own library includes Laird, Weathercasting; Lee, Weather Wisdom; and Watts, Weather Handbook, and there are others of a similar vein.
The bottom line is that we are going to fall well short of what was considered adequate education and training for meteorological personnel (Draghici).
The odds are pretty good that the high school has at least basic weather station equipment: barometer, thermometer, sling psychrometer, rain gauge, weather vane and anemometer, and it might have continuously recording apparatus (mine did). An individual farmer might, too. However, at present there are no personal weather stations in the Mannington area registered with Weather Underground; the closest are in Curtisville and Worthington.
Meteorological Instruments in Canon
In Fall, 1635, Peter Boglonovich’s weather station at “the Dacha” is equipped with a thermometer and a barometer (1636: The Kremlin Games, Chap. 65).
In December, 1635, the weather station at the USE’s Tetschen airfield has just a mercury thermometer and a crude barometer (Flint, 1636: The Saxon Uprising, Chap. 15).
The Meteorological Observer Network
While a single observer may make short term forecasts for his own location based on his or her own observations, multiday and regional forecasting requires a network of observers who can communicate their observations to analysts, who in turn communicate their “nowcasts” and forecasts to the public (or some particular segment thereof).
The network will be much more effective if there is standardization of the instruments, of observing methods (including observing time), and of reporting (terminology and encoding) of observations. This is to assure both a minimum level of accuracy, and consistency between reports from different locales or different observers. The speed and accuracy of communications are also important but will be discussed in a later section.
It is absolutely necessary that all observation reports include the date and time of observation. Synchronization of observations is tricky. In the 1850s Smithsonian network, observers were told to use “mean time” of their station. These could differ by several hours at the extreme western and eastern ends of the network. Standard time was initially created by railroads for the sake of timetables. In the USA, national observations weren’t simultaneous until the 1870s.
We can get some sense of how proposals for a meteorological network will be received in NTL by looking at how weather observations were made later in the OTL seventeenth century.
The first international meteorological network was created by the Medicis in 1654, and reached its peak level of participation in the period 1655-60. Observers in Florence, Vallambrosa, Pisa, Cutigliano, Bologna, Parma, Milan, Innsbruck, Warsaw, Osnabruck, and Paris participated at one time or another. The main stations were at Florence and Vallambrosa. The network was officially shut down in 1667 for political reasons—too close an association with Galileo—but the main stations continued operation until 1670.
The observers were monks (Benedictines or Jesuits) and followed a precise observational schedule (5-8 readings a day, covering both daytime and nighttime, but unevenly). The thermometers were “Little Florentine Thermometers” provided in duplicate by the network secretary, with one to be hung on a north-facing wall and the other on a south-facing wall. Observations were sent, daily or weekly depending on the distance, to the Grand Duke of Tuscany. (Given the communication delay, the network is perhaps best considered one for collecting climatological data.) The reports included the date, reading time, north and south thermometer readings, and “notes on the weather.”
In 1659-61, William Balle showed the Royal Society his weather diary, which used a tabular format to show the place, date and time of observation, daily temperature (“water glass”), pressure (“height of quicksilver”), and “weather” (e.g, “clear,” “gloomy”) (Vogel). In 1663, Robert Hooke read a paper (published in 1667) to the Royal Society, describing a proposal for a more elaborate tabular scheme for recording weather observations (Crewe). This had columns for recording the day and hour of the observation, the age and sign of the moon at noon, the direction and strength of the wind, the temperature, the humidity, the pressure, and “the faces or visible appearances of the sky, the notable effects, and general deductions to be made. . . .” (Egerton). Not only temperature and pressure, but also humidity and wind strength, were quantified. Hooke also proposed that weather observations be “made to a common standard” and collected internationally (Crewe).
While there were individuals who kept weather diaries in sixteenth-century Europe, they did not have tools to measure temperature, pressure, or wind strength. However, they could quantitatively report wind direction, as was done by, for example, Tycho Brahe and David Fabricius. One diarist who was still alive at the time of RoF was Karel starsi ze Zerotina (1564-1636) in Moravia (Pfister). In Britain, many of the weather diarists were physicians, clergymen, or landed gentry.
Some of these diarists made daily entries; others were more haphazard in their observing practices. For the meteorological network, we will need individuals willing to go out and take readings, even in harsh weather, on a daily basis.
For later networks, it is particularly interesting to note who organized (and paid for) the network and who the observers were, as this may provide some inspiration for our writers. The Meteorological Society of the Palatinate was organized by a politician-amateur meteorologist. It collected weather data thrice daily from 39 volunteers in eighteen countries, 1781-1792, but just for later study (Monmonier 18).
A network was established in New York State in 1825 by the Vice Chancellor of SUNY; it relied on observers in the 62 academies under state supervision, and these were given a standardized thermometer and rain gauge, and instructions for their use (39). The Board of Regents told them that they must submit meteorological observations before they could receive public funds (Fleming 19).
In 1837 the Franklin Institute successfully lobbied the Pennsylvania legislature for financial support for statewide meteorological observation. There was to be a set of instruments (barometer, two thermometers, and rain gauge) for each of the 52 counties, and a Philadelphia manufacturer agreed to supply them at $16/set. Standardization (the permissible deviation was 2%) proved difficult, as did shipping the instruments without suffering breakage. The Institute preferred to recruit, as an observer, “the principal of a College, Academy, or Lyceum.” (60).
In 1834, the US Navy ordered yard and ship surgeons—not captains—to keep thermometer and barometer records. The motivation was to investigate the effect of the weather on health (Fleming 61). In a similar vein, US army stations were ordered to keep weather journals in 1842 (70).
In the British network organized (1861) by the Meteorological Department of the Board of Trade, the observers were telegraph clerks, who weren’t paid for observing, but did receive instructions and standard instruments. Observations were made 8 AM daily, except Sunday (Anderson 112). Telegraphy was the department’s biggest budget item (114).
The Smithsonian network (1847-70) had several hundred volunteer observers who took measurements thrice daily and mailed monthly reports to DC. They fell into three categories: those without instruments, those with just a thermometer, and those who also had at least a barometer (sometimes also a psychrometer and rain gauge). There was also a separate, sparser telegraphic network. A word to the wise: the mail network generated far more data than the clerical staff could process (40ff).
In 1851, 47% of the observers had scientific, technical, or educational backgrounds, while only 8% were farm workers. By 1870, the values were 16% and 37% respectively (Fleming 92). While the Smithsonian observers weren’t paid, they enjoyed several perks: they could obtain free scientific advice and were given greater access to government and Smithsonian publications (88).
The instruments were provided to telegraph offices and exploratory expeditions; others could sometimes get them on loan or at reduced cost. Some instruments were broken during transit, and there is an instance of a thermometer stolen by Indians (86). Not surprisingly, there were observers who didn’t follow observational instructions or filled their reports with “hieroglyphics” (83).
In 1870, the telegraphic weather service became a function of the Signal Office of the U.S. War Department (49). The original meteorological network model was centralized; all observations were reported to the central office, and it issued any forecasts. In the U.S., the Signal Corps authorized the NY district office to issue local forecasts in 1881. Thereafter, more district offices, and even local stations, prepared local forecasts, and some even produced daily weather maps for posting at public places (53).
The military wasn’t very responsive to the meteorological needs of the farming community and in 1891, the function was transferred to the newly created USDA Weather Bureau (Id.).
Ultimately, of course, it will be desirable to have some automated weather stations for remote or harsh locations. Christopher Wren (1632-1723) and Robert Hooke (1635-1703) together designed, and Hooke built, the first automated weather station, the “weather wise.” The first working model is from 1669, but they continued to tinker with it. This contraption “made recordings at the Royal Society with trip hammers that made marks on paper to record wind direction, wind speed, temperature, humidity and pressure. Not surprisingly, it spent more time being repaired or developed than actually working. . . .” (Crewe). I have not been able to document other pre-twentieth century automated weather stations.
Station Bulletins and Plots. Currently, weather observations (bulletins) are reported in one of several standard international formats so they are concise and can be understood without linguistic skills (Stegman). WMO’s SYNOP bulletin format for a land station is twelve four- or five-character code groups, and for a ship 15-18 (plotmanual). There are also (to name a few) METAR (for airports), TEMP (for weather balloons), and most recently BUFR.
At the analysis center this information is first plotted on a very large weather map (or a set of smaller maps) so that all of the station information is visible. Several standardized “station plot” formats have been developed in which the numbers and accepted graphical symbols are arranged in a predetermined configuration (plotmanual; NWSSSP).
It’s not very likely that these formats are fully explained in Grantville literature, but some kind of system needs to be developed in the NTL and the ex-pilots probably know more about the historical formats than anyone else in town.
In part 1 (Grantville Gazette 72) I discussed measurement of temperature, humidity, and precipitation, and in part 2 (Grantville Gazette 73), pressure and surface wind speed and direction. But there are other meteorological variables of interest.
The observer may and should also report on the appearance of the sky—the percentage covered by clouds, the types of clouds seen, and their apparent rate and direction of movement. The vocabulary of clouds was developed by Luke Howard (1772-1864), and there are definitely books in the Grantville public and school libraries that illustrate the cloud genera (e.g., “altocumulus”). They are less likely to drill down to the level of species (e.g., stratiformis vs. floccus) or variety (opacus, translucidus, etc.) (Pretor-Pinney) but it is doubtful that these minutiae will be useful from a forecasting standpoint.
Observation stations may be on land or on the water, and with the right equipment they can make upper air as well as near-surface observations. They also be manned or unmanned (automated).
Automated stations require recording instruments (meteorographs). All must connect the sensor so that a change in the meteorological variable causes a movement of the recording element. Often some sort of amplification is necessary. The recording element must in turn produce a record: e.g., by causing a pen to move across or intermittently press against paper, scraping a smoked or silvered surface with a stylus, or causing a spot of light (natural or artificial) to traverse photosensitive paper. Finally, the recording medium is mounted on a drum that is rotated at a uniform rate by clockwork (MiddletonMM 6ff). Generally speaking, meteorographs have to be periodically recalibrated against the corresponding meteorometers.
Automated stations can be expected to receive maintenance visits only occasionally, and are likely to be used primarily in harsh environments where it is difficult for a human to live, and thus must be engineered with robustness in mind.
Land Surface Station Observations
On land, both natural features and manmade structures can affect the readings obtained by meteorological instruments. Thermometers, wind vanes, anemometers, and precipitation gauges must be appropriately and consistently sited, as discussed in the prior two articles. Another point to be wary of is that wind direction is reported relative to true north by meteorological stations and relative to magnetic north at airfields (WMO2008, 5.1.2).
Air pressure varies with altitude, and local gravity (and thus the weight of a mercury column) with both altitude and latitude, so it is very important that all barometric pressure measurements be reduced to mean sea level pressure before they are reported to the central office. It should be noted that the altitude needs to be determined by non-barometric means, e.g., surveying with level, chain, and theodolite relative to some reference datum. (There are complications in long-distance surveying that are outside the purview of this article. )
An interesting question is, how do you get weather reports from inside enemy territory?
First of all, you can have a mobile weather station that travels with your forces. It can be manned or automated. Second, you could drop (by parachute) an automated weather station into enemy territory (and it could double as a radio beacon for air missions) (MiddletonMM 271).
Marine Surface Station Observations
Incorporation of marine observations from the North Atlantic into the network will be very advantageous for weather forecasting for Europe, because in the northern hemisphere, frontal systems tend to move from west to east.
Marine observations worldwide will be used for determining prevailing winds (strictly speaking, that’s climatology, not meteorology) as a function of season and location, which in turn will make it possible to find more efficient sailing routes.
In this period, ship captains usually kept logbooks in which weather observations were recorded. Indeed, some navies and merchants marine required that such logbooks be maintained and turned in at the end of the voyage.
Taking this a step further, Maury organized an international conference in 1853 at which the participating nations agreed that “cooperating merchantmen” would keep a standard log in which they would record the time, position, pressure, temperature, and wind direction and strength (Moore 206).
While merchant vessels and even warships can report weather observations as they progress across the ocean, it was found advantageous to station ships in one place to monitor the weather. The first such “weather ship” was deployed in 1938, and the last was withdrawn in 2010. Wikipedia has a map showing thirteen weather ship locations in the North Atlantic. Typically, they spent two-thirds of their time at sea.
I think it would be hard to justify a dedicated weather ship in the 1632 universe. However, if the ship would also be useful as a radio relay station, that might shift the economics in its favor, depending on the value of the radio communication link it supports. Or perhaps it could also double as a lightship, if stationed in shoal waters.
Historically, weather ships were replaced by weather buoys. These were first deployed by German U-boats during WW II. They may be moored or drifting. Naturally, they must be automatic in operation.
Unlike the land observer, the nautical observer will likely be asked to describe the sea conditions. It is helpful to determine the time interval between crests. To do so, you go up to the crow’s nest, pick a distinctive mark (foam patch, clump of seaweed, drifting object) one or two ship-lengths to windward, and count the number of wave crests that pass under the mark within a set time from the first wave crest (AM3/2 p185). Obviously, this assumes that the mark is still or moving only slowly.
Determining wave height is tricky. If you are accompanied by another ship, you can observe the height of a wave striking her, relative to distinctive parts of the ship’s structure. If your ship is alone, then you go amidships and judge the height of the trough and crest against your ship’s reference points.
It is desirable to take the sea temperature. Franklin mapped the Gulf Stream in 1770. Knowing whether you were in it or not could make a difference of two weeks at sea for a voyage up or down the East Coast (Moore 77). Later it was recognized that average sea temperatures in the tropical Atlantic were relevant to predicting the ferocity of the hurricane season.
One complication is that the apparent wind (what is felt by the sails, and also by the weathervane and anemometer) will differ from the true wind if the ship is moving. If you know the true course and ship speed then you can calculate the true wind from the apparent wind. However, that will require some navigational advances.
The nautical observer may try to estimate true wind direction “by noting the direction from which small wavelets, ripples, and sea spray are coming” and the true wind speed by observing the sea conditions and correlating them with the modern version of the Beaufort wind scale. (NAVEDTRA10363, p48).
Upper Air Observations
Clouds occur at different heights, and by studying how clouds move, you can determine whether the air at high altitude is moving in a different direction than the surface wind. Espy, in his Hints to Observers on Meteorology (1837), pointed out how cloud movements could be used to estimate upper air winds (Fleming 61). “In the late 1890s [Bigelow] used cloud data from 140 telegraphic stations to construct three-level wind maps describing for various weather types (e.g., “New England Winter High”) the movement of air at the surface as well as among both ‘lower’ and ‘upper’ clouds.” He was systematic, dividing the country into 96 squares (Monmonier 69).
Forecasters also like to know the upper air pressure distribution. The upper air patterns are simpler (less terrain perturbation) and thus easier to use in prognosis.
In theory, mountaintop observations can also be used to supplement data from nearby lowland stations. However, the mountain disturbs the atmosphere and thus the mountaintop data doesn’t give a true picture of the upper atmosphere overlying the lowland.
Direct aerial observations of upper air conditions may be taken by unmanned kite, balloon, or rocket, or by manned kite (uncommon!), balloon, aircraft, or airship. (Balloons include “kytoons”, essentially blimp-shaped balloons with rigid tail fins. These lift like a kite in a strong wind—MiddletonMM 229).
Unmanned Probes. With an unmanned conveyance, there is no one present to fix a glitch in the instruments or to take readings at intervals. Unmanned upper air observations became more useful once continuous multi-variable recording instruments (meteorographs) were developed.
Until the meteorograph could communicate from the air, the recovery of the data was delayed until the vehicle returned to earth and was located. Middleton comments in “thickly populated central Germany a loss was exceptional,” but that even in Ontario, nearly three-quarters of meteorographs were “eventually” recovered (MiddletonMM 230). (Elsewhere Middleton notes that the recovery could take days, weeks, or even years—243.) However, even when the meteorograph was padded and its descent retarded by parachute, it was often damaged, perhaps beyond repair. Hence, their designers had to balance the competing goals of lightness, accuracy, robustness, and inexpensiveness.
Several expedients resulted in more timely transmission of the sounding data. I have already mentioned in the earlier articles how some meteorological instruments could report their readings at a remote location. When this was accomplished by electrical means, the sensor could be in the aerial vehicle, and the readout on the ground, connected by a very long electrical wire. This was called a wiresonde.
The next important development was the radiosonde (1927), the combination of a meteorograph with a radio transmitter (Brettle; MiddletonMM 243). A battery is also required, to power the transmitter. The transmitters communicated the data by means of (1) time interval between pulses, (2) Morse code, (3) variable radio frequency, and (4) variable audio (modulating radio) frequency (244). I would think that amplitude modulation was also a possibility. Middleton recommends that the transmitter have a range of at least 100 miles.
Manned Probes. With manned vehicles, the problems are that the height of ascent is limited by oxygen supply, a greater lift force is needed to support the weight of the observer, and the return to earth must be gentle enough not to injure the latter.
Sounding Altitude. At what heights do we need weather observations? Heights are often expressed in terms of the pressures for the international standard atmosphere; the earth’s surface is a little over 1000 hPa. The ENIAC numerical weather model relied on measurements at 500 hPa, where frictional forces are negligible and half the air mass is below you; that’s 18,000 feet [the mean height of the 500 hPa pressure according to the International Standard Atmosphere, but the true height can vary]. “The [pressure] level best suited for determination of convergence and divergence is the 300-hPA level.” (ISA 30,000 feet) (NAVEDTRA 14010, 7). The most common of the lower atmosphere plots are for 850 mb (ISA 5,000 ft) and 700 mb (ISA 10,000 ft). These constant pressure (isobaric) maps are contour maps in which the contours are isoheights (lines of equal height).
It is also possible to generate constant height maps in which the contours are isobars. Pilots of course will be most interested in the constant height chart for whatever is their expected cruising altitude; aircraft often try to follow isobars. In 1903, Bigelow produced pressure maps for 3500 and 10,000 feet (Monmonier 71).
Altitude Measurement. A general problem for upper air observations is uncoupling pressure and height. The most common form of altimeter really measures barometric pressure. If we want to generate maps showing heights at a standard pressure, or pressures at a standard height, we need a non-barometric method of determining altitude.
If the observation platform is tethered, then in theory this can be done by trigonometry if we know the payed out length of line (thanks to markings on the line or counting turns on the windlass) and the angle of inclination of the line.
An optical rangefinder might be a possibility. In stadiametric rangefinding you sight on an object of known size and measure its angular width. In coincidence rangefinding the optics combine partial images from two sighting points at opposite ends of a base line and you determine the distance by bringing the images into coincidence. Stereoscopic rangefinders are similar, but there are two eyepieces instead of one. In any case, the aerial observer must have a well-defined object to sight on, preferably close to directly below the vehicle so you are measuring the vertical distance. Or the ground observer must sight on the aerial platform, but then you need to have a way to synchronize the altitude and pressure measurements.
Another approach (used historically on German airships, see Dick) is to drop an object and time its fall. This worked only over water, where the splash could be observed. Note that dropping anything is like dropping ballast and on a balloon or small airship may result in a significant height change, so take the readings before you make the drop. You also need an accurate timer.
All optical methods are of course dependent on atmospheric clarity and adequate lighting. Searchlights were used with the “splash” method at night.
The sonar altimeter measures the time it takes for a sound emitted by the observing platform to bounce off the earth and return. A primitive sonar altimeter, the Echolot, was used on the ZR3 in 1925 (Draper 6) and on the Graf Zeppelin. The sound source was “a gun, loaded with a blank cartridge, . . . fired downward through a sleeve in the control car.” On the Hindenburg, the sound source was a compressed air siren (airships.net). A whistle and megaphone have also been used.
The first sensor was a carbon powder microphone; the listener watched a metronomic timer and stopped it when he heard the sound (Draper 7). On the Graf Zeppelin, the sensor was “a bouncing type of light indicator” that would vibrate in response to the echo (Dick). The bouncing light was produced by a rotating mirror, and when a sound was received by a carbon microphone, the reed of an electromagnetic oscillograph was excited and moved a lens, deflecting the light (Draper 56). In a later version, the echo striking the microphone caused an electromagnet brake to stop the rotation (57).
It is possible to imagine a completely automated device that can be placed on an unmanned platform, and perhaps even records the transit time alongside the barometric pressure.
The sound must be sharp and shortly defined so that at low altitudes the echo is not confused with the generating sound. The echo must be louder than the vehicle noise; this is more of a problem for aircraft than airships or balloons. Absorption and spreading losses limit the sensible upper altitude (but the louder the sound source, the better). A typical upper limit is 800 feet for cruising airplanes, 1600 for gliding airplanes, and 2400 for airships (Draper 84). The accuracy of the timer limits the accuracy of the altitude, given the speed of sound, an altitude difference of one foot corresponds to about 1/600th second. (I don’t know whether the accuracy of the early sonic altimeters was sufficient to warrant considering the effect of the variable density of air on the speed of sound.)
The same echo timing principle appears in the modern radar altimeter, but that senses the echo of a radio wave, traveling at the speed of light, and thus requires much more sophisticated engineering. I will leave it to others to predict when this will appear in NTL. (Laser and GPS altimeters are even more remote prospects.)
Before leaving the subject of altitude measurement, I would comment that as long as the vehicle remains below cloud level, it is possible that its height could be measured from ground stations, by the combination of optical measurements and trigonometry. Two ground observers could be stationed at the ends of a much longer baseline than that possible on the vehicle, and thus the rangefinding would be more accurate.
Let’s examine more closely the various direct observation modes.
Unmanned Kites. In 1749, Alexander Wilson sent thermometers aloft on paper kites. “Each thermometer was bundled in strips of cloth to prevent the delicate instruments from breaking when they hit the ground. Each bundle had a slow burning fuse and a white ribbon attached. As the fuse burned through the string that bound the bundle to the kite, the white ribbons would surf the wind on a downward descent and signal to the experimenters.” (Robinson). In 1847, “a six-sided meteorological kite in 1847 [was developed] that allowed meteorologist to raise and lower weather instruments using a pulley system” (Id.).
For a time, kites were eclipsed by balloons, but it was found that tethered balloons were harder to control than kites, and free balloons might be carried so far away that their instruments weren’t recoverable. Moreover, kites were cheaper and could be flown in stronger winds (but not in a calm!).
Kite string was replaced with piano wire in 1887, and kite designs were modified for increased stability. In 1894, a kite was used to lift a thermograph. By 1899, the Weather Bureau was able to send instruments up above 10,000 feet, with box kites attached to a main line at 2,000 foot intervals. The kite would be held at a specific altitude for 5-10 minutes to let the instruments stabilize. The line, up to 40,000 feet long, could be wound up by a steam-driven windlass, and a dynamometer was attached to the reel to measure the pull. In the early twentieth century, the Bureau had high-wind, moderate-wind (12-30 mph) and light-wind (8-10 mph) kites. There are illustrations of the Marvin-Hargrave kite (68-square foot lifting surface) and the two pound kite meteorograph in EB11/Meteorology.
In 1898, there were eighteen kite stations in America.The basic limitation of the kite was that its data wasn’t available until the kite was brought back to earth. Hence, it was finally replaced by balloons carrying radiosondes (see below) (Monmonier 74). In theory, of course, a kite could carry a sufficiently light radiosonde, but I haven’t found any historical instance of this. However, a wiresonde was used with a kite in 1917 (MiddletonMM 243).
Manned kites. Cody’s war kites were intended for military reconnaissance, and a Cody kite lifted a man to a height of 2600 feet, but I don’t know whether any kite was used to carry a meteorological observer.
Manned Balloons. In 1862, Glaisher took numerous meteorological instruments up in a 90,000-cubic foot balloon. (Moore 253). On his first ascent, up to almost 17,000 feet, he didn’t report any health issues, but on his third, he reached 29,000 feet and blacked out for a time. At 35,000 feet his companion managed to pull the valve cord. They descended rapidly (which is also dangerous) and made it back to ground (Moore 261ff). As information about ballooning diffuses out of Grantville, one can imagine that some down-timers will learn how to make a balloon without also becoming aware of the dangers of high altitude.
It is worth comparing these heights to those of the jet stream; the polar jets are at 30-39,000 feet, and the subtropical jets at 33-52,000 (Wikipedia). However, at 20,000 feet, Glaisher’s instruments began to malfunction.
Some manned balloon observers used a telegraph connection between the balloon gondola and the ground for communicating observations in near-real time.
Unmanned Balloons. There are basically two types of balloons used by meteorologists, pilot balloons (uninstrumented, but tracked) and sounding balloons (instrumented, and called radiosondes if they have radio transmitters).
Pilot Balloons. The use of pilot balloons was pioneered in 1909. The balloon movement was observed with a theodolite. Its disappearance into the cloud cover could be timed, giving an estimate of the cloud height, and the measurement of the balloon’s azimuth and angular altitude gave the wind direction and speed (Monmonier 74). There were about 75 pilot balloon stations covering the USA in 1933 (76).
The theodolite for land use is essentially a wide-angle telescope that can pivot on two axes and can measure horizontal and vertical angles. A special theodolite for shipboard use had a Cardan suspension to keep the instrument level, and was really a combination of a sextant and a conventional theodolite (Middleton 179).
If only one theodolite is used, the calculation of the direction and speed of movement of the pilot balloon is dependent on the accurate estimate of the rate of ascent (181), and this is the greatest source of uncertainty.
At first the balloon accelerates freely, but as its speed increases, air resistance becomes significant. The speed becomes constant when the drag force equals the buoyant force.
However, both forces also decrease slowly with the density of the atmosphere and thus with height, and not quite in lock step (since one is related to area and the other to volume). Hence, the rate of ascent is not quite constant. With some simplifying assumptions (note 4), the velocity is inversely proportional to the one-sixth power of the air density, which itself decreases exponentially with altitude in the troposphere (0-10 km) and more gradually in the stratosphere (20-50 km) (MiddletonMM 172; DennyWB; Yajima 2.100). Hence, the theoretical “equilibrium” ascent rate slowly increases. For a typical sounding balloon with a one-meter radius at ground level (Gallice), the expected rate for a hydrogen balloon is 5 meters/second, and is reached at an altitude of about 5-7 km. The rate increases to 5.5 m/s at about 20 km. (DennyWB).
However, there can be countervailing factors (see note 4), and in the tropopause, 11-20 km, the location of the jet stream, the ascent rate decreases (Gallice). For many years, the ascent rate was considered practically constant (MiddletonMB 172).
Two theodolites were used mostly in mountain regions (where vertical currents could be substantial), and they make that estimate unnecessary. The baseline must be adequate, say, one mile (186). Even so, the values are accurate to perhaps 2 degrees direction and 2.2 mph in speed at low altitude, and twice that at 5 km altitude.
Pilot balloons are preferably made of either natural rubber or neoprene. Natural rubber deteriorates as a result of exposure to sunlight, and neoprene doesn’t fare well in cold, so the former is preferred at night and the latter in daytime. By trial and error, it was found that the best colors were clear against a blue sky, red with broken clouds, and blue or black when overcast. For night flights, the balloon may be tracked by hanging a “Chinese lantern” (essentially a candle with a white tissue paper shade) 10-20 feet below the balloon (174). Larger balloons are easier to track visually, but more expensive.
A ceiling balloon is a small pilot balloon that is observed merely to determine when it disappears into the clouds and thus the height of the clouds. They are cheaper and therefore can be released at an airfield more frequently than a normal pilot balloon.
To obtain wind data from above the cloud base, pilot balloons could be equipped with a radio. Instead of using one or two theodolites, you would use one or two radio detection finders with highly directional antennas and a similar angle measuring capability.
Sounding Balloons. A “pilot balloon” provides information solely by being tracked from the ground. The next step beyond that is a “sounding balloon,” which carries at least one meteorograph. The Scott expedition launched (1911) meteorological balloons over Antarctica; these were one-cubic meter gutta percha balloons filled with hydrogen produced by adding calcium hydride to water. The recording (temperature, pressure) instrument weighed 70 grams, and was attached to the balloon by a device containing a slow fuse (15 minutes). The expiration of the fuse detached the device which dropped to the ground (without a parachute to slow it down) (Burton).
The development of compact, light, cheap transmitters and meteorographs led naturally to the evolution of sounding balloons equipped with radio transmitters (radiosondes); the radio then serves the dual purpose of permitting the ground station to determine the position and movement of the balloon, and of transmitting the meteorological data to the ground. It is advantageous to hang the radiosonde below the balloon—as much as 200 feet below—so heat absorbed by the balloon is not transferred to the radiosonde (Denny 100).
While unmanned, untethered balloons equipped with radiosondes are the backbone of the modern upper air sampling network, they are one-use devices (the balloons burst when ascending high enough) and radiosondes are expensive ($50 in 1960) (Edwards 217). In the modern USA, the recovery rate is about 20% (Denny 100).
Middleton suggested that one could design a “controlled altitude free balloon” that, like the Japanese “balloon bombs” sent against the Pacific Northwest in WW2, would have a pressure switch to drop ballast if the balloon descended too low and to valve gas if it ascended too high (196).
Balloons generally. While a sphere is the shape best able to withstand pressure, and has the smallest surface area for a given volume (thus minimizing initial weight/lift), it requires additional structure to suspend a heavy load, and when launched the balloon is only partially inflated and does not have its final spherical shape (increasing stress and drag). Therefore, in scientific ballooning, only small balloons with light payloads are spherical (Yajima 20). In contrast, a “natural shape” balloon maintains “the same type of shape from partial inflation through to full inflation” (21). Calculating the shape initially required a digital computer (22), but Grantville’s balloonist, Marlon Pridmore, probably has tables that establish the shape.
Zero pressure balloons (45ff) have a venting duct at the base and, after the balloon obtains full inflation, if there is further expansion some lifting gas will “overflow” (underflow) and escape by the vent, thus maintaining a “zero” (in practice perhaps 1%) pressure differential. This minimizes the stress on the envelope and thus permits use of very thin envelopes. However, venting reduces the lifting force and depending on temperature conditions may cause the balloon to descend. If it is desired to have a balloon remain aloft after sunset, one must consider that the temperature will drop, reducing buoyancy, and automatically drop ballast to compensate.
Super-pressure balloons (48ff) are closed (except for safety valves to forestall bursting) and once fully inflated, further ascent causes an increase in the pressure differential across the skin. While these balloons are capable of long duration flights, the problem was to develop a film strong enough to withstand a much greater pressure differential (perhaps 20% in the stratosphere) without greatly increasing the film weight (and thus the maximum altitude). One may also have dual-balloon systems (51ff) combining a small super-pressure balloon to control altitude (despite the day-night cycle) and a large zero-pressure balloon to lift the payload. These relax the weight constraint on the super-pressure balloon and would be feasible in NTL much sooner than a pure super-pressure system.
Balloons are usually characterized by weight instead of launch or burst diameter. The envelope of a ceiling balloon might weigh 10-30 grams; of a pilot balloon, 100 grams; and of a sounding balloon, 300-3000 grams. Small balloons may be made by dipping a mold into the latex emulsion, but larger balloons require putting the latex in a divided hollow mold and rotating it to achieve a uniform thickness (MiddletonMB 167).
The maximum altitude is set by buoyancy considerations and thus in part by the combined weight of the balloon and payload; the buoyant force decreases with altitude but the weight is constant. A balloon may fail to reach that altitude because it bursts prematurely. A partially-inflated balloon will expand as it rises. Folds in the envelope will disappear, and the envelope material itself will stretch and thin. If the balloon climbs faster than it can compensate for by volume change, it must either vent lift gas (“zero-pressure” balloon), or experience an increasing pressure difference between interior and exterior (“superpressure” balloon).
The stress on the envelope increases with the radius as the balloon expands, and the envelope’s ability to resist the stress is reduced as it thins. Modern natural rubber balloons of 300- to 1200-gram sizes burst at a film thickness of 3.3-3.5 microns. If the rubber density doesn’t change as the balloon stretches, the burst volume is directly proportional to the envelope weight1.5 and inversely to the minimum thickness1.5 (Yajima 155).
Larger balloons generally make faster ascents and have higher burst altitudes. A 300-gram balloon might have a diameter of 4.1 feet and a volume of 36 cubic feet at launch, and a burst diameter of 13 feet, bursting at 82,000 feet. For a 1200-gram balloon, figure 6 feet at launch, 28 feet at burst, and burst altitude of 109,000 feet (Kaymont Balloons). These burst altitudes are achieved with balloons made from pure latex (dust excluded!) and by very precise molding methods; in the 1630s we aren’t likely to do so well.
A typical scenario for a stratospheric balloon is 90 minutes from launch to burst, and then another 30 minutes for the parachute-borne payload to return to earth (McNamara).
Rockets. There is a model rocket club in Grantville. Rockets can be equipped with black powder “motors” and used to carry a radiosonde. A rocket can be prepared for launch faster than a large balloon can be inflated, and it will reach its ceiling much faster. It can also be used where you don’t want to handle hydrogen. It also can reach higher altitudes than a balloon could. But sounding rockets are much more expensive than a balloon of the same altitude capability (McNamara). Usually, rockets carry “dropsondes”—the weather is observed on the way down, via parachute, rather than on the way up. Weird tech note: in the 1950s the U.S. Navy developed balloon-launched rockets.
Airship and Aircraft. We can assume that the crew of the NTL aircraft and airships will be expected to make weather observations and radio the information to friendly ground stations.
Like ships, these vehicles will be observing the apparent and not the true wind. But airships, at least, can hover while they make weather observations.
Particularly with aircraft, we have to correct for aircraft speed. There is excess pressure at the nose and reduced pressure at the tail. There are corresponding adiabatic temperature changes. In addition, along the side there will be frictional heating. If the aircraft flies through a cloud, some of the heat will go to evaporating the liquid water present. The increased pressure and temperature lower the apparent relative humidity (MiddletonMM 236ff).
With both aircraft and airships, we will have to be careful where instruments are positioned so ship operations (e.g., heat from engines or burners) don’t perturb readings.
Sferics. Lightning causes static that can be heard on unused radio bands. Hence, as Popov (1894) demonstrated, a radio receiver can be used as a lightning detector.
Determining the distance and direction of the lightning strike is trickier. You can use several well-separated detectors and triangulate (with the observers synchronized so they know they are hearing the same strike).
Rossby said, “Communications are the alpha and omega of meteorology” (Fleming 96). In 1854, a storm heavily damaged the French-British naval forces off Balaclava. Le Verrier reconstructed the course of the storm, and realized that a telegraphed message would have given the naval commanders a day’s warning. This led to the institution of a daily telegraphed weather bulletin in France in 1858, and a similar system in Britain in 1860 (Monmonier 44).
Physical Delivery. Messengers, traveling by foot, horse, or rail, can deliver weather reports, making use of an established post horse or railroad station network. And let us not forget carrier pigeons, which were used quite effectively during the siege of Paris in the Franco-Prussian War.
Still, there are obvious limitations on the speed of communication. NTL early 1635 Russia has several weather stations, and Peter’s central station, at the Dacha, receives observational data from the others just once a week. The same messenger also carries his reports to Moscow, again just once a week (1636: The Kremlin Games, Chap. 65).
Optical Telegraph. In 1684, Hooke proposed communication by optical means; messages would be transmitted by displaying deal boards of various shapes and orientations (or lanterns at night), and received by viewing them in a telescope (Moore 26).
An actual optical telegraph network was built in France in 1793. It was from Paris to Lille, 230 kilometers long, with fifteen semaphore stations. The fifteen-foot-tall semaphore tower had two signaling arms, each of which could be placed in any of seven positions, and the post could be turned to any of four positions. The transmission rate was 1-3 symbols per minute, and one symbol could be transmitted the full length of the line in ten minutes. The first message reached Paris in less than half an hour. The network was expanded and “at the beginning of the 19th century, it was possible to wirelessly transmit a short message from Amsterdam to Venice in one hour’s time” (De Decker).
The advantages of the optical telegraph over the electrical telegraph is that there are no wires; hence we avoid the cost of the wires and their maintenance, not to mention the risk of communications being interrupted because hostile soldiers cut the wires or local farmers steal them to resell the copper. The disadvantage is that the optical telegraph cannot transmit in visibility-obscuring bad weather or (if it lacks light signals) at night.
It is conceivable that pyrotechnics or signal lamps would be used. See Cooper, “Life at Sea in the Old and New Time Lines, Part 4: Lights Across the Waters” (Grantville Gazette 71). I am not sure how visible they would be in bad weather, but of course the very powerful lights of a lighthouse are intended to be seen from several miles away, even under adverse conditions.
Electrical Telegraph. In 1843-44, Morse constructed a prototype 36- mile line from Washington to Baltimore (Moore 163). In 1848, in England, there was a brief experiment with collection of weather reports (wind, weather type) by electrical telegraph. At 9 AM, readings were taken at twenty-nine telegraph stations and telegraphed to Glaisher in London, who published them the following day in the Daily News (Moore 191).
Later that year, the Daily News decided that collection of weather data by telegraph was too expensive and in 1849 turned instead to railroad-borne mail (while we think of telegraph lines paralleling railroads, in England at the time, the railroad network was much more extensive). This time the station masters made the reports, and the railroads carried them for free (Moore 193).
By 1854, the telegraph coverage of England was much expanded, with less than 10% of the population living more than ten miles from a telegraph station (210). Readings were taken at 9 AM and by 10 AM telegrams were arriving at the meteorological office in London. By 11 AM, the morning reports were sent by messenger to The Times, the Shipping Gazette, Lloyd’s, etc (274).
In America, too, cooperation was sought from telegraph companies, and in 1858 the Smithsonian was receiving limited descriptions (weather type and wind direction) from thirty-two stations (Monmonier 41).
Note that high winds can bring down telegraph lines, and that geomagnetic storms can cause interference (as they did in 1859).
For the practicality of building NTL telegraphs, see Boatright, “So You Want to Do Telecommunications in 1633?” (Grantville Gazette 2).
Radio. Peter Boglonovich mutters to himself in fall 1635, “What’s the use of a weather station if it doesn’t have a radio?” (1636: The Kremlin Games, Chap. 65).
Radio weather reports are being received in Grantville as of March, 1635. In Offord, “The Vice President’s Plane is Down” (Grantville Gazette 26), the chief of police is told that the barometer is falling in Fulda and a “westerly front” is approaching Frankfurt am Main.
Insofar as the capabilities of NTL radios are concerned, see Boatright, “Radio in the 1632 Universe” (Grantville Gazette 1) and “Radio in 1632, Part 3” (Grantville Gazette 9); Carroll, “Marine Radio in the 1632 Universe” (Grantville Gazette 52).
What we won’t have. There are some observational tools that twentieth-century weather forecaster had that we simply will not have in the 1630s. These include ground radars (for wind profiling and rainfall detection) and satellites.
Weather reports. Besides communicating the weather observations to the analysts, the analysts must also communicate the forecasts to the consumers. If the information is confidential (e.g., for military or private economic purposes), it is likely to be encrypted, or communicated by more secure means (telegraph lines, heliographs or semaphore lines, or messengers), or both.
In 1848, the London Daily News began reporting what the past day’s weather had been in many English cities, based on telegraph- (later railroad-) borne reports (Monmonier 154).
In the U.S., naval stations began broadcasting abbreviated weather bulletins in 1914. However, the limitations of radiotelegraphy (broadcasts in Morse code) also limited the demand for the product. As radiotelegraphy was replaced by radiotelephony (voice broadcasts), the picture changed; in 1922, 98 stations in 35 states were voice broadcasting Weather Bureau reports and forecasts (Fleming 51).
Weather Maps. Weather maps are used both by meteorologists and consumers, but their interests differ. Meteorologists want to know pressure distributions, whereas consumers care more about precipitation and temperature. Both care about wind strength, but only meteorologists and sailors are concerned with wind direction. Meteorologists have to look initially at all reporting stations, whereas consumers just care about the weather where they live.
Edmund Halley plotted the trade winds in 1686 (Moore 141), but that’s climatology, not meteorology. Daily weather maps were proposed by Brandes in 1816, and he published a “geographic table” in 1826 (Monmonier 18ff). The first true weather map was that of Loomis (1843), and it was color coded to show the weather type (clear, cloudy, rainy, snowy, foggy) in different regions. There were arrows for wind direction, and dotted lines to show points of equal pressure (isobars) and temperature (isotherms) (Moore 145). However, it was not a contemporary report; it showed conditions from a year earlier. In 1851, Glaisher produced daily weather charts that were put on display at the Great Exhibition (Gribbins 257).
The first daily newspaper weather map (showing yesterday’s weather) was published in 1875. A “drill-pantograph” was used for “transferring lines from a hand-drawn weather map to a printing plate” (Monmonier 156). Newsworthiness was increased in 1876. when the morning papers began carrying maps based on the observations at 6 PM the evening before (157).
In 1909, 112 U.S. weather stations were printing their own weather maps, but 79 of them did so by mimeograph (164).
Determining how soon facsimile machines can be built is outside the purview of this article, but in 1877, Caselli’s pantelegraph was used to transmit a weather map across a 120-mile telegraph line (The Manufacturer and Builder, 9:88, 1977).
Choice of Observation Locations
Whether a station is placed at a particular location will depend on its economic and military importance, the sensitivity of the local economy and military operations to local weather conditions, the proximity of the location to the existing communications network (preferably radio stations, telegraph lines, and railroads, but roads and shipping lanes are also of interest) and to geographic features that facilitate communication (e.g., hilltops to emplace optical telegraphs or radio transmitters, salt water for ground wave radio), the relevance of weather experienced at that location to weather later experienced at more important locations (e.g., it is on a “storm track”) , and finally, politics (Is it in an area friendly to the government that is planning the station?).
For the USE, I figure that the backbone of the network will be weather stations associated with the railroad stations. A telegraph line parallels part, but not all, of the first railroad line. Other sections are served by radio. A railborne messeger can be the backup for either.
It takes time and money to construct railroad and wired telegraph lines, so to cover other parts of the USE, we must rely on radio or on optical telegraph. Magdeburg of course has radio, and by the end of 1635 or early 1636 I would expect radio transmitters to be erected at the major ports, such as Stockholm, Copenhagen, Bremen, Hamburg, Lubeck, Wismar, Rostock, and Stettin. The provincial capitals and the towns with major universities are also likely to be part of the radio network, which will range as far east as Prague (and, further north, Riga).
Some sense for the disparate elements of the USE’s meteorological network may be gleaned from Carroll, “Time to Spare, Go by Air,” Grantville Gazette 28: “A military plane on a high altitude flight came on the air and started relaying weather reports from far-flung postal stations, ships, hams, and the very few airfields where there was someone who knew anything about weather.” Reports come in from the Netherlands, Nordhausen, Wolfsburg. Hannover, and further north. “Then Magdeburg Tower started transmitting reports from railroad telegraphers, which the air force co-pilot dutifully repeated. . . .” These observers were at Halle, Feuchtenthal, Eisleben, Aschersleben, and of course Magdeburg. This reveals that the Netherlands are part of the USE network, and the Trans European Airways stories show that the Venetian Republic, Tyrol, and Tuscany are cooperating, too.
As to the density of the USE network, well, it depends. Along the rail lines, probably every 25-50 miles; see Carsten, “Railroading in Germany” (Grantville Gazette 7). Further west, probably every 50-100 miles, depending on antenna height and station power. Further east, scattered stations 100-500 miles apart, communicating by sky wave.
It would be nice to know common storm tracks into Europe, and the location of the semipermanent weather features that affect European weather.
Van Bebber (1891) identified five tracks for low pressure centers over Europe:
I: from the Atlantic, northeastward across the northern British Isles to northern Scandinavia (Ia), and thence southeast (Ib), east (Ic), or northeast (Id)
II: from the North Atlantic eastward over central and south Scandinavia to northeastern Europe
III: from the North Atlantic south-east over southern Scandinavia to Central Eastern Europe
IV: from the middle Atlantic across the southern British Isles and southern Scandinavia (IVa) or northern central Europe (IVb) to northeastern Europe
V: from the middle Atlantic across France of the Bay of Biscay to northern Italy (Va) and then northeast (Vb) to Finland, east (Vc) to the Caspian, or southeast (Vd) to the Mediterranean.
Weather diarists in individual towns may have some sense of the direction that storms come from, and it may be possible to piece together this information to postulate typical storm tracks, at least over the USE.
Stations in North America are probably not going to be especially useful for forecasting European weather. While weather systems do generally move from west to east, extratropical storms exiting North America tend to veer poleward, paralleling the eastern coast of Greenland and perhaps hitting Iceland (Wing).
The 2002 CD version of Encyclopaedia Britannica is based on the 1998 paper encyclopedia. The article on wind provides maps of world distribution of mean sea level pressure, and primary and secondary storm tracks, for January and July. Insofar as Europe is concerned, the July map shows storm tracks from Newfoundland to Iceland, from Iceland to southern Scandinavia, and from west of Ireland to northern England. The January map shows storm tracks from the Eastern US to Iceland, from southeast of Iceland to northern Scandinavia, from east of Scotland to the Baltic, and from southern England to Italy.
Obviously, that doesn’t show all possible storm tracks. Occasionally fronts come down from the pole, up from the south (dropping desert sand on cars) or over from the east. But there are up-timers who had a tour of duty in Germany or Belgium, and would have some relevant recollections of pan-European weather reports and forecasts on the news.
The EB2002 CD article on Europe briefly discusses the roles of the Icelandic Low, the Azores (and Bermuda) High, the (winter) Mediterranean Low, the Siberian High, and the Asiatic Low. Unfortunately, I was not able to find any reference to the North Atlantic Oscillation (relating to the teleconnection between the Icelandic Low and the Azores-Bermuda High) or the Polar Oscillation, which would have been useful for long-range forecasting. But perhaps the pilots in Grantville would have heard of them.
The British Isles would be a possible early warning line for storms approaching Germany from the west, but the political situation is problematic. Jack Carroll has suggested a station at Cape Clear (southern tip of Ireland). I would also look closely at the possibility of establishing stations in locations not closely monitored by the London authorities; e.g., the Aran Islands or the Isle of Man.
Stations in the Hebrides, the Shetland and Orkney Islands, the Faeroe Islands, Iceland, and Greenland would be very helpful in monitoring the approach of storms to Europe from the Northwest. Additionally, a station in Iceland would help monitor the strength of the Icelandic Low (one of the poles of the North Atlantic Oscillation), and a station on Svalbard or Jan Mayen Island would help to monitor the Polar Oscillation. All of these stations would necessarily be radio stations and would most likely be reporting at night by sky wave.
Some Europe-bound storm tracks do emanate from the Atlantic, 45-60oN, 45-30oW, and head toward England or Spain (Wing). We would be dependent on shipboard observations (and hopefully the ships will accurately report their locations).
It would be nice to get weather reports from the Azores, the second pole of the North Atlantic Oscillation. However, these are under Portuguese control. While the Azores were the last part of Portugal to accede to Spanish domination, the overt resistance ended before the Ring of Fire.
It is conceivable that weather reports from Cape Race, Newfoundland will be useful, as the jet stream often passes over it, and we would want to have a radio station there anyway.
What powers other than the USE will do is even more speculative, but they would have less access to vacuum tubes for building continuous wave radios. They may still build spark gap transmitters, but those have shorter ranges, and their broad spectrum means that they can’t be close to each other without causing interference.
Still, it is clear that France and even Spain and parts of Italy have radio “telegraphers” by 1636. See Hunt, “Prison Break” (Ring of Fire IV) and a number of passages in 1636: Cardinal Virtues. Russia, too: Flint, Goodlett, and Huff, 1636: The Kremlin Games.
The difference thus is one of degree; the non-USE radio networks are sparser. It is possible that they will be supplemented in some areas by optical telegraph or even post horse messengers.
In part 4, I will discuss what the options are for weather analysis and forecasting.