Advertisement
If you have a new account but are having problems posting or verifying your account, please email us on hello@boards.ie for help. Thanks :)
Hello all! Please ensure that you are posting a new thread or question in the appropriate forum. The Feedback forum is overwhelmed with questions that are having to be moved elsewhere. If you need help to verify your account contact hello@boards.ie

My research -- index values in a theory of external signals "astro-climatology"

  • 10-02-2019 11:24pm
    #1
    Registered Users Posts: 14,279 ✭✭✭✭


    So, how to start a thread like this? As many readers of this forum know, I have worked on a research theory for many years in which I seek to understand weather variations from the perspective of external signals that are (in some rational framework lending itself to prediction) capable of producing responses in the atmosphere. I called this theory astro-climatology when I first started the work in the 1980s. If I had a chance to rename it, I would call it something different but not sure exactly what. The name astro-climatology is meant to imply that the theory somehow lies at a boundary between astronomy and climatology. But really it should probably be called astro-meteorology since I think the climate of various parts of the world is already well understood and what I am really trying to uncover is variation in the day to day weather events that form parts of those climates.

    Naturally the use of the prefix "astro-" can lead some to imagine that what I doing is astrology. There is no real connection. Most of the processes that I envisage are not of the straight-line mystical cause and effect sorts of things that astrologers produce. Rather, the majority of the index values in my research are complex patterns of response of the atmosphere at chosen specific points to presumed variations in the solar system magnetic field and geomagnetic field, which are then presumed to be interactive with the atmosphere. An astrologer would not recognize the concepts involved and in my approach, I am not looking for something that precise but rather an indication of trends or tendencies.

    The work has been summarized already on the internet on the Net-weather forum. That summary is in the science forum and was last updated a few years ago. I would recommend that anyone interested in this theory and research take a look at that over a few days, it does not lend itself to a quick read. Or you may prefer to follow this discussion and have a look over there a few weeks into this once some of the basic ideas are more familiar.

    I should say from the outset that the theory and the research are in a developmental stage and I don't expect this to lead to extremely accurate forecast results in my lifetime. I have seen some advance towards a non-random result that encourages me to continue.

    Another basic premise of this discussion is that the research was developed mainly for North American climates and then I attempted to extend the concepts to better results for Europe. At the same time I have investigated how the theory might work in eastern Asia and parts of the southern hemisphere. I have not had the time or keen interest to look in depth at the tropical zones.

    As I present some of the findings and discuss them, I could mention that my research for this part of the world has been mainly an analysis of the CET temperature and associated CEP precipitation data, but more recently, I have acquired and number crunched pressure data taken for a point close to Dublin (54N 6W) at six-hourly intervals from 1851 to end of 2014. This basically supports some of the CET and CEP findings although that pressure point is some 200 kms northwest of the CET zone. You could expect the pressure data for the CET zone to be very highly correlated in any case, especially over that length of time.

    My interest in Irish weather was therefore an extension of the decision taken to plunge into the day to day details of UK weather over on Net-weather (in 2005) and within a couple of years I had been invited by some members of both forums to join this forum, with the results being a daily forecast thread and getting involved in monthly contests and storm discussions.

    My rather naive belief was that if I just saturated my research working time with in-depth study of the Irish and U.K. weather, then I would start to get a really good feel for how the theory works in practice in that region, as I had done in my two home locations over the time of the research (central Ontario from about 1980 to 1995 and British Columbia from 1996 to 2019).

    This is almost enough of an introduction. I am going to take this thread fairly slowly so we don't rush into too many things to discuss at any one time. But the first point I wanted to make was that the research model or what I call the cumulative index value can be applied to any location with sufficient data resources, and analyzing why it shows the signals that it does can be handled in a separate investigation. In other words, like a lot of index values in climate and weather, a user (including myself) does not need to know why the index shows what it does, the user can start using the index values from the point of developing them. Whether they can be trusted to show anything non-random is then a point for investigation on both the validation front (was the forecast accurate or not?) and the theoretical front (can we see a reason why this index value shows what it does?).

    So to give the first approximation of an overview, the index values being used now amount to over a hundred in total, some of which are generated by the Moon's postulated interactions with an atmospheric grid, and many others which are independent of the earth-Moon system and originate in the solar system magnetic field. So take note then, there will be discussions of somewhat unrelated index values that can be related by taking partial segments of data (a non-lunar index can be studied by lunar orbital variables within the sample, and a lunar index can be studied by relation to different input from the solar system magnetic field variations).

    I think for now, I will leave this discussion with this opening, and spend perhaps the first week or so of the interaction getting into more detail about the grid being used (probably the first thing that should be explained) and some of the basic concepts involved once you understand that grid. It will all be new material unless you independently thought of any of this (and for your sake, I hope you didn't, this is not for anyone who wants a nice comfortable career in meteorology or climatology, let's say).

    Could I ask if people would just confine questions to material presented, rather than jumping ahead of where I am at in the discussion, to ask about things not mentioned or discussed. That will make it easier for the readers (who may number in the low single digits) to follow what we're discussing.

    I am hoping one result of this discussion will be my own education, this is not meant to be a lecture where I teach and you learn. This theory needs all the help it can get and I am pushing 70 years old (sometimes I have to review that, I am one of those older people who still feels about 45 inside). A clear and obvious conclusion is that this research will be difficult to bring to a workable stage (reliable long-range forecasts) and may never get all the way there. I am quite open to the concept that these findings are some portion of the total picture of atmospheric variation but not a high enough portion to make it a workable system without cross-breeding it with other theoretical approaches. So I keep my mind open to learning whatever I can about those other approaches to see if (a) they might be translatable to my own theoretical structure, in other words, unrecgonized consequences of astro-climatology under other names (I suspsect the JMO is one such creature), or (b) they are simply variables that can go into the equations on their own merit even if they don't have any connection to external signals.

    This is probably never going to turn into an equation-driven precise science like most of physics. It is going to remain a sort of technology or empirical study that makes advances in small and irregular ways, to give an example, perhaps being on the right side of normal in its predictions 60% of the time in an early phase and 70% at some later phase with the goal of reaching 80% and the necessity to reach 90% to get into the ballpark of what would generally be seen as reliable long-range forecasting.

    Okay, will promise to post again soon with some concepts used in this theoretical research, so we can begin to look at some signals (index values) of interest to Irish readers, and discuss what they mean and what hope there might be for basing any level of forecasting on these findings.

    Oh, and to answer one question that is bound to arise, do I use this theory very much day to day in making forecasts? Yes and no, I think it helps me assess some of the details of forecast scenarios to have this theoretical approach, but at the same time, I tend to base my forecasts on the same general approach that is used by all forecasters, guidance by numerical weather prediction models and various real-time observations. How I might see a pressure pattern or radar or satellite image might involve different concepts internally, but the consequences for a forecast are probably not so different that I could really say that I am "using the theory" to predict the short-term weather.


Comments

  • Registered Users Posts: 14,279 ✭✭✭✭M.T. Cranium


    This is a link to the Net-weather discussion thread. Most of that was posted in 2013-14 but there were a few later follow-up posts.

    https://www.netweather.tv/forum/topic/75437-roger-smiths-developing-lrf-model/?page=1

    (and as most already knew, MTC and Roger Smith and Peter O'Donnell all the same person).

    Now, as to this grid ...

    Workers in meteorology are used to the lat-long grid and sometimes divide the globe into sectors on some arbitrary basis that helps with their research.

    My theoretical approach rather early on pointed to a need for a grid that followed timing considerations and resembled the geomagnetic grid although not duplicating it. So what was devised was a "meteorological grid" sketched out below:

    This grid consists of two meteorological poles that are broadly similar to the locations of the magnetic poles (the southern one is more offset than the NMP by its location away from the ice plateau of Antarctica), then nine timing lines located radially and equidistant, with meteo-latitude determined by a correlation of upper level wind flow with the requirement to keep flow lines broadly perpendicular to the timing lines.

    You will see that timing line one is in eastern North America. According to the research, it curves southeast across the Atlantic into tropical Africa, crosses the equator in Africa and extends through the west-central Indian Ocean to the south meteorological pole. This timing line got the number "one" for two reasons -- I started the research in Ontario and noted timing similarities of astronomical events and low pressure over the western Great Lakes, and then discovered evidence that this timing line was also the one most likely to be selected for direct response of larger scale features to external forcing.

    Some time around the mid-1980s I had gotten as far as theorizing nine timing lines and the concept of directionality of progressive and retrograde signals related to solar system dynamics.

    Nine timing lines were postulated after trial and error analysis of where low pressure areas would be located at event times (for example, the full or new moon). They were numbered eastbound for no particular reason, which means that timing line three is the timing line closest to Ireland. Its equilibrium position runs from south tip of Greenland well to the south of Iceland and southwest of Kerry into southern France and then into the west-central Mediterranean. It ends up going across the equator south of India and Sri Lanka and continuing on through southern Indonesia curving back to the south past eastern Australia and into the south meteorological pole. Most of that need not concern us again in this discussion, except to say that broadly speaking, there should be some resemblance in inverted form between patterns over western Europe and eastern Australia, with seasonal differences factored in. I haven't had much time or opportunity to study this.

    So timing line three has this equilibrium position, but it can at any given time be oscillating east-west within about 20 degrees of longitude, so that on occasion it is as far east as Scotland to Netherlands, and on other occasions, as far west as central Atlantic to southern Portugal.

    The concept of directionality works like this -- postulated responses to events in the solar system magnetic field will take place to east of timing line one when behind the earth in its orbit, and to the west when ahead of the earth. This means that prograde events must be caused by sectors moving slower than the earth (so they gradually fall behind in this directionality concept), while retrograde events must be caused by sectors moving faster than the earth (so they advance from east to west). In the earlier stages of the research I began to find correlations between these effects and "field sectors" that were radially connected to planetary orbital positions. I will discuss that in a later section, after we look at some of the lunar signals in this research.

    Now, what is the reason why we might have a system of nine timing lines for events that are associated with the Moon's interaction with the earth? Let's consider ocean tides for a moment. Ocean tides are predictable phenomena that demonstrate that the gravitational tidal pull of the Moon on the earth's oceans lifts water levels as the Moon goes overhead and also behind the earth in its 25-hour (apparent) orbital cycle, and that there are larger ranges of ocean level variation when Sun and Moon are either colocated or opposite one another. But early research concluded before 1900 showed little if any similar response in the atmosphere. Why? Because the atmosphere does not have coastlines to stop the air from moving on. If there were a global ocean, tides would be a much different sort of affair than in our segmented land-sea planetary surface.

    However, I believe that the early research neglected to ask a follow-up question, where would the equivalent tidal energy go in the atmosphere? Just because we can't measure it does not mean it is not there doing something. And in fact, as bad luck would have it, the definitive study was done in Berlin, which just happens to be about equidistant between timing lines three and four. Had the same research been done in Chicago, I am pretty sure some early researcher would have noticed a pressure correlation with lunar phase, although not with daily lunar tidal peaks. So for the atmosphere, I have postulated that the tidal energy is dissipated over an interference pattern creating nine timing lines, but at the same time, the triggering events for low pressure (equivalent to high tides in the ocean, it is really the thickness ridge normally out ahead of low pressure that is the direct equivalent) are alignments of the Moon with the Sun, the gravitational equator and the larger planets. Each of these alignments creates a system of low pressure areas crossing timing lines. If two events happen to be less than a critical time apart, the low pressure areas will be double-centered.

    I noted also at an early stage of the research that there was quite a coincidence between average motion of lows in the main westerly belt, with the time required for an atmospheric wave to circumnavigate the earth at the same pace as the lunar orbit. This turns out to be around 13 degrees of longitude per day (360/29.5 = 12.2, while 360/27.32 = 13.1). Those numbers are the periods (in days) of two principal components of the lunar orbit (synodic and sidereal). This was an indirect piece of evidence that low pressure was being generated by interference patterns between the Moon and some fixed gravitational source. It turned out by examining the evidence that the sidereal peaks (declination peaks that I have called northern and southern Max) are as strong as the synodic peaks. The planetary conjunction peaks and three fixed-star interference patterns are relatively weaker. The planetary events would have mean periods between 27.32 and 29.5 days in the range of 27.5 to 28.0 mostly. The fixed star events will vary from the declination max in period only very slightly over long intervals, for the time of this research, no difference applies.

    So to conclude this session, the model is based on the idea that low pressure events will travel along to reach timing lines as generated by the timetable of astronomical events. Perhaps the nine-timing-line system originates from the fact that the average time separation of events is about three days, so the atmosphere's most stable response is to generate a nine-wave response if it is trying to incorporate a moving large-scale wave feature of earth-moon resonance.

    This leaves unanswered, why would the lows track at various angles to the grid and at various meteo-latitudes? This appears to have only small connections to the earth-moon origins of the interference pattern, and incorporates the other part of the research model, the atmosphere's response to solar system magnetic field variations. Those determine the route that the lunar interference waves will have to take, so the Moon is only contributing the timing portion of the outcome, and perhaps to some extent a range of latitudes, but the other factors in play create the path. And since those may vary from timing sector to sector, a diagonal path is a possibility (diagonal with respect to the meteo-latitude lines not to conventional latitude).

    Just for future reference, I have a system to define any given location by timing number as follows. The equilibrium position of the timing lines is given the value .50, so that a location on timing line one is at 1.50 in the system. Some place one quarter of the way east to timing line two would be at 1.75. The system starts at 1.00 and ends at 9.99 or 10.00 = 1.00. There are no values lower than 1.00.

    ====================================

    Click on this image to see the grid ... nine timing lines and three representative lines of meteo-latitude. Near the northern extension of timing line one, I have hatched the timing line in green and black to show the space travelled by the North Magnetic Pole (and therefore the North Meteorological Pole) since 1840. The last third of this represents the space by which the NMP has moved since I first formulated the grid. Therefore the grid has migrated by about 500 kms but I generally assume that the timing lines have not shifted much if at all except near the NMP. I will discuss this shift again later. As you will see this schematic of the timing lines is drawn over an actual prog chart, as I could not locate any blank hemispheric maps and my other maps of timing lines are not hemispheric. Later in the thread, I will produce a better map for discussion of events around timing line three and Ireland. Also, these are average positions, the timing lines oscillate east-west by about 25% of the space between them (so a total of 50% of the space between midpoints).


  • Registered Users Posts: 14,279 ✭✭✭✭M.T. Cranium


    So I should just repeat for emphasis, this grid is not presented as some sort of discovery, although you could say it is ... the only significance claimed for it is that it gives us a framework for studying the travelling low pressure systems around the hemisphere. The assumption is that once the timing lines are reset to current locations (and they don't oscillate in sync, some may be east and some may be west of equilibrium) then low pressure systems at times of astronomical events will be crossing them at various latitudes. The study of those various latitudes will yield some understanding of the second portion of the theory, setting up ridge-trough locations (and in some cases cutoff highs and lows) that represent the input of the solar system magnetic field.

    After a day or so, I will go into the basics of the lunar orbit and the astronomical events that I find significant to the model or theory. This will be interesting just as a recap of lunar orbital dynamics. Following that, I hope to give some proof from my research that these effects are real, after all, anyone is free to theorize anything they want (for proof, see rest of internet) but it helps if you can show that what you theorize actually takes place in nature. That's what the index values are designed to do.

    What if you don't live conveniently close to a timing line? Then you get used to the idea (as I have, living some distance east of TL 9) that your weather events will arrive with a lag from a timing line upstream, or in advance relative to the one downstream. But it stands to reason that the closer you are to the mean position of a timing line, the more orderly the progression of events relative to the timetable of astronomical events. Ireland is normally just a little downstream from timing line three. This means that in general, your peaks of cyclonic activity (troughs might be the better word) should come a few hours after astronomical events.

    You could expect, therefore, that there would be just slight lags between those triggering events, and lowest pressures, or strongest winds. But with the nature of low pressure areas being to pump up warmth and moisture ahead of their central cores, the peaks in temperature and rainfall should occur about the same time if there's a six hour lag to pressure minimum. Toronto (for which I did all of the initial research, as it has 179 years of data now) is just about in the same relative position to TL 1 as Dublin would be to TL 3 and given the way the meteo-latitude lines run, also at a similar meteo-latitude in the system. It's not a long way from Dublin to the CET zone. So in broad general terms I found similar results over long periods of time for weather stats at both locations and with the pressure data already mentioned for a point near the Irish Sea coast north of Dublin (54N 6W).

    I will go into this in more detail but anyone who can't wait for that should check out the Net-weather thread, that will have all but my later pressure analysis findings.

    What causes the east-west oscillation and what's the result of that?

    I think that what causes the oscillation is relative changes in the mass or angular momentum of the solar system. If the mass balance shifts ahead of the earth, timing lines will be drawn back to the west. If it falls behind the earth, timing lines will be drawn to the east. When a retrograde episode is about to start, Atlantic timing lines in particular will shift east. The effect is mainly that you get the events a bit faster, before the astronomical timing (because they are on their way to the eastward-shifted timing line). But with split flow and blocking more likely in these cases, you may also observe that many of the events miss Ireland by large amounts to south or north. When on the other hand a retrograde session is nearing its end, timing lines shift west. This tends to distort the lines of meteo-latitude also, and these would be times of anomalous southwesterly flow, with events arriving a good deal later than at equilibrium (up to a day or even 36 hours lag time can be introduced). Timing lines can sometimes get bunched closer together, or separated out more than usual, by the differential displacement effects. Since timing lines five and six represent conditions on the far side of the solar system (timing line one always depicts conditions in the solar system sector being traversed by the earth), oscillations are probably weaker on those (which may account for the less variable climatic regimes of eastern Asia).

    At this point in time, with the NMP rapidly moving west at high latitudes, there is likely to be some westward pressure on all timing lines, especially those running through the North Pacific and east Asia. I have the feeling (and so did other climate researchers in the past) that the hemispheric circulation is partly anchored by the NMP, which means that as it approaches northern Siberia or perhaps even northern European Russia in the next 30-50 years, the mean position of lowest heights in the arctic may shift away from northern Canada and into that zone, or at least try for a bipolar orientation. That should mean long-term cooling in some parts of the European arctic. The changes between 1980 and 2019 may account for some of the warming of the far northern Atlantic which is now close to the point of furthest longitudinal separation from the NMP (50 years ago that would have been near the New Siberian Islands). But in another 30-50 years, the point will shift closer to Greenland.

    I promise not to go on longer-term tangents but there are all sorts of potential applications of this general theoretical concept to long-term climate shifts, including the details of the ice age circulation. Some of the variables in my research model had much different orbital characteristics tens of thousands of years ago. It stands to reason that if they have predictable impacts on an ever-shifting geomagnetic grid, you could get radically different weather patterns without any change in overall heating of the atmosphere. For one example, some of the retrograde features now studied seem to be shifting poleward in winter. What if they shifted the other direction? What if strong blocking highs routinely moved out of southern Russia across central Europe into the mid-latitude zone of the Atlantic? This would really alter the winter climate at least on a frequency basis.

    (next post will start to deal with lunar events, the categories low pressure systems crossing timing lines, and show some of the features of the index value analysis related to them ... but it's previewed by the material in the other thread linked).


  • Registered Users Posts: 14,279 ✭✭✭✭M.T. Cranium


    So, I will make a start here in trying to explain the reasoning and evidence for this first concept about lunar interference patterns in the atmosphere resulting in travelling low pressure areas.

    First, I should exclude tropical cyclones. My research has indicated that there may be cause and effect within the other portion of the research model, in the form of disturbances caused by Mercury and Venus in field sectors. That involves some terminology that I haven't introduced so for now just file that under "something to be discussed later." I don't think the earth's Moon is any primary cause of tropical cyclones or anything that happens along the ITCZ unless it might be some diurnal tidal peaks at work there.

    So what exactly is the theory on this? Let's start with a simplification that is not quite a true approximation of reality. Let's say the Moon had no variations in orbital distance (therefore orbital speed), went around the earth in exactly 30 days, and toured the fixed star background in 27. (the actual numbers are 29.53049 and 27.32166). Let's also say that there were no other planets in our solar system, just three fixed points of gravitational energy, the galactic equator (traversed twice every 27 days) and the Sun. Then with a system of nine timing lines, if the result of the tidal force was to create both one eastbound travelling wave and one set of constantly moving interference patterns, the resultant lows would be likely to pass each timing line about every 3 days. (eight events in 27 days and the two other events, full and new moon, on a moving timetable relative to that fixed timetable). With no other planets, it might also be the case that our solar system magnetic field would be steady-state (no discernible sectors of differential flux energy) and so the moving lows should then travel roughly the same path all the time, perhaps with a seasonal north-south oscillation.

    The actual situation is more complex but not by all that much. The fixed gravitational sources are identified as follows, starting from Northern Max which is when the Moon simultaneously crosses the galactic equator (in the vicinity of Orion and Gemini, where we see a midwinter full moon) and achieves highest declination. Before going into the rest of the list, I should point out that the Moon, unlike satellites of all other planets, travels around the earth in the ecliptic plane (the plane that the earth travels around the Sun) and not the earth's equatorial plane which is the case for satellites of Mars, Jupiter, Saturn, Uranus and Neptune (Pluto is no longer considered a planet but I believe Charon orbits Pluto in its equatorial plane too). Why? Probably because the gravitational pull of the Sun on the Moon is relatively stronger than on those other satellites. Consider that Jupiter is 318 times the mass of earth, and the force of the Sun's gravitation out that far is 1/25 or so what it is for our Moon. So a satellite of Jupiter feels 318x25 times as much pull from Jupiter in relative terms (if it is at a similar distance to our Moon, and Io fits that description). Saturn's factors are similarly 96x100 so the relative force is even greater there. This doesn't matter to the theory except to say that the Moon climbs considerably higher above and below our equator as it follows the ecliptic plane, than those satellites would. And so there is a considerable declination reached, in the range of 18 to 29 degrees, both north and south of the equator.

    Those peaks just happen to occur when the Moon is passing the galactic equator (what we call the Milky Way when we see it in dark rural skies). The northern max position is in the direction away from the galactic centre and the southern max position is towards the galactic centre.

    Now another aside before moving to the menu of sources -- why is there a range of declinations? This is due to an 18.6 year cycle of precession of the nodal points of the Moon's orbit which is not exactly in the ecliptic plane but inclined to it by 5.1 degrees. So we add that to the 23.4 degrees of axial tilt of the earth to see that there is a declination range of about 18.3 to 28.5 degrees. That cycle last peaked in the year 2006 so we are around year 13 of it, with the Moon reaching declinations close to 23 deg N and S. The place where it is currently above the ecliptic is after northern max (where the full moon in March will be) and the place where it is currently below by 5 deg is opposite that, where the September full moon appears. These positions, the nodes and extremes, migrate in the opposite direction to the Moon's orbit around the orbital plane (every 18.6 years). The planet Mercury does something similar but takes tens of thousands of years to complete its cycle.

    So then, after leaving the Northern Max position (which is contemporaneous with full moon around the winter solstice), the Moon passes close to the massive star Regulus around 3-4 days later depending on orbital speed, and Spica around 6-8 days later. About two days after that, approaching the southern max position, it passes fairly close to Antares (usually a larger separation but it can be an occultation around years 2 and 3 of the declination cycle). This is a different event than the first two in that a second significant source, Aldebaran, lies almost opposite Antares in the sky, so that there is a combined event that I label "A" in the model. The Regulus event is RC (for Regulus conjunction) and the Spica event SpC (for Spica conjunction).

    Here I need to add the detail that every gravitational source postulated to be significant in the model has to be treated as both a conjunction (where Moon passes the object) and opposition (where Moon is on opposite side of the earth from that object) set, just as high tides occur with the Moon overhead and behind the planet, in effect, underneath our feet. (various angles apply here obviously, and the 18.6 year cycle means that at different points in that cycle, the Moon displays different separations and can sometimes occult the sources -- move in front of them from our perspective.)

    These "stellar" events are the weakest in the model but not as weak as you might suspect. In early stages of the research, I identified the RC event as a "northern resonance" thinking that it was the N Max event from upstream arriving at the timing line to complete the continuous flow of energy required to sustain an interference pattern. That may in fact be mostly what the RC event really is. But then for a while I was thinking that the SpC event was a second resonance or perhaps a forcing created by the Moon's crossing the earth's equator (which it has to do at some point between RC and SpC events). The A events tend to come 2 days before N and S Max and so are quite energetic since those are high energy peaks and a low will be forming to arrive downstream at the next timing line. So there, I am not too sure that the energy can be directly related to the actual source(s).

    Then we reach southern Max with the Moon crossing the galactic equator. In the early winter, this will happen just before new moon (in the current timetable, we happened to have one of those rare combined full moons and northern max peaks on 21 Dec, so the next S Max occurred around 4 Jan and the new moon around 5 Jan. Here's a principle to keep in mind -- northern and southern max make steady progress forward against the timetable of full-new moons, so that a northern max will overlap the new moon closest to 21 June, and a southern max will overlap the full moon event at that time (this year, I think the new moon is later than the summer solstice and the full moon a few days earlier).

    Another tangent if you will indulge me, but something very significant for longer-term research and also something touching on archaeology because as you may know, the ancients were very interested in the lunar orbit and even seemed to be tracking declination cycles. This tangent is to point out that we just happen to live in an epoch where the Moon's declination maxima overlap the galactic transits and where all of that occurs at the solstices. As recently as Roman times, the Sun crossed the galactic equator in November and May, so that northern max (if a gravitational event and not a declination driven result) would not be northern max but instead would be something like outer galactic equator conjunction. There might not be any discernible energy peak from the declination max. Going back further into antiquity, the Sun crossed the galactic equator at increasingly early times relative to the seasons until we reach a point where summer was when the Sun crossed the current S Max position. All of this probably hurts your brain as it hurts mine, but to the extent I have time and inclination to study it all, I find interesting possible correlations with ice age climatology as understood (which may be too strong a word).

    So the system I am describing to you is not fixed, it slowly changes over the centuries. It is just a coincidence that around the time weather records started to be kept, the N Max was moving into the solstice range (it was probably overlapping full moon around 16 Dec in the Maunder, once the calendar was shifted to Gregorian, before that it would be the 5th of Dec). We have enough daily data that I can reconstruct some faint evidence of a shift in solar-galactic waves in the temperature patterns for both CET and Toronto. For example, the January thaw phenomenon has been slowly shifting to later dates. Anything created by long-term interactions of this sort should be shifting later in time like the causative interactions are shifting.

    Alright, to get back to the tour, after S Max, the Moon traverses a relatively empty portion of the sky. But there are RO and SpO events as it moves opposite those sources. About two days before returning to N Max, the Moon passes Aldebaran, a component of the A event generator so these events are not AC and AO but just "A" both times in the system.

    Well, back at northern max but the Moon is not quite full yet, as the earth has completed about 1/13 of its orbit and it takes that extra 2.3 days to reach the next full moon. So that late January full moon is about midway from N Max to the RC event. The February full moon will be very close to overlapping the RC event (Regulus is in opposition around Feb 18th).

    Now add to that fixed timetable (which repeats 14 times a year and a bit, to include 13 lunations (full to full moon) in the same interval), a number of variable-positioned planetary gravitational sources. At this particular point in time, both Jupiter and Saturn are close to the S Max position and closing in on a pass which we see every 19.7 years. I've mentioned the significance of that in the solar variation cycles which are quite close to the halfway point generated by mutual alignments (on opposite sides of the Sun). Venus conjunctions will always be within four days of new moons because of its inner orbit, and Mercury conjunctions even less separated (although this is a weak event that I do not track after some trial and error). Uranus, Mars and Neptune also show faint energy peaks in the analysis of lunar events.

    Here again, the conjunctions (for Jupiter, labelled JC, for Saturn, SC, etc) have analogous opposition energy peaks (JO, SO etc).

    The reader, if patient enough to continue on, may be wondering then, what can be the physical justification for proposing relatively similar results from objects of such different mass and distance from earth? I am (only too) familiar with the objection that classic gravitational theory does not allow for this to be remotely possible, and even the full and new moon "events" are called into question.

    My answer to this is that with the peaks appearing so similar in their relation to the timing and arrayed in terms of a slowly falling intensity when compared to classic mass over distance squared, that the best fit of the results suggests that the effect is being generated at (get ready for this) the twelfth to fifteenth root of mass over distance. Why this is so, I have no idea, but speculate that this is how gravitational waves are generated. So while there is still a hierarchy based on mass and distance, the reduction for growing distance and falling mass is very subtle. My only hope here is that first the theory itself would prove to have some validity as a predictive framework and then some scientists more able to study gravitational waves and physical processes on cosmological scales would take up the challenge to verify what causes this effect to work at such an arcane relation as twelfth to fifteenth root (I would be more specific but the range of intensity variations produced by different angles of the Moon's separation from sources makes the precision rather indefinite) of mass over distance. However, we already have the situation that gravitational force is mass over distance squared, gravitational energy is mass over distance. So it is distance that apparently can be the moving party in the various manifestations of two-body interactions. Why not some third concept that is scaled down further?

    Here's one example of how the mass-distance considerations are levelled. In conventional terms, Jupiter's gravitational force on the earth (and Moon) is 318/95 x 100/25 times that of Saturn (masses are 318 and 95 units and distances 5 and 10, so the inverse square of distances 100 over 25). That works out to about 13, so we can say Jupiter has 13 times as much pull on our home planet and its satellite than Saturn manages. But taking the twelfth root of mass over distance, those values (in arbitrary terms) are twelfth root of 318/5 and 95/10, or of 63.6 and 9.5. (those are the ratios of gravitational energy by the way) ... so the square roots are about 8 and 3.1 in approximate terms. The fourth roots (square roots of those) would be close to 2.75 and 1.75. The eighth root (root of the preceding) would be about 1.7 to 1.4. By the time we then get to twelfth root, we are looking at just a slight differential, something like 1.5 to 1.3. And taking an arbitrary stellar mass and distance on this same scale, the fractions derived are substantial (around 0.8 on this same scale). This greatly levels the playing field for all significant masses at all sorts of variable distances.

    Well, another question might be, what about all the discrete gravitational sources near the N Max position, such as stars in Orion, Sirius, Gemini? Two things about that, one, the separation between Moon and these sources may exceed a critical limit which seems to be 10 degrees, and two, their masses may just blend in with the larger component of galactic mass (vs its greater distance). In any case, separate energy peaks do not appear in the research for these nearby to N Max sources. The S Max situation is different in that you just have the galactic equator and no massive nearby stars other than Antares which is already handled as part of the A energy peak.

    That's about it for this session. The more interesting part comes next. Obviously, anybody is free to theorize any fantastic or bizarre hypothesis they imagine, but if you can't find evidence to support it, there is no point in discussing it. For this, I have evidence. Otherwise I would not believe this to be possible either. What sort of evidence could you find?

    This goes back to the grid discussed in a previous post. Once you establish that a location with weather data is close to a timing line (ideally just east of one) then you can begin to investigate the data for signs of pressure, temperature, wind and precipitation peaks (or troughs) that reveal the timed passage of low pressure systems that would be the postulated effects of these postulated causes.

    How does that work in practice? I will get into that in more detail next post, but as an opener, would say this -- if the Moon had no orbital distance variations, and travelled at a constant speed around the earth, then the sidereal (fixed star background) events should line up almost perfectly in 27.32166 day time intervals and the synodic (full, new) should also line up every 29.53 days and change. They won't quite do that because of the distance and speed variables in the lunar orbit, amounting to roughly 10% faster and slower motion, getting the Moon to some events up to 1.5 days earlier than average and others 1.5 days later. This perigee to perigee cycle takes 27.55 days so it is a bit longer than the sidereal cycle. It moves around in the same direction as the Moon's orbit over a period of 8.85 years. In recent months the lunar perigee has come just after Northern Max, which is why the recent full moon in January was a "supermoon." Therefore, also, the Moon is moving faster in orbit around there, so the time difference this year from N Max to RC events is as small as it can be. About 4.5 years from now, it would be a longer interval by over one day in four to five days (moon would then be at apogee).

    A really precise study needs to take the data and continuously reset it to intervals that centre on actual event times, not the mathematical averages. However, that takes a lot of time and effort. The first order investigation is to look at the mean periods, then start case by case studies to see if the signals (which are already there in the coarser moving data columns) can be better focussed. The general answer to all this is, if you find a signal of roughly 10 units in the average of all data, it's probable that the actual signal is 12-15 units when you align all the data more precisely. This is quite often what I find when doing the more detailed analyses. But some of the signals are quite robust even in their coarse form. I have found some particular lunar signals near timing line 1 for example that reach values like 15 mb pressure waves, 3 C deg temp spikes, 4 x random precip and considerably higher wind speeds.

    So, will get into that sort of discussion for timing line 3 where the results are not quite as great in magnitude, although some of the pressure waves are similar. The magnitude of signals will always be related to the climatic variability of the location chosen to study them. I would imagine a study of timing line seven on some island in the Aleutians would yield more impressive pressure signals than any of the other variables. But when you get into east-central North America, you're dealing with a climate with large variability so the signals are quite a bit larger than in western Europe.

    (at this point, if you can't visualize the path of the Moon in its orbit, this might be the time to click on the previously supplied net-weather link and at least read through to the early point where I sketched out the lunar orbit, however, you'll be seeing where the various planets were in that orbit six years ago.)


    Some questions I also investigate in the research ...

    (a) if the events are indeed simultaneous at timing lines with astronomical events postulated to cause them, is there some other mechanism that might be producing this effect? Is it possible that the external signals are mistaken for processes within the atmosphere on these time scales? Is it possible that the only real events are full and new moon and all the rest are resonances on different timetables?

    (b) what causes intensity and latitude fluctuations, once it is established that the timing lines are a valid timing focus?

    (c) are there historical shifts in locations of timing lines and therefore the grid itself, that can be discerned from data analysis?


  • Registered Users Posts: 14,279 ✭✭✭✭M.T. Cranium


    I am working on an excel file of pressure readings for Ireland (at 54N 6W) with some analysis, but it may take a day or two longer to get it into the sort of shape where you can follow the discussion. I may have to create an external website to download the excel file, it is rather bulky.

    Probably best to start the discussion with evidence of pressure variations conforming to the theory outlined, then work our way through evidence of temperature spikes related to the lows, precip data and relevant wind analysis in that order.

    The sort of process that I follow, as mentioned briefly, is to array the data as closely as possible (given the decimal values involved) to the periods being studied, then see if any second order variability can be detected that might firm up the timing even more.

    By doing that sort of analysis near timing line one (in North America) it came to light that there was apparently a westward shift of this timing line in the late 19th century as events seemed to be arriving somewhat later for several decades (up to a lag of one day) then evidence that the timing line shifted east and may have been further east than it ended up in recent decades, around the 1940s and 1950s. I have not really gotten so deep into the Irish-UK data that I have found anything similar to that, in fact timing line three seems to have been just about where it is now for the entire 250 years of CET daily data according to the timing of events.

    When I have this excel file of pressure ready to post, I will return to this thread and we can have a look at some case studies. I'm really interested in any feedback from a discussion of the evidence, more so than general discussion about the theory being valid or not, since at this point I'm more or less in for the long haul on validity and looking for improving results.

    By the way, this is not necessarily related to the theory but the pressure analysis shows an interesting drop in mean daily pressures in Ireland just about the time of the autumnal equinox, and it's quite a sharp drop then. There is a fairly large rise in mean pressures during February to reach a peak towards early March. I think some of these kinds of findings go more into the climate base portion of the model than any forecasting application.


  • Registered Users Posts: 14,279 ✭✭✭✭M.T. Cranium


    Just testing to see if I can upload the excel file which is ready for use now.

    Well, my file is way too big to upload here. Going to plan B now, creating an external website where it can be uploaded. Back with the details on that when it's set up.

    There is plenty of interesting easy-to-follow graphical content in this file, and it should make for a good discussion. Thanks for your patience.


  • Advertisement
  • Registered Users Posts: 14,279 ✭✭✭✭M.T. Cranium


    Success, the excel file has been uploaded onto the Net-weather research thread. The link is here:

    https://www.netweather.tv/forum/topic/75437-roger-smiths-developing-lrf-model/?page=3&tab=comments#comment-4016019

    Assuming that you can access the excel file from that location (it's the most recent post until somebody comments, bottom of page three), here is a guide to what you will find in the file.

    1. The daily pressures, four times daily, taken from NOAA 20th century archive maps as viewable on wetterzentrale.de, for the grid point 54N 6W are listed in column G. They appear without decimal places x100, an example of what they should read in mbs is shown in cell H2.

    Date and time information run down columns D and E.

    2. My analysis of 20-year intervals of average daily SLP can be seen in columns I to P. The averages for 2011-14 appear in column S. There are averages for the entire data set (weighting the columns so that the last four years have appropriately smaller input) in column Q. Column R has only the monthly averages for each of the four principal times 00z, 06z, 12z and 18z. The analysis follows the date and time codes for 1851 to end of February, then inserts 29 Feb (which did not occur in that year) and so the rest of the data appear one day after their date code from column D. Since 366 x 4 = 1464, the year runs from row 2 to row 1465 (in the daily pressure analysis).

    3. Then my analysis of lunar events follows in columns AD to AN. This is based on the following system. As the data begin with 1851, the new moon occurred on 2 Jan 1851, so that I took the averages for each lunation (29.5 day intervals) starting then and jogging all other years into line so that a selected new moon date was in the same position. The choice was based on keeping the first full moon in January (the first of two if there were two full moons in January) in the first lunation. This means that the first lunation averages data starting around 16 Dec to 15 January in earliest cases and from 16 January to 14 February in latest cases. It is therefore a sort of analogue to January without being entirely composed of January data. And the lunations continue from that starting point, using a 13th lunation whenever necessary to return the data to the right starting point (in practice, seven out of nineteen years will require a 13th lunation).

    Graphs of the results for each lunation appear down the right side of the data columns (approx cols AS to BA).

    A composite annual graph is shown, and a comparative graph of mean daily pressure for calendar dates and lunar-adjusted dates on a broadly similar timetable.

    That one is quite interesting, it shows that the second-order variations in the annual pressure correspond to variations in the lunar pressure.

    4. A final section of this excel file (column BI) shows the adjustment of lunar data from synodic (new moon to new moon) to sidereal (in this case, from one southern max to the next) by averaging the data out for the 13 lunations. This shows clear evidence in graphical form of pressure troughs at the S Max and N Max times in the sidereal cycle. Some of the other sidereal pressure variations are better seen during the study of the various lunations rather than overall in the data.

    I am going to leave the discussion at that for now, take some time to look through this link before I get into case by case analysis of what the pressure variations for each lunation show.

    Will start to post some commentary over the weekend. Please post any questions that arise as to general assistance issues like what is shown here or where do I find such and such in the file. Anything more related to the results, I will defer answering until we reach that stage of the discussion later on.


  • Registered Users Posts: 14,279 ✭✭✭✭M.T. Cranium


    So, I hope readers have been able to locate the excel file with daily pressures and analysis (see previous post).

    Let's start out by examining the graph of daily average pressure (1851 to 2014) which can be viewed around cells U-AA 9-18 in the file, to the right of the data generating the graph.

    The most obvious characteristic is an annual curve that approximates the temperature curve for the year although it peaks in June rather than July-August.

    Some second order variations appear interesting whether from this research perspective or any other weather interest.

    There is a relative minimum around mid-January followed by generally higher pressures in February. Another minimum occurs towards mid-late March.

    The rise through spring is fairly steady and a plateau around 1016 mbs is reached in June. Then there's a bit of a minimum by late July into early August that coincides with the time of maximum heating. It probably signifies a statistical peak in "thundery low" type disturbances against an otherwise fairly static mid-summer trend.

    Some of the highest pressures of the year occur in early to mid September as the trend returns above 1016 mbs for a while. Then towards late September, pressures fall off quite rapidly and by early October a different regime is evident with the average closer to 1010 mbs.

    If you examine the 20-year averages (in columns I to P), check the data around row 1098 (1097 data points into the year, as there are four per day, that is 274 days in, or around 1st October which is indicated to right of the data columns).

    On average this late September pressure slide ends on 1st October (Q1099 is 06z 1st Oct, 1012.56 mbs), rebounds slightly and hits a second minimum around Q1133 (18z 9th Oct, 1012.54 mbs). After that, pressures reach an absolute low of 1009.56 mb at cell Q1203 (27th Oct 06z). This is in fact the lowest average of the calendar year until a few data points in December drop lower.

    These features seem semi-permanent in the data. For example, the late October minimum has had these average minimum values and timings over the 20-year intervals used:

    1851-70 ___ 1007.3 ____ 26th 12z
    1871-90 ___ 1009.8 ____ 26th 18z
    1891-1910__1006.9 ____ 29th 06z
    1911-30 ___ 1001.6 ____ 30th 12z
    1931-50 ___ 1008.7 ____ 25th 18z
    1951-70 ___ 1008.4 ____ 28th 06z
    1971-90 ___ 1013.4 ____ 27th 06z
    1991-2010__1006.9 ____ 23rd 06z
    2011-14 ___ 1004.9 ____ 25th 12z

    Another relative minimum occurs on 7th December (1009.4 mbs) and this one also shows up fairly regularly in the 20-year intervals. A final relative minimum (1009.5 mbs) takes place on 30th December and the intervals show this one somewhere close to that date. From 1891 to 1930 the averages around 29-30 Dec are below 1005 mbs.

    So this is the background against which we will be measuring the lunar variations in pressure through the year.

    I will return to that (more significant to the research) analysis after a short break, wondering if there are any questions or comments about the annual averages before moving on ...


  • Registered Users Posts: 8,219 ✭✭✭Gaoth Laidir


    I'm with you so far. Just wondering where you got the 6-hourly data for that location and how does the reliability change the further back you go?


  • Registered Users Posts: 14,279 ✭✭✭✭M.T. Cranium


    A forum member kindly sent me the file, about two years ago, and it took a while to work out how to set up the research files. The pressures are supposed to be readouts from the NOAA 20th century archive maps (as they are called, I think they started out that way and worked back from 1900 into the 19th century too). You would be familiar with those maps over on wetterzentrale. Whenever I check a map out it seems accurate as to the readout values.

    Probably the older maps are reasonably accurate over the region we are investigating here, I wouldn't be as confident about pressures shown a long way from land although ship reports were numerous back in those earlier years.

    We get into a more interesting part of this tomorrow or whenever I post the next bit, as some of these annual variations seem to be linked to lunar cycles in ways that may give hints about processes.

    One thing we can clear up before getting into that, lunar dates don't cluster in any non-random way. As you may know, they return to similar dates every 19 years. The actual math on that is worth a look. The dates would be just slightly ahead every 76 years when the number of leap years is not an issue, let's say on average about six hours or one data point earlier. However, any 76 year interval in this research that spans 1900 would be 18 hours later on average because of one missing leap year in 1900. Since there are three missing leap years every four centuries, they have a larger effect on the 19-year cycle than the faster pace does, the cycle would move forward one day every four centuries but the missing leap years push it back three, so for a net change of two days (for example, dates after 456 years would be two days later, if they spanned three missing leap years as in 1820 to 2276, but 1851 to 2307 would jump to three days later.

    Now also data eight years or eleven years apart will have average time differences of 1.5 days (but eleven years is often a day different as it's frequently three leap years intervening). The rule is that eleven years from now, dates will be 1.5 days earlier and eight years from now, 1.5 days later. When you combine all these near-misses with the repeating cycle in nineteen years, you find that lunar dates are fairly evenly spread out over the calendar. I have looked at data in intervals of similar lunar dates when something interesting shows up to see if the effect is concentrated on a range of dates. Using that method, I found one particular event that had quite a rapid growth-decay signal for specific ranges of dates indicating that a particular time separation of two events was critical.

    But having lunar events (such as new moon dates, full moon dates, northern and southern max dates, etc) scattered at random through the annual calendar means that lunar signals would not necessarily ever show up in an annual pressure analysis (or any other weather element analysis). There could be a large signal but it would occur at random on various different calendar dates, all cancelling out. However, if an event is strong for a brief portion of the calendar year, then it will leave a mark on the annual calendar year data.

    This seems to be the case for the late September pressure fall and the October first minimum that we worked through earlier.

    Anyway, if you wanted to have just the data that I started out with, basically you would need to download the linked file and erase everything in columns to the right of G, as that was the file in its original form. I have created everything to the right of G, and so far have also added nothing lower than the last daily pressure readings which are way down around row 239620 or some such number.

    The specific lunar date groups (19 groups) are not in the download file, if I find things worth showing in those, I may at least copy the blocks of data being discussed. I find having more than two analysis groups in a file this large slows down the computing speed to the point where my computer freezes up regularly so the file we have is probably about as large a segment as I would want to open up on my system, not sure how it's working for anyone else on their devices. I hope it doesn't crash.


  • Registered Users Posts: 14,279 ✭✭✭✭M.T. Cranium


    Starting to take a leisurely look at the lunar data now ... I am going to take this quite slowly because even for those few who might be quite interested, time may be limited and the subject matter is a bit complicated, so small steps ...

    I will remind patient readers that the link to the excel file is back in post 7 over on Net-weather where I can upload the massive file (here I could maybe upload a couple of days worth).

    So assuming you have that open and available, let's go into the lunar "normalized" annual data columns which begin at column AD. Each one is a 20-year segment and labelled at the top in row 1. The data file as you may recall starts in row 2 because it has its own headers in row 1.

    First point, what is meant by "normalized lunar year?"

    Well, if we happened to have a more co-operative satellite that would perform twelve orbits of the earth every calendar year, then our calendar months would be lunar months. There would always be a new moon on some selected date, whether January 1st or perhaps a bit later, and a full moon around the middle of each month. And if we could order up such a system, I suppose it would be good also to have a 360-day year and a 30-day lunar month. This by the way is the "synodic" cycle, when I speak about the "sidereal" cycle which is shorter, I am referring to position of Moon vs fixed star background.

    In reality, we have 12.3 lunar months, or "lunations" in each calendar year. This is why cultures that use a lunar calendar (such as the Jewish calendar in use in Israel) need to add a 13th month every so often to keep their variable 12 or 13 month years in sync with the solar timetable of the seasons.

    A lunation has no particular start or end but should run from defined lunar phase to next same phase. A lunation can be this new moon to the next one, or this full moon to the next one. It's just a unit of time in lunar terms. Its precise length is 29.53049 days. But the advancing perigee cycle makes that an approximation. Certainly the half-points are quite variable, a full moon can be 13.5 days after a new moon or more than 16 at extremes (the average is 14.77). If perigee comes halfway from new moon to full moon, then the duration will be short for that half-lunation, and the following half-lunation will be relatively long. Other minor factors intervene too, the gravitational pull of other planets, the speed of earth in its orbit, related to axial tilt. But we crunch the numbers mainly by averages rather than precise data input. When I do a more precise study using the data input, then I find slightly sharpened peaks and troughs but you don't get so much additional precision that it really justifies that effort. The more significant variables in any case are lags in timing caused by oscillations in the grid, and the hit or miss nature of latitude of storm tracks.

    That was a tangent, back to the actual business at hand ...

    Since the data begin in January 1851, the data sets are defined by lunar dates in 1851. Almost ideally, I would say, the first new moon in January 1851 was on the 2nd near 11z, so fairly close to data point 6 (in row 7).

    The data analysis then proceeds (as I mentioned earlier) by taking the first new moon in each winter solstice period no earlier than the 18th of December. This sets up the first full moon of the data analysis to be a full moon in January. The range of dates is actually closer to 3 to 31 than 1 to 29 as I had intended, just because most of the data jogged later after missing leap year 1900, and luck of the draw in terms of where perigee was falling in a few close-run cases where January had both an early and late full moon.

    This presents no problem in practical terms, you could start this analysis anywhere in the calendar year and it would cover all cases. But the method of selecting start points on the first day before a new moon in the eligible time range 18 Dec to 16 Jan means that we are normalizing the data set to approximate the calendar year. There are more or less the same number of dates in each row before as after the 1851 data, and the timing for 1851 (new moon on 2nd) is the reason why we start on the day before new moon. I allowed the mean period to determine where the data sets began, although at frequent intervals I would check actual dates from an astronomical ephemeris to make sure the dates actually were within a small tolerance hitting new moon dates around the range of 5 to 8 data points into each annual segment (keeping them all aligned would get into dropping or squeezing data points here and there).

    So what we come up with is an average of pressures for dates of similar lunar phase as close as the Moon's irregular timetable allows to overlap the calendar year. The data for normalized 1st January is really a selected average of data in the time range 17 December to 15 January. But it's always the day before a new moon.

    Starting the data set near the winter solstice means that the shorter sidereal cycle also begins from its southern Max position on 1st of January 1851 but actual southern max data positions will be scattered up and down by about one day due to the ongoing separation of the sidereal from the synodic timetables. A new moon near the winter solstice will have a coincident southern max event. By mid-January, the southern max is about 2 to 3 days before new moon. Those differentials can best be studied by taking the 19 similar sets of lunar dates arranged from earliest to latest (in this case, 1851, 1870, 1889 etc is 10th out of 19) and looking at averages to see if patterns can be determined for gradual increasing separations of synodic and sidereal events. I found that you could see a clear pattern of the separations in January. By February, the separation is becoming 4-5 days and more easily shows itself in the synodic data analysis.

    Now, let's see what this analysis of 20-year intervals and overall averages contains. We could spend a long time looking at this. However, I wanted to get an overview of what is shown. To the right of all the lunar normalized year data, there's one large graph that shows the annual cumulative pressure pattern. It is 13 lunations long because there is an overflow of data into the 13th lunation. I have presented the 13th lunation as an equal set so it contains 12/19 of the data of the first lunation data set. When I examined the 13th-only data (the stubs of years 2, 5, 7, 10, 13, 15 and 18 in the 19 years starting 1851) it looked similar but as you might expect with 7/19 of the data set size, more variation from high to low. So I don't think there's much scientific difference between the findings for lunation 13 whether we consider it to be unique data or derived in the same way as the other 12 lunations.

    What this means is that the normalized lunar year runs for 384 days and days 367 to 384 should be broadly similar to days 1 to 18.

    The normalized lunar data over the year can then be directly compared to the previously calculated annual averages for calendar days (no lunar input) back in section 1 that was graphed around cells U to AB 9-18.

    I repeated the data column (Q) that formed the basis of that graph so I could graph both the lunar and calendar year data together.

    That result is below the 13 small graphs of the lunations and you can see it by scrolling down from the large "lunar year" graph past all those small graphs to cells AU to BH 300-325.

    The calendar year is in red, and the lunar normalized year is in blue on this graph -- same time scale, the calendar year is shorter by 18 days but all data compared are for (within 15 days of) the same dates.

    What jumps out of this graph is the obvious fact that the lunar variations are superimposed on the annual curve, as you might expect. They quite often vary up and down by about 2-3 times the variability of calendar year data, but not always. It made me curious to see whether any of the annual ups and downs were more tied to calendar dates, or clusters of lunar dates that left their imprints. The results are somewhat ambiguous but I did find one very clear case of the lunar cycles making an imprint. We'll come to that in perhaps the third day (or time) of this discussion.

    That's already a fair bit to absorb. You probably noticed scrolling down through the graphs of lunations that some of them show some rather distinct signatures of events. This is where I plan to end this first day's discussion, we can get into details on day two ... the rationale for the twelve lunations (and the 13th which I posted beside the first one since they partially share data) is that the normalized dates shown for each time frame (lunation) refer to the 1851 calendar and therefore the average calendar year connection to this data set. It does not mean that this is what the lunar pressure curves should look like every year on dates mentioned. It means that once you've established the time differential of the year in question, this is what they should look like (individual cases will vary all over the spectrum but these tendencies should be observable). For 2019, our first new moon in January was on the 6th and our first full moon on the 21st. This means that the dates for the lunation graphs (and the entire data set) normalize to 5 January onward, and 4 days should be added to all dates shown to place the analogies into 2019 time frames. Last year would have been more like 15 days, just about at the last gasp of the data in the first lunation, so referring to about 16 Jan 2018 to 4 Jan 2019 (if you took the first 12 lunations, which you would because 13th lunation strictly speaking is only referenced for years that require additional time to get back to the next starting point, 2018 dates being later than 2019 in the system, this was not the case).

    Next time the discussion will be about lunations 1 to 7, and some of the specific patterns found in the winter and spring quarters of the year. Lunations 8 to 13 will take us through summer and autumn data.

    (the average sidereal cycle of pressure can be seen around column BI, but differences from lunation to lunation are quite important as the interplay between three factors, synodic time scale, time of year, and timing of perigee, are all quite significant, meaning those average signals for the sidereal cycle are overwhelmed by other considerations from month to month).


  • Advertisement
  • Registered Users Posts: 14,279 ✭✭✭✭M.T. Cranium


    I got into some unexpected trouble with other computer matters here this week, sorry I have not been pushing ahead with this discussion.

    There will be a post up later today about the first half of the 13-lunation cycle as discussed. Meanwhile, I did post the complete list of minimum pressures over on another thread I started about annual low pressures.

    I have done a census on those dates (164 minimum winter pressures plus 51 additional minimum annual pressures that were not also minimum winter pressures), 215 events in total. The count is very similar around both full and new moon dates, a peak occurs around the three day interval starting two days before these events and ending on the date, with a secondary peak about two days later (which I interpret as frequent northern and southern maxes in the peak months of Nov and Dec which follow new and full moons). The peaks are not astounding but certainly on the order of a 2:1 bias vs random scattering for the dates closer to full and new moons (the term syzygy applies to either/or but that word is a bu**er to type in the dark).


  • Registered Users Posts: 14,279 ✭✭✭✭M.T. Cranium


    Looking now at the first half of the year in terms of average pressure during the lunations as described in previous posts. The link to the excel file being used is back in post #7 above. Be patient, it takes a while to load up.

    The discussion will now proceed to a series of graphs of data for lunations (1 to 7) found in columns AU to BA and starting with lunation 1 betwen rows 29 and 43 of the file. This graph shows data from column AP. (the data from AQ are used to generate a graph at the bottom of the lunation section, and are just column Q repeated)

    Lunation 1 then is a cross-section of lunar-date-centered pressure data that falls in the interval of 17 Dec-15 Jan to start, ending 15 Jan-13 Feb. As explained earlier, I have tables of data available outside the linked file which break down the data into 19 groups of similar lunar dates scattered at roughly 1.5 day intervals through this section of time. I can use those to analyze further the trends spotted in the lunation graphs. Eventually I will have that set of data on a linked file, as there are also some perigee and lowest pressure files that would not fit into the 10 mb limit of the current working file.

    In lunation 1, note that the most prominent pressure feature is a minimum around data points 48 to 60, the absolute minimum occurs at data point 52 (row 53 of the file), which would correspond to 13th of January in the 1851 data used to centre these lunations. That is 12 days after the new moon on 2nd. Within a tolerance of about a day then, this data point is very close to the northern max position of the Moon, after a new moon in the time range 18 Dec to 16 Jan, and 2-3 days before full moons in mid to late January. So it's the Northern Max event rather than the January full moon that centres the pressure signal. And looking at the 19 timed groups of data that advance from small 1-day to larger 3 or 4 day separations with later full moon timings in January, it's evident there that a second pressure minimum begins to develop towards the end of the data. I have done some investigations to determine whether there is any evidence of a critical time differential which is perhaps inhibiting to low pressure (meaning a time differential that is too large to create one event and too small to accommodate the motion from timing line to timing line). There is some faint evidence for this, around a 10-15 January full moon. Certainly by a 20 January full moon you get into an ideal 3-day separation for N Max and full moon, and data analysis both in this set and the Toronto data show a peak of intensity for a N Max - Full moon double hit when full moons are later than about 18th of January. The earlier pressure minimum around the start of the lunation (southern max average data point 3 and new moon average data point 6) is less pronounced than an analogue in the slightly earlier lunation 13 data (graph beside lunation 1).

    Lunation 13 in my data set is an equal portion of the data which includes the 7/19 cases of data absent from both lunation 12 and lunation 1 graphs (because the next year starts 20-30 days later) or the 12/19 cases that are also present in the lunation 1 graph.

    Not shown, but I did generate a lunation 13 data set that leaves out the duplicate data. The only difference it made to the appearance of the graph was to sharpen up the amplitude especially around the mid-lunation N Max pressure fall which by the way is similar to the lunation 1 set. Since lunation 13 timing is generally ten days ahead of lunation 1, that means the initial new moon-S Max combos are in mid to late December and tend to be simultaneous (syzygy and dec-max in both combinations overlap right at the winter solstice nowadays).

    Moving to lunation 2, for data that generally speaking overlaps February with a spread of plus or minus 15 days ... here we see that the new moon event around early February has a decent signature with a minimum of 1011 mbs early in the data set. The next pressure trough is timed for around the N Max event nine days into the lunation. In mid-February, a northern max will occur four or five days before full moon. In eastern North America, these are often quite stormy events, but it appears to me that in the European winter climate, the February full moon runs into the countervailing influence of an annual peak for blocking high pressure. There would be individual cases of deep low pressure couplets of N Max and Full moon, but the average response seems to be a sharp pressure rise after the N Max followed by a weak relapse at full moon. It should be noted that the February full moon will be coincident with the RC event, with the minimum separation occurring with a 21-23 Feb full moon date.

    If you could go into the file and move over to the left through the lunation data to columns Z to AC, you'll find a guide to some colour coding that I added to this portion of the file, showing time intervals for this mid-lunation-two (somewhat similar to mid-February) pressure rise after northern max and before full moon. I show with green and yellow highlights the intervals where pressure in each 20-year interval has risen by 3 mbs, and it tends to be an event over just 1 to 3 days. The last instance failed to rise 3 mbs so there's only a high point shown for that column (other columns without highlights are data averages, you can see the same effect in those). So this is one example of the second-order variability that has a lunar origin, making some impact on annual data (the mid-February pressure rise is evident there also).

    For the rest of the lunation 2 data, pressures tend to be rather uniform from about full moon for about a week until another marked pressure fall near the SpC, A and S Max events (around 19-24 days into the lunation). This seems to be a signal for breakdown of any blocking, as the next lunation shows.

    For lunation 3, corresponding generally to March, the trend is falling pressure, with local minima apparent every seven days as the separation of the major events is now about at its maximum, N Max coming seven days after a March new moon and about the same timing separation before a late March full moon. The general falling trend reaches its ultimate conclusion with a southern max signal that appears to show that with the Moon's tidal forces concentrated south of the equator, the seasonal pressure buildup is reversed temporarily.

    For lunation 4, there is a sharp signal of a pressure wave peaking between new moons (in the time range 15 March to 13 April) and N Max coming about six days later. The April full moon seems to be a time for equal chances of blocking or zonal flow since looking at the 19 sets of data I can see some with sharper pressure signals than others. A weak signal then occurs at the mid to late April southern max event which in the temperature analysis is a peak for warmth (in both Europe and North America). The weather appears more likely to be settled in the interval between a late April southern max and the next new moon.

    In lunation 5, N Max is beginning to close in on new moon (all these lunation graphs are timed to place new moon at about day 2). This set of the data is mostly from May data although we are starting to shift a little ahead of the calendar now so there's more April data and less June data involved in this set. Clearly, the lunar influence on this part of the spring is to promote disturbed or unsettled weather around the new moon to Northern Max interval. It's interesting that folklore says to plant at this time of the month in the spring (when the waxing crescent of the new moon can be seen in the evening skies). What this analysis shows, supported by temp and precip analysis, is that you can expect a better chance of rain at this point with settled weather to follow. The full moon and southern max signals are evident but in a generally higher pressure trend overall. That higher pressure maintains itself towards the end of lunation five before breaking down.

    Lunation 6 would apply to data mostly in June but with a significant amount from late May, and a small amount extending into early July. So this lunation will capture a time where new moon is closely followed by a northern max, and where full moon (mid-lunation) is likely to be almost coincident with southern max. Once again, as in lunation 5, there's a relative minimum of pressure at the new moon-northern max combo except that now they are joined together. Then comes a second wave around the RC to SpC events. The mid-lunation signal for the full moon - S Max combination appears to be delayed more towards the RO timing. In North American analysis, S Max to RO time frame is often a warm spell to heat wave generator. My theory is that the slight southward pull of tidal forces releases desert heat into the plains states around this time with stronger low pressure likely in the lee trough. The European case would be different, more like a slow but steady fall of pressure in the Icelandic low around this time. Something for further study and analysis in any case. These are some of the higher average pressures of the lunation year, in the days around the June full moon.

    Finally, for this session (leaving part two of the year for another day), lunation 7 shows conditions in late June and July. By this point the overflow into August is minimal as we have shifted five days ahead of the pace of the calendar now. This lunation is a bit different, it shows the breakdown of the June high pressure likely coming with the late June to early July new moon (N Max is now moving off the left edge of this graph) and a steady fall in pressure from there to the mid-July S Max - full moon where there's a relative minimum that continues to build up to an ultimate low point around the SpO event. From there to N Max (now 2-3 days before the late July or early August new moon) pressure recovers slightly.

    ______________________________________

    Next session will look in detail at lunations 8 to 12, basically the late summer, autumn into December portion of the year. There's one big feature to discuss in that session.

    I could note in passing that an analysis of average pressure vs the 27.55 day perigee cycle shows mostly a flat signal with a minimum in pressure around lunar apogee. I think the only likely significance of lunar distance in this complex question of pressure variation might be which events are strengthened by lunar perigee. We are currently at a point in a 9 year cycle where perigee falls just after N Max. It moves later into the sidereal cycles every year, reaching S Max around 4 years from now. So 2018 was a year with stronger N Max potential if it is assumed that intensity of events varies inversely with lunar distance from earth.

    Also, I will remind readers that 2019 runs 4 days later than the mean timetable of these graphs, as lunar events in 2019 are on average four days later than in 1851. So if you're looking at any given lunation, add four days to the dates shown to get a more precise bearing on what part of 2019 they might be applied towards. We have just passed that mid-Feb N Max full moon combination. The next new moon will be on the sixth of March. That would place the southern max around the second of March. This coming year, Jupiter is closing in on the S Max position and Saturn is moving away from it but rather slowly; Jupiter will overtake Saturn in 2020 around the position of "S+2d" in this timing (two days after southern max). It should mean that summer of 2019 southern max (with full moon added) events will be multi-energy-centered events with ripples of the additional JC and SC energy embedded, and winter 2019-2020 could be highly energetic with the new moon southern max events of Dec and Jan having very closely timed additional energy from those other events.

    At some later point in the analysis, I will show some evidence for separate pressure signals of JC and SC (also JO and SO) events. The general rule is that they are about 50% as large in terms of signal strength of the major events, both in pressure and temp-precip responses.


  • Registered Users Posts: 14,279 ✭✭✭✭M.T. Cranium


    Continuing on then, and please note, it would be necessary to have followed the posts before this one to understand the references to an excel file available at a link ...

    Lunations 8 to 12 >>

    The data for lunation 8 come mainly from late July and all of August. We are just about at the point where no data is needed from the following calendar month as the central dates are about a week before that month begins (so a small portion comes from early September).

    The graph for lunation 8 appears around AU-BA 184-200. The first third of this period from new moon to about SpC event a week later is generally a time of higher pressures. Marked signals then occur for lower pressure around August southern max and full moon events. From some of my tropical storm research, this is a time in the lunar cycle when extratropical transition is more frequent than average. So some of this low pressure may be contributed by recurving former tropical storms from the central Atlantic. Pressures then rise again with the next significant signal for low pressure coming around N Max which is the trough about 4-5 days before the time sequence ends in this lunation.

    Lunation 9 would refer to September data and a fraction from late August. It would occasionally pick up a few late data points in early October but the sharp pressure fall shown is something that normally happens in the period between the autumn equinox (22 Sept on average) and the first of October. The new and full moon events in this time sequence appear to provide the primary energy for low pressure and the southern and northern maxes are weaker than usual. I think from inspecting some maps that there's a tendency for them to run so far north that they don't impact the Irish Sea pressure field very much, while the new and full moon times seem to collect a lot of the tropical storm remnants and other debris from the central Atlantic that is being pushed east more frequently by the rapidly strengthening monsoonal flow out of the eastern Canadian arctic in early autumn.

    Lunation 10 contains an interesting anomaly that I mentioned earlier. After a time of relatively high pressure lasting into early October, there is a sharp pressure fall timed approximately even with the October full moon time period (this would be a five day separation to N Max on average, and varies from six for earlier dates in the set, to four in the later portion of lunar dates). As you may know, you can identify the location of data in any graph by clicking on the upper right corner of the graph box, and if you do that, you're going to find that the data presented for lunation 10 is taken from rows 1065 to 1182 of column AP which is the mean of all data 1851-2010. So a steady drop in pressures from data points 70 to 85 refers to rows 1114 to 1129.

    So if you could then scroll down to the data in that range, from AD-AG and AJ-AM, as AH and AN are averages of those sets of four (80-year intervals). The sharp drop is evident in each of the eight columns at the same point, with some variation. This sharp drop manifests in the annual pressure data (calendar based) and yet it is apparently mostly generated by one portion of the lunar cycle that could be two weeks before or after that actual calendar timing. This suggests that it is a stronger occurrence at lunar dates near the average of the set. For 1851, the full moon in lunation 10 takes place on October 10th and northern max on the 16th. (1114 to 1129 would be days Oct 6 to 10). Applied to this year the time period is Oct 10 to 14.

    A check of the 19 data sets that are timed across the range of dates in each lunation shows that the strongest signals are in fact near the middle to end of the segment after weaker or contrary indications at first. So there's something happening when a full moon occurs in a time range of about October 8 to 24 to generate a strong low pressure response in the central to eastern Atlantic (this is the time where full moon is overlapping the SpO event). Usually a second similar event will happen at the N Max event about four or five days later. This is probably what most weather observers would take as being the time of "the return of the autumn Atlantic stormy season." So it depends to some extent on lunar dates how strong this onset will be.

    The graph for lunation 11 shows that there is then an even deeper low pressure signal likely to occur after the following new moon in late October or early November, at the southern max event. This would be timed on average 3-4 days after new moon in this lunation. Pressures reach a lower average here than they do in the calendar year (when the apparent mean signal is around 27 October as you may recall). The pressure tendency in November at full moon and N Max is not similar to October but instead stays fairly high with smaller pressure signals. The pressure trough at the end of lunation 11 is the late November or early December SpC event that takes place about three days before new moon.

    Lunation 12 then tracks the pressure variations in late November and December. The new moon range of dates would be on the order of 12 November to 10 December to start out this segment. Southern max would be four to one day(s) later depending on where the new moon fell in that range. So on average this segment has about a 10 data point separation for S Max (around 20 on the scale after the new moon around 10). This is clearly a relatively disturbed period and anecdotally it seems to be a time for stalling lows and deeper troughing than has been seen up to this stage of the late autumn. This makes sense if the tidal forces pulling strongly southward are a contributing factor.

    Pressures then recover somewhat for a week or so before troughing again with the full moons close to early December dates. These will have northern max "partner events" within a day or two on average, some of the later data in lunation 12 approaches the winter solstice where the timing is coincident. (the stub of lunation 13 already discussed has more of those).

    The actual data for lunation 12 seem to show a series of active lows likely at 3-4-day intervals from about the SpO event through full moon, N Max and towards RC before this settles down, but I think that perhaps my timing on these graphs needs a bit of editing work as a check shows that the dates are drifting slightly to the right of where I had intended them to sit (average duration of the lunation segments must be a data point short on average), so the sharp trough is not the RC event as it appears to be, but probably the N Max event. I may try to edit this graph and repost it as a source then edit some of this material tomorrow while I still can do so. But in any case, that does bring us back to the start of lunation 1 (or sometimes 13 into 1).

    Now might be a good time to revisit the graph at top of column BI where the 27 day sidereal cycle is averaged out. This shows a very clear dominance of pressure waves related to southern and northern max, and secondary troughs for the SpO and SpC events. If these are just travelling resonances and not related to Spica as the model postulates, then the timing suggests that they are the arriving pressure waves generated near timing line 1 (not 2, the separation is too long) or possibly even the Rockies lee trof so you could say timing line 9 (which has the same value as zero in terms of westward progress in the grid).

    ====================================

    I will take a break until we've had a chance to digest all of this and perhaps some will have a few questions or comments on this part of the research. Eventually there would be a second excel file showing some other studies on other time scales that represents the other portion of the research, rotating solar system magnetic field effects. It might be a couple of weeks until we get that far.


  • Registered Users Posts: 14,279 ✭✭✭✭M.T. Cranium


    I hope to move on to the second part of this pressure analysis which involves other external sources (other than the Moon). Just wondering if people had any questions about this part, or comments. Seems unlikely to me that it could be random or insignificant that these pressure variations are fairly regular, and affect several different climate zones in similar ways (meaning timing lines).

    The temperature and precip signals that go with these low pressure waves are about what you might expect as shown in the discussion of them over on Net-weather. Some of the full moon - northern max combinations are quite strong in December and January when you break down the data into similar dates (of lunar events). I've found some segments that approach 4 or even 5 deg and 5 to 10 times random precip indicating a frequency of significant low pressure systems. Knowing that about one-third of the potential cases miss any given recording site north or south, this adds to the potential intensity of the events if this depleted sample can produce that strong a signal against what is likely to be opposite tendencies during the missed events.

    I certainly wish this system would work "like clockwork" with very predictable signals that always happen. But I'm sure if that were the case, somebody would have noticed this a long time ago. Perhaps somebody did, a very long time ago.


Advertisement