I transcribed these from personal reel-to-reel tapes shortly after, with all the inherent difficulties of room acoustics at a distance, etc. Paragraph formatting and punctuation in such instances also become arbitrary. ============================================================================ Roger M. Gallet National Oceanic and Atmospheric Administration, Boulder ======================================================== I'm very glad with the new schedule. That way I will try to make it a very few minutes. I've been ... to say ... First I ... concerned with morphology, prediction, and ... structure, and, in general, by that we mean makeups of observation by radio. Afterall, we have known sporadic E in the "good old days", like David Whitehead tell us, essentially by the radio properties. And, it seems to me, that in some of these meetings the danger is that we are loosing from view radio observation properties. And so, I would like to mention one. But, before that, we have a good summary of ... we had essentially two papers concerned with radio properties of sporadic E. One of a statistical nature, quite comprehensive and by Harnishmacher for work in Europe. This, I think, is essentially giving more precise information on many things which we already more or less know with exact information. We will find most of that information in the Monograph by Ernie Smith and Matsushita, and the several reviews from the preceeding sporadic-E conference. I want to present one thing to you. I was quite impressed by the double-peak in the average diurnal variation. That is one property which seems to be widely ... but indicated by the ionospheric ionograms and the statistics by Harnishmacher. But it was also definitely indicated by several of the papers on properties of propagation. For example, on the paper reported by Wilson, the statistics part of it did indicate that the intense sporadic E also shows this two peaks during the day. I'm not sure if I'd realized that before, if I had realized that it was that result. So, I think we all know, more or less, that statistical property of sporadic E from the seasonal viewpoint. There is a strong statistic temperate latitude, a very strong summer effect. And we know the diurnal variation and ... of the geomagnetic variation. In the past in the radio properties are many, many data concerned with sporadic E. They are concerned with the probability that sporadic E will produce either a blanketing or a maximum useable frequency. So these observating terms should be useable, should be ... it should be possible to use this property for ... predictions or problems of prediction of radio propagation. However, we have from George Haydon of a paper by Margo Leftin, presented by George, it seems that even at present time, which is almost a standard for the people concerned with radio propagation, even within that time, 25 years of ... being in the business, it seems that the use of the statistical properties of sporadic E for practical long-distance radio communication is still very unsatisfactory. If you forget sporadic E completely, then the maximum useable frequencies are, in general, too low. We understand about that. But, more effectively, it seems that it is still unsatisfactory. I would like to pass with just one more remark to suggest that is partly due to maybe bad use of the statistics - of the distribution of the probability of either blanketing or maximum useable frequencies to appear. So, sporadic E, not so perhaps, because I have the feeling, I may be wrong, but I have the the feeling the relationship between the vertical property as observed vertically on an ionogram and the transmission, to a large extent, which can be used either for calculating the blanketing by sporadic E for the transmissions through the F-region or for calculating the maximum useable frequency. I have the feeling that the relationship is not as simple as for the other layers. For the other layers, of course, we can use Martyn's Theorem and the Secant Law. But, for sporadic E it is very doubtful that the Secant Law is as simple as that. It is probably, on the one hand, more complicated from the geometrical viewpoint. It is not just a simple Secant Law, and probably also it is essential that we have passing relevance to the intensity as a function of frequency. So that it means something which has been, perhaps, overlooked in the other session, and which can be discussed still more. It seems to me that it seems to be bad now that there is ... Perhaps because we forget a little bit these radio properties ... We must ... series, of course, very impressed by the big success of the wind shear, wind-shear theory. Which explain, of course, the presence of relatively sharp layer, very sharp layer, of high electron density imbedded in the normal E-region. A very long time ago, already, was observed, first I believe by Jackson and Seddon, and then was summarized and discussed by ..., and then by many other, that these peak of ionization, as it is observed by rocket, correspond, more or less, to fbEs, more to the blanketing properties of the E-region than to the maximum useable frequency, either foEs, or others, been discussed also by Reddy and some other. But, if we have such a large gap of ..., relatively large gap, between these two regions ... But, first, let me say that is understandable. It follows if it is a blanketing property related to the maximum electron density. It is evident that a frequency larger than fbEs will go through the layer without too much problem. However, we could continue to have reflection under these frequencies, but also we have semi-transparency. Now, following the semi-transparency, that's not possible with a sharp layer. When we are very impressed by this wind-shear theory, but even so, in the wind-shear theory or in the observation, in fact, the layers are hardly thin enough to permit a wide range in frequency for semi-transparency or a progressive diminution of the reflection coefficient. At the same time, on the one hand, we have semi-transparency on the F-region. On the other hand, we have, also, a diminishing coefficient of reflection. These kind of measurements, which were made by ..., and then quite beautiful. I had one of the treatise by Chatterjee, a long time ago, already. This kind of record, where you have, I don't know if you can see, as a function of frequency, the reflection coefficient and the transparency coefficient, and, in some cases, the sum of the two, for accounting for the total power, show, that in this region, there is some kind of gap. So, if what the means, perhaps, we are forgetting this kind of properties. We seem to go one way or another. But it seems that you cannot obtain properties like that, which are observed, and which should in practice be ... You cannot obtain such properties just on one thin layer extending everywhere. And it is never thin enough. And, if you want that, you tend to semi-transparency by a thin layer, even from theory, it has to be impossibly thin. It has to be much less than 100 meters. In fact, a few tens of meters is already enough. So, what I believe. I believe that certainly in this region the maximum electron density, because the thin layers are never thin enough. And, certainly maximum electron density is not that uniform, even over several 100 kilometers. Probably, and we have it observed by David Whitehead, we need to have a great "patchiness" of that maximum electron density. But if you have patchiness, then you will have some kind of scattering properties. And, it seems, apart from this paper by Whitehead, which mentions the point, and only mentions, really not discuss, not with any new data. Apart from the paper, we seem to be losing the point that some of the property, not all, but some of the properties of sporadic E requires scattering. So from the viewpoint of these semi-transparency, but also, sometimes, it is very obvious in some of the kind of ionograms. So, if we ever have another kind of sporadic E seminar, maybe we could devote one session for a sort of better, more-balanced, systematic summary of the properties. And since we have learned about sporadic E not just from rocket, not just by ... they were not discovered by a rocket. I suggest that maybe we could have a small session during which we have a very fast, but perhaps very-systematic, and very-fair summary of the properties. One thing, and, perhaps, I will welcome that with Ernie Smith, is the conversation. Someone mentioned, I believe, "the good old days", that was you, that in the good old days we have learned some of the properties not just from ionograms but also from these wonderful movies. Now, I do not recall, I have seen many of these movies, ionospheric movies, but, I do not recall to have seen one for sporadic E. However, I recall, I remember very well, to have seen this wonderful case where you have a sort of stratification starting in the F-region and going down to the E-region. I think it will be wonderful to see things like that, and, perhaps, that will avoid that we make big mistake in theoretical work. I have the feeling that theoretical work right now is quite biased by only one way or another. I, perhaps, should know that. I have been quite interested on the scattering aspect of sporadic E. Really means that some observed properties were due to no new source of ionization, no intensification of electron density, and I believe I was wrong at that time. But at the same time, I was emphasizing the effect of if you have turbulence within a layer of ionization, then you will have density fluctuation of the electron density, and that, in turn, will permit to have fairly intense scattering properties. These scattering properties will be amplified by, in a region where the refractive index is less than 1, especially close to foE, or at oblique frequency close to the maximum useable frequency, by the E-region. So, that when you are close to that, the scattering is enhanced by the fact that ... is smaller than 1, or close to zero. So, if anything, the summary of my few remarks is that, on the one hand, I feel we are losing from view some of the supposed-to-be well-known radio properties of sporadic E, and as a risk, we run the risk to have a too-simple and too-biased view for trying to explain all the properties of sporadic E. And, further, I believe, that reflects on the poor use that we make of the presently avilable information for practical application in telecommunication. (applause) K. Tao: Thank you for your cooperation in keeping in time. Do you have any questions and comments? Q 1: Well, just for the sake of precision of description. I'd like to ask for a more careful definition of what is intended by the term "scattering". Reflection from a metal sheet at these wavelengths is coherent, in the sense that few amplitude-phase fluctuations are reproduced by such a thing. Scattering from a region which contains irregularities smaller, or of the order of a wavelength, introduces a lack of coherence. And, we do equate the term "scattering" in the section that it ... with a serious lack of coherence in the received echo. Is that what you have in mind? RMG: Yes. Yes, but also, also I think that the semi-transparency, the very, very large range - that is impossible by just a thin-layer theory and semi-transparent. Q 1: Yes, but it's very possible, as you can see on any cloudy day, for there to be partial transmission of the sunlight and partial obscuration of the sky by clouds. That would be a very plausible model to agree - RMG: Yes. Q 1: with the so-called partial reflective properties of sporadic E. As to return to my question of scattering, I'd like to know whether you would want to include the total reflection from isolated, dense clouds - large, isolated, dense clouds - as a kind of scattering. There there is a partial loss of coherence, depending how many clouds are in view at one time, but the individual reflections are very much - RMG: Personally, in the past I have not believed, or I have not taken into account, huge fluctuations, very large fluctuations of the electron density. Now, of course, I could see very large fluctuations of the electron density, such as 1:2 or 1:3. But, I think it one essential ingredient, the variations we do not have one big, one value of the maximum electron density and uniform everywhere. Even on a relatively small scale, the scale of less than the result in the E-region. We do have large fluctuations of that density, and then we can have ... certain regions total reflection but of some observed scattering. Q 1: I think it needs to be said that it's for questions of this kind where the radio experiments can really bring something of value. RMG: Yes, correct phase measure. Q 1: Well, any one of a number of different approaches can shed light on the nature of the irregularities, whether they're small or large. Whereas, this becomes a small a ridiculous sampling problem with rockets. So - RMG: I was concerned with keeping the time, I will admit, and not making another paper. (laughter) Just making a plea for balance, that's all. Tao: One more question. Q 2: Yes, this is really my own poor recollection, but the question of whether this non-blanketing is blobs, intense blobs as suggested by Whitehead. There is some evidence for that. In the sense that, do you remember that ... in his papers a long time ago, I'm sorry, I don't recall the details, that showed phase-plot records where you could get the regular E-region return, which was changing relatively slowly at first, and then superposed on that, underneath it as far as the group path was concerned, a blob, which apparently approaches, and then receded. Now, this, however, I don't know whether there was ever any effort to correlate that appearance of that blob with non-blanketing sporadic E. It certainly doesn't block very much, superficially, very much like conventional types of sporadic E because it was usually below the regular E-layer. So, I'm just wondering if you, or Dr. Whitehead, would shed some light on that. RMG: Yes. In fact, I'm glad you bring that, because I have really in mind all sorts of experiments that have taken place with the results ... and Chatterjee ... It seems to me that it was done in something like the 1950's to 1955, or so. Q 2: ... experiments were quite different. He was concerned with how to measure the collision frequency in the E-layer by the absorption. RMG: I was not talking of that - Q 2: OK, Yes - RMG: But of Chatterjee, I recall was partial reflection coefficient and the semi-transparency coefficient as a function of the frequency. So there's where one discussion perhaps of the ... but, in details measured, I think the principle is very good. I have the impression that we are losing that view, and certainly I have the ... but, I forget. I cannot agree more with you. I think that we do not have enough, but we have lost this kind of experiment in the last ten years. ============================================================================ J. David Whitehead University of Queensland, Brisbane ======================================================== Now I'm looking to Seminar Session II. This session, which had a number of papers comparing the electron density profile with the winds using the wind-shear theory, ... checks out the ... I think the most noticeable thing is the amount of disagreement in the various ... First of all, Dr. ... uses values a of 10-8, ... found values of about 10-7. I think another value was 6 × 10-9. Is that correct? So that there is a very wide range of values therefore, from 10-7 to 6 × 10-9. A great range of values appear to be required in order to explain the profiles. Now, if this ..., in one sense, is very unsatisfactory, 'cause we're dealing with the same ionosphere every time. In another sense, though, this would indicate change in coefficient, and that may be just one of the points that I was trying to make ... in the ... talk. And, it seemed to me to be one of the major things about the theory that we ought to be better understand the origin, and production, and loss of the metallic ions. Maybe that if you required it for recombination coefficients - you could put down the proportion of metallic ions that happen to exist at that particular time. So that I'm not trying to make any case at all for these effects. It may be the ... time ... very available ... for metallic ions. Another area of disagreement shows very clear, that ..., which lead to values of the vertical velocities, need to be a factor of ten greater than the ... For example, in Kato's example ... And yet, the horizontal winds are pretty much the same as far as the collision frequency observation are concerned. I would still like to know what the answer is ... There seems to be something that could well be sorted out. Now, from MacLeod's paper, he was dealing with the 3-dimensional ... I think that this is very important, results of the work of Harris. It's becoming very obvious that we do need to have a 3-dimensional wind, and, preferably, of course, we need to have a direct measurement of the vertical wind. When you tend to calculate the vertical wind from the horizontal wind, and you make allowances, I believe assumed to be the ... to well with the electron density ... Therefore, it seems to be quite essential that we should have direct measurements of the vertical wind. However, as Dr. Smith and Carl Rosenberg discussed, it is a very difficult thing to do that. It does lead to complications in the way in which the trail is laid out. Complications in that the trail experiments observed from a relatively cheap and easy way ... one wishes for ... a bit more complicated. But, nevertheless, it is such. It may not be worthwhile for a 3-dimensional region, it must be vertical ... So far as the structure's concerned, we have two papers, I believe ... paper by two - . And, in the first one we discussed the apparent tilts that are observed by a rocket. Remember that this is only a sampling at different places. And, so, for example, it would be quite easy to get a situation where you have your sporadic-E layer like that, and your rocket going through the ... same and much the same heights. And, therefore, having a very-small average tilt. And, yet another measurement, which is a large measurement, shows a large tilt. However, it was said that that was essential at the last sporadic-E seminar ... I think we have the radio measurements too short, and even ... Where it's not very tilted. But, I'm getting out of the partial reflection, so if you will ... This triangular shape ... In fact, we need a bit more theory to work out exactly what happens ... how this type of profile gradient converges in this incidence ... situation. You've got to bear in mind, I think, that the major ... are equals who see sporadic E ... or don't want to use sporadic E as a ... And we should, therefore, do the proper calculations. It's well within our competence to perform the calculations. We know what are the theoretical properties of ... In their ... paper four, Smith talked a great deal about a layer which he identifies as being the E2 layer. And, I guess I didn't ... He showed that a wave descending from the upper E-region down to the body of the E-region. And showed, incidentally, that the electron density increased as the layer descended, increased in the layer, thinner and thinner, as it descended. But this seemed to be at variance with the results which Matsushita talked about, this business that the fbEs decreased as the layer descends, and whether this ... somehow that part of the statistics, or whether this was simply one observation. This ... whereas, maybe we wont worry about it, or there is an inconsistency there, between those two results. There is something else which is worth just mentioning here, which was mentioned in the Session, and that is the electron density at the minimum in each case in the E-region at night. This minimum that occurs at about 120 km about midnight can be used to get a value for the production rate at 120 km. The reason is that we know, roughly, the sort of winds that exist at this height so, and so we know that at, say 120 km, what's happening is that whatever is being produced there is being diverged and pushed apart by the winds which produces them. We know the rate of divergence, and we know, also, the electron density there, and it's quite easy, then, to get the rate of production. Now, you don't have to remember all this because it's in the paper by ... Carl Rosenberg presented some new data, we ... on the shears that exist in the E-region winds. And these figures are extremely valuable. He gave us maximum shears and also the distribution of some information about the distribution of shears that existed. And these figures, of course, fall within the theory. There's one thing which doesn't yet appear to have come out in the wind measurements. It hasn't, in fact, been mentioned at all ... can forget may be terribly important in what is causing sporadic E, and that is the question of preferred heights of sporadic E ... positions, another subject that needs to be ... about. However, if you look at the distribution of values which Reddy listed in his paper, he didn't actually reproduce this distribution, if you look at the heights measured by rockets, you find that this is a very "peaky" distribution: peaks at certain heights separated, if I remember correctly, by about 7 kilometer. And there were, it seems to me, sufficient measurements to show that it appears preferred heights of sporadic E. Now, if this is so, it means that you almost certainly got to have a standing gravity wave producing sporadic E. It's very obvious that rather than this ... you could always show in the tidal effects, it could be that tidal winds bring the sporadic-E layers down to the 120-km region. I think that may be the most ... thing for preferred heights, the heights presumed to be determined by standing wave ... pattern ... but this ... additional work ... determine look at ... We can see that there are a number of trails ... there seems to be no evidence from the trail measurements that there is such a structure. Nevertheless, it would perhaps be a good idea to look at the heights of occurrence of sporadic E once again. Ionograms are no good, except ... measure of 5 kilometers in the ... of sporadic E measured using the ionosonde height for plot purposes, and ... shape. We also have ... errors of about also of about 5 km. So that the difference in height between the ionogram height and the rocket height is something like ... km ... At last, the paper by Cloutier. He used a result which still puzzles me a great deal. A very intense current, current density which was far exceeded current density that you usually get from simple picture. And, also, the current flux didn't seem to be consistent with the sporadic E. With sporadic E production, if you're going to be pushing your ionization together like this, and this is the J x B force, essentially. So what you've got to have is the current flowing like that. You've got to have a shear in there. This shear would normally be produced, of course, by the wind shear which drives these currents. And, I could make these currents, this one in this and that direction. I could get a measurement that was described here. ... a current flowing all in one direction. Now that evolved the wrong picture here ... It did seem to be inconsistent with the theory, and, therefore, very insufficient. Perhaps I might have further comments from the others? (applause) Q 1: Yes, I think that's not totally fair to Cloutier. I think that current he is interpreting that current as being simply a focusing of the, I guess it's the Sq current, for it. Now that would be additional to this current system you put up there, I believe - JDW: All Sq currents will have the same effect. You should really take the total current in which you expect to have a shear to produce a sporadic-E layer. It doesn't matter whether you include Sq currents or not. Sq currents simply can assist or detract from the proper currents that are produced by a warp of the wind system. But it's the total currents, the total J x B force that is acting on the ... Q 1: How large are those currents that we are talking about, relative to the Sq current? JDW: Well, that's a difficult question. Q 2: I suspect it's more. JDW: It could be, if that the metallic ... ions keep. ... couldn't be at all sure that they could be swarming. Actually, we're quite sure of the same result. If they were molecular ions you'd require a substantially large current. Or, perhaps, he was looking at this from a current system, in which case, of course, he still would be finishing up with a quite unexpected current distribution, which is much too narrow and much too intense. Q 2: There were observations in this data which - JDW: There was only one major observation. Q 2: There was a displacement of two ... by between there small oscillations. JDW: Yes. ...... Q 2: The question is, do you attribute ... are these going to reproduce any inconsistency ... The answer is one ... Q 1: Yes, I think he was probably ... asserting that if there is enough sporadic E, that most of the integrated conductivity was in the sporadic-E layer and not outside of it, which, the point is perfectly valid, of course, that it would get most of the current flowing in there. JDW: But, incidentally, if that is so, you get quite an interesting situation where the local electric field is therefore determined, to a very large extent, by the local wind flowing and the sporadic E height. And this can lead to a sort of feed-back process, almost, so that electric field tends to preserve the sporadic E at that height. Now, no one has looked at this anymore since Lynch did in the ... one of the first papers on the wind-shear in the JDP as a means of preserving the sporadic-E layer after the wind has, more or less, switched off. It may be worth looking at that ... again. Tao: Thank you. ============================================================================ Tao: Our next speaker, Dr. Layzer. First, paper: "Beyond the Wind-Shear Theory", and afterwards he will give a review of Session III. David Layzer Harvard College Observatory, Cambridge ======================================================== ... by the experimental people at this conference. And there was one ... which received a great deal of attention on the theoretical side. Let me to begin with, just show a summary of some of the properties of sporadic-E layers that I'm talking about, and this will ... the kind ... of sporadic E that especially concerns us. These properties, obviously, are not common to all sorts of sporadic E. But, mainly, I'm concerned with the narrow layers which show a typically a kind of ion profiles that Les Smith has shown you and left from the earliest rocket measurements. Seddon, to the later work of Smith, have been associated with intense sporadic E or blanketing sporadic E. And, particularly, let me remind you that we're dealing with phenomena that has two distinct vertical scales. In the first place, in overall thickness ... predominately below, let's say 110 km, of 1 to 3 km; but also, a second scale, very important, which ..., unless calculations are not sufficiently kept in mind, a scale defined by the gradient of electron density at one of the two edges. These are not typical, in that, normally, the sharp region does not appear at both edges of such a layer, but only at one. But, in some cases, it does appear at both. And this sharp region, amounting to of the the order of sometimes 105 electrons/cc in the space of a hundred meters, may well be responsible for a major part of the phenomenon. Again, we have really a hierarchy of horizontal scales, and the statements greatly simplify the observations. But, we do know that the largest scales are on the order of at least 200 km, and there are small scales on the order of tens of km. Again, the long-lived layers of that I am concerned with are those that have been shown to now to consist mainly of long-lived metal ions; and here we have some of the correlations which make the phenomenon really fascinating. And, which, so far, in theory, have very little attention paid to them. First, the positive correlation of the horizontal component of B, which is what lead Whitehead originally to propose the wind-shear hypothesis. This had been found a little earlier by E. K. Smith, and is described in his thesis. But, the connection and the geomagnetic significance of this was clearly realized by David Whitehead in 1959 or '60. These three correlations have been very much discussed at this meeting, and I feel that mention ... Now, the simplest assumption that one can start with, I think, that the best assumption is something which I think the wind-shear theory ... That is somewhat more general, mainly that these layers are formed through the vertical redistribution of neutral ionization, and this process is governed by two equations, of which the first takes an extremely simple form for metal ions, because during most of the process diffusion is not important, production is not important, and, in the first approximation, the loss processes are not important. And this is a rather trivial equation to solve in the sense you've written down B for the terms of the characteristic. The second equation was derived in 1953, I believe, for the first time by D. F. Martyn, and connects the vertical component of the ionization drift, W, I call it, with the two horizontal components of the wind velocity. Actually, there is a third term which takes into account the third component, which is verified by approximation. Considering the plane parallel problem, I'll ignore that: just consider the horizontal winds. And these two coefficients have expressions whose evaluation depends on assumptions that one makes about the way currents flow in the ionosphere. Expressions have been written down as contents for that. And Wp is the initial contribution that has also been discussed resulting from the electrical polarization field. The theory for this, as I say, was worked out completely by D. F. Martyn in 1953. This contains all special effects which occur about corkscrews and so on. The ... brief point that David Whitehead called attention to was that in the region below, I think, 115 or so km in the hodographs, the second term dominates. The coefficient of that term is proportional to the horizontal component of the magnetic field, x is the northward direction, B is the magnetic field. And this suggested, partly, perhaps, suggested to him that one could account for the major part of the phenomenon of sporadic E by using this equation and taking into account the fact the wide east-west component of the neutral wind varies with height. And that conjecture has turned out to be very well-founded. Now, the wind-shear theory then, as such, is based on the assumption, that is normally fulfilled, that the coefficients in that expression of W vary slowly, that Wp varies very slowly, which is always true. General arguments for that computed by Baker and Martyn back in 1953. And, therefore, we have this sequence of deductions from wind shear to enhanced or reduced electron density, as the case may be. Now, the results. The first place I think we now have certain convergence of opinion on what the results are to date. But there's been poor general agreement, especially at the large f-scales appropriate to the observed wind patterns between the minima and maxima. And, these scales suggest we are within a few km. And, also, that a favorable wind shear, that is to say, a wind shear that leads to a convergent ion drift motion, is certainly a necessary condition for the occurrence of strong layers. And, as Bill Wright has pointed out, where this condition is not fulfilled, layers are found to be growing weaker. So that they are transient. What has not been predicted as now what seems to be a consensus of disagreement, at least between Whitehead and myself, I'm not sure about others, that the calculations based on observed wind profiles at fail to agree with the "fine structure" shown by these profiles at these heights. And here we're talking about the very fine kinks that I have. This is the exercise I want to confine my attention to. That there's a great difficulty in understanding sporadic character of the phenomenon. And, finally, as the papers have shown in this conference, that there is no clear way of understanding the correlations, either seasonal, diurnal, or with magnetic activity, particularly the seasonal and diurnal correlations which are such a striking part of the phenomenon. That is to say, the corresponding periodicities doesn't seem to be present with the required amplitudes in the winds. Well. We seem to be getting further with ... Let's look at the wind profiles and see if there isn't any feature in them that seems to have the required scale as the shears will have to be known. Now, the first slide is the one that Dr. Constantinides showed earlier. ... this shows some recent. In particular, I want to call attention to features which are shown, as I've already said in this conference before, on high-resolution analyses of winds, that I call "corners". Just picking one, this at random, these heights are 99.6 km, 99.7, and 100.8. And notice the change - these are hodographs, the horizontal wind is obtained by connecting the origin of the picture, which is the cross here, to the point in question. So this wind vector is the horizontal wind projected onto, let's say, the ground. And we see that as we mentioned before, the arrows connecting successive points are the wind shears. So that what happens at such corners, the wind shear changes, both in direction and magnitude, because between equidistant points in height occur at different layers. These aren't equidistant, actually, but in general that is what happens. These features as winds shears do not produce any significant effect in the profile. That is, to say, you could just as well smooth this out so far as the wind-shear calculations are concerned. Ok, now, the question arises what happens physically at a corner, and this is part of my talk where I have something to say that hasn't appeared in print. I've discussed one possibility, of which I now no longer exert to be the most probable one, and that possibility that hadn't been excluded on physical grounds, likely describes the analog of a shock wave. It's a surface discontinuity through which there exists matter flow. There is one other possibility consistent with the laws of hydrodynamics, namely, that discontinuity in which there is no mass flow. Z is the direction through the surface of a discontinuity in which there is no pressure change. But, remember, the vertical gradient of the horizontal loss of ... is produced there, so those are discontinuous. What else is discontinuous? Well, the density and temperature, in general, are also discontinuous. And this kind of surface is known as a contact discontinuity - it's a particular kind of contact discontinuity, and it is also known that this particular kind of contact discontinuity is hydrodynamically stable. The necessary condition is that Dr be negative, that is to say, that you have a light fluid over a heavy one rather than the other way around. So, we, unfortunately, don't yet have any measurements of the quantities D of the density and temperature. So far as I know, they're insufficiently fine scale, but, perhaps, there may be measurements that I'm not aware. They are very much to be desired. There are some indications of discontinuities in the density that fall in the sphere of measurements, but I believe that so far are the required accuracy should be, is essential to the conclusion concerned. All right. Suppose that this what happens at corners, and from measurements it appears that if indeed it does, then we have the possibility that ... source to the variation of W, which had entered the equation of continuity, can give rise to convergence or divergence of the ion flow. Because the coefficients in the expression now can vary. And, this is a sort of complementary situation for one that Whitehead considers in which all variations due to Vx and Vy. Under favorable conditions, one can get trapping in a way that is illustrated here. Let's say, as we look at the bottom figure, with the line W = 0, W is the drift velocity, passes through the little meander, or discontinuity, if you like, quasi-discontinuity, in the W profile, in this particular way. Then in there we build up and, in a good interval of time, in this way. With this characteristic scale equal to the characteristic distance, Dz, with this small feature, which, as I should have mentioned, is observed, it were 100 m or so. In some cases ... somewhat less. On the other hand, if the line W = 0 falls outside, then very little effect will be produced, as shown here. So, let me call this situation "trapping", that is where the line W = 0 falls within the profile. And, if you now ask what are the conditions that must be fulfilled in order for trapping to occur, these are three: In the first place, the term Wp has to be sufficiently small not to swamp the effect. You might say, well, by coincidence you could have the right value just to bring the line W = 0 at the proper place. The difficulty with that, that may be rectangular, I suspect that wont work, because Wp through Ep varies in an irregular way, perhaps not over the period required to build up such a layer. So that, in fact, the actual polarization field needs to be either small or steady. ... Secondly, the wind shear, of course, has to be favorable. That is to say, you have some ions there to accumulate. I should emphasize, by the way, that this trap we caused is intrinsically weak because the converging power of this kind of discontinuity here is probably rather smaller than that of a favorable wind shear of the normal magnitude. We need long times to operate in order to be effective, several hours. And this, of course, requires metal ions. The third point, which also is ... if this is in fact small, which may not be the case, if this is in fact small, then for the line W = 0 to pass through this kink, we need to have a node of the favorable wind shear - this is to say, if it is only matters, as in Whitehead's original case, when we actually need to have a layer forming near a node of the profile. So this, now, in this theory, turns out to be a necessary condition. I've noticed that it's not a necessary condition, as I've remarked here, that the wind-shear theory, may not, and in this theory as well, find remarks about the polarization field that are not right. But, still, a very special value of this is required. Also, the wind-shear theory, there is a concept the whole effect is due to wind shears doesn't restrict the polarization field. So this is a way in which the sporadic character of the phenomenon now can predict. We have found some further necessary conditions, and the obvious thing to do is look at those conditions and see whether those conditions which need to be satisfied in order that strong layers can be formed are related in some way to the diurnal, and seasonal, and magnetic correlations, if we'd like to understand. And, I suggest that ... the key, in effect, lies in the fact the magnitude of the polarization field with produces this additional term. One can crudely have what I've represented, here is the magnetic geometry associated with the problem. I've tried to represent the two kinds of regions - those that are connected with the opposite hemisphere by a magnetic line of force ... and with those whose obvious magnetic lines go out into the interplanetary medium. And it is a remarkable feature of mid-latitude sporadic E that the latitude at which that separation occurs is precisely the latitude at which all the statistical properties of sporadic E change discontinuously. This remarkable correlation, which was brought to my attention by Dr. Smith, Ernie Smith, and which I think is extremely important. 1971 USU Es Conference Review Sessions
(TAPE EXPIRED)