Документ взят из кэша поисковой машины. Адрес оригинального документа : http://srd.sinp.msu.ru/nymmik/models/quaghebeur_1.pdf Дата изменения: Fri Jan 26 21:10:41 2007 Дата индексирования: Mon Oct 1 19:12:06 2012 Кодировка:

Note to the B. Quaghebeur refrees report (comments on Memorandum

)

In this note I comment only the scientific problems. The referees remarks as the editorial or technical errors were taken into account, if it was possible, without any comments. It needs do remind, that Memorandum is not official document of ISO (unlike as the Specification).
No 1.1.2 2.2.1 Comment Correction for increasing/decreasing from Sun. How? Just a quadratic law doesn't seem appropriate.. In the purpose, it is written that this is the for scientist and engineers. Why getting negative on the engineering approach? Also, if events can be split up into distinct but overlapping events, how do you approximate the fluences of the individual events? Reply This problem turn out from topic of Specification All the regularities, on which the Specification based, work only in case of "physical" events. For the "engineer" events you can not find any such kind of regularities.... In most of such cases, the part of fluence, what remain in area of second event, is always less than 10 % of the basis region fluence part. In the rare cases, when it is not so, we use the semilogarithmic approximation. Principles of "physical events" were defined and used by Logatshov et al. for the catalogues "Bazilevskaya et al. Basis principle of that definition need the compare the high energy particle increasing with the X-ray burst on the Sun. In any cases, You can observe new increasing the E>30 MeV fluxes after first event. Of cause, sometimes we can to except the new increasing of the low-energy fluxes as caused by interplanetary shock, connected with first SEP event. In most cases, the interplanetary shock, what we observed on the Earth orbit don't consist the particles E>30 (60В100) MeV. So in cases, when we observed a new increasing (second maximum) of high energy proton fluxes, we accept it as the second independent event. I recommend you to analyze the events from the catalogues of the Bazilevskaya, Sladkova et al and compare them with the Feynman catalogue technical events.

2.2.1

2.2.1

What is now the definition used for "physical event ?

2.2.2

In the calculation of mean, only two are used. Why?

Fig.2.5

I think, the asterisks clearly show they do not belong to any curve.

2.2.3

From the data shown in Fig. 2.5.... Sorry, I don't see it!

2.2.3

For proving the dependence of the number of SEP event on the solar activity...........

Numbers 40.1 and 39.4 for the last three or four solar cycles, respectively (the mean value is 39.8). What is the problem? You can write 39.8±0.4. These differences are not essential for further conclusions. There are large errors caused by small statistics of events and the form of event distribution function. There is written "From the data shown in Fig. 2.5 it can be seen, that for 4 SA intervals, the fluences were recorded with probabilities, close to the expected mean value (0.5), for one interval below the mean value, and for the interval 100< <128 such actually observed fluxes were expected with the probability of 0.05." Is it wrong? Excuse. 2.2.3 - here we discuss the problem of SEP flux independence on solar activity. Please define you question more precisely.


2.2.4 2.2.5 2.2.5 2.2.5 2.2.5

The total sunspot number in the SA cycle The first paragraph speaks about maximum flux, the rest of the section about fluence ..at most 300 events and a bit further: 354 events. Is my understanding of `at most' correct? I also don't think it's a good idea to plot probabilities on the logarithmic scale. At last experimental value used has a probability of ca. 0.007, while the plot has a minimum of 0.0001. This is a factor of 70!

.... is the sum of mean monthly sunspot number in the SA cycle The text is corrected. "at most" is replaced by `not more, than 400'. It is a only possibility to true analysis perform. The models deal with the SA duration missions and probabilities for them 0.001 (JPL-91). Please, calculate yourself, how many events we must measure, if we want to determine the fluences for that case for example with the 10 %. statistical errors without any extrapolation. I think, something about 106 (corresponding probability is 10-6)! Therefore, the probability scale must start in this case from 10-6. And for that case, what should be difference in fluences, predicted by different models. And how do you can see it without the logarithmic scale? I don't understand. Explain you remark, please, more detailed. You may ask about that M. Xapsos. In his paper, there is not any number about this. If you analyze the data Fig.1. from his paper, you may propose, that min=3106. The problem of events database is very complicated. The one ­ two errors in the event selection or their parameters really don't change something in the model parameters, but may take up position of hard criticism about both: data base and model. Therefore we are multiple cases recalculated the base. The last version is the part of the Mottl's doctors work, but I have recalculated the parameters of the events and there appear small difference in the some event parameters and some event selection. I need minimum a week to clear all the differences, but I had not this time in last some months. In last year I had more than 30 reports on the different conferences, working groups. It is really very hard working loading, moreover, one my eye finished to see anything. But the event selection problems need the analysis of the INTERNET monitor data, what is very hard problem for eyesight. I hope, I can decide this problem in November. Now it needs to prepare my COSPAR's invited reports for publication. About what limiting values do You ask? About Xapsos? If Xapsos mode values calculated by us for different SA cycles, then why do you ask about Specification? Is it important? There is not any function (2) and therefore not any explanation. For displayed on Fig. distribution were calculated 4000 versions, each content 300 events ­ close to number is measured up to now. Sorry, there is nothing to correct. Eq.s 2.5 and 3 are different expressions of the differential spectrum. In Memorandum the Eq. 2.5 is co-orginated with Eq. 3 in Specification.

2.2.5 Eq. 2.3

Link to cpme (CPME?) data. See remark above. What is the value for parameter min?

2.2.5

.....when selecting the SEP events.... Why don't make clear how and which events are selected?

When looking at the limiting values for the fluence given, all of them are less than the one from the specification. Explanation? Fig. 2.8 Why the numbers 1 and 3. What is 2? , which were calculated ...where we calculated.... Is the 300 SEP sufficient? 2.2.6 In Eq. 2.5, compared to Eq.3 of the standard (or Specification R.N.), somewhere a derivatitive (derivative? R.N.) regarding to E got lost.


Fig. 2.10

Should contain balloons, not ballons.

Fig. 2.11 Fig 2.12

Label on the y-axis: particle cm-2s-1sr-1MeV 1 /nucl) What does this /nucl mean? Caption

-

The values given for the power-law spectra...... Fig. 2.13 after Eq.2.7 Fig.s 2.13 and a cyrillic character 2.14.... Calculated from the integral ..........

Thank you, but Memorandum is not official document and should not never been published. Memorandum serve as subsidiary information to understand the Specification. If You mean that I have time to correct in Memorandum all small mistakes in Figures, You are mistaken..... 1. Please quote the expression right. On the y-axis is another expression. 2. Please look at the spectra, displayed on the Figure (He, O, Fe). 3. Please look at the label on the x-axis. I can't find in my Memorandum such kind of mistakes What spectra (Eq.) you have in mine?????? If the Eq.s 2.7 and 2.8, then this Eq.s are from quoted papers. ???

Fig.2.13 also shows the differential spectrum, recalculated from the integral one, plotted according to the data of neutron monitors. (it is the text of the Memorandum)
Integral spectra, calculated from differential data of the GOES-7.... Initial neutron monitor data are always integral fluxes, initial GOES data ­ differential fluxes, measured by differential channels............ There are 3 kinds of energy spectra: 1. Spectra as approximations of the fluxes, measured in the certain moment of time; 2. Peak fluxes ­ particle fluxes different energy reach the maximum in different time; 3. Fluences ­ the total flux of particles in SEP event. ISO 15390 is the Standard from 1. January 2003. Read, please the Kings papers annotations first sentence: "The probability with witch any given solar proton fluence level will be exceeded during a space mission is computed for missions to be flown during the active phase of the next solar cycle." Before you write the comments, read please the annotations of the papers, analyzed. It is not difficult! Excuse. What is DH? Have you never read: http://www.spenvis.oma.be/spenvis/help/backgroun d/flare.html page 12-13 and figure 6. See also Memorandum Fig. 6.29. Thank you! The data Fig. 2.1 (black dots) were calculated some years ago and reflect some corrections, in following canceled. The Fig. 2.1 was recalculated and respond now to the data 4.1 and 4.2. According to events, determined in catalogues of Bazilevskaya et al. 1986, 1990. The MSU model used from data set another models, the part of events of 30 MeV fluences only, because the distribution function for this part of data set (random) is in agreement with the GOES uncorrected data distribution function. (see the text later). Dear colleague. Read bit later on the same page!!!!!!! It is very difficult to answer on the all such kind of questions.

after Eq.2.7

.....calculated from the differential ones

Corresponding to a certain moment of time. What does this mean?

2.3 3.1

About ISO 15390 ... The King model does not regard active or quite Sun period.........

3.3

The supplemented JPL-91 model (SPENVISESA). No comments. CAN DH SUPPLY SOME???? The ratios in Fig. 2.1 are not reflected in event sizes with fluences in Fig. 4.1 and 4.2. Can it explained? Determined us and calculated - how? If JPL-91, and other models, overestimate the the fluences using the same data set, why not the MSU model ?

4.1

k(F30) from F30106, but the standard uses F30105


The value given in Eq. 4.2........... Here , 183 SEP events mentioned, while on page 15, 300 is mentioned.... Further, mention is made that 101.32........ Fig. seems! to contradict the results given in previous section..... ... uses Fo ....... Different from given in standard....... What is meant `tilt values'? The largest event mentioned is the one on November 12, 1960. But the database only starts in 1974....? ......... What is meant by...... combine theses events with the data set ? `The values differ insignificantly.... And C30 increases by 26%. Fig. 4.4 Fig 4.6 Fig 4.6 Caption : complete data set: Which one is meant now? Some normalization by Wolf numbers? At the ordinate 0.12, the model and experimental data differ by a factor of ca. 2. It looks strange......that the same factor (10) can be applied for both peak fluxes and fluence ..the data set is now 1973.8-2004 and not 19742003 as in text.. If would be nice to see the correlation coefficient. But I don't get the values of the dashed lines given the values in Eq. 4.8a and b.

Fig. 4.3 Fig. 4.4

See some lines later!!!!!!!!! For different SEP model characteristics were used different events sets. 183 were physical events, what were used for development of the regularities, inherent to the SEP events. This is technical error. Must be 100.32.. On the Fig, 4.3 demonstrated functios spectral index is 0.32, Therefore the factor is 2. Once again! Read bit later........beginning of the next page..... = spectral indexes of distribution function (different from Eq. 4.4 ) According data Fig.4.5 the distribution function of the 30 MeV protons used the data set 1954- 2003!!! The data 1954-1974 are used only in case of that distribution function ­ read, please, text of the Memorandum more carefully! It mean, that in case of 30 MeV protons fluence distribution function we add the events of 19-20 cycle to the data set of 21-23 cycle. For the MSU model the value of C30 is not significant. In case of random value of SEP event F30 generation, Monte-Carlo method uses and Fo only.. According to markers explanation on the Figure, it means 1956-2003. It is not normalized..... Full data set 1974-2003 Of cause! Here were used the same methods, as for proton fluences distribution (see section 4.2) ???? Factor 10? About what do you ask? On the Fig is written 1973.8-2004. (not 2004)­ it is the exact interval. And as is written 2003 ­ it means all year 2003......Sorry, to explain that kind of things. C=0.81, but this value in the highest degree is the result of the Gamma determination exactness. The true C>>0.81. The statistics of the events is too small to install any functional dependence in range of F301.0·109. The attempts to find any functional dependence for the range of large fluences were not succeeded. These Eq.s are quite good for model. Read more carefully the section 2.1, please! All the functional dependencies, used in the model (distribution function excluded) are based on the GOES-7 and GOES-8 (uncorrected) data. ??? Thank You. Eq.s 8a in Specification and 4.10 in Memorandum were written irregular. Corrected: =10log(A)-1 The degree of authenticity all this equations are always checked by comparing the model outputs with the experimental data. It is only criterion of the Equations used in the model. All the versions (there was about 50 unpublished versions of the model development) were checked by this method. Correlation coefficients and so on didn't help me in the process of the model development. The correlation coefficient values when there are large errors in the SEP characteristics determination,

Fig. 4.7, 4.8 Fig. 4.9 Fig. 4.11

Fig. 4.11 Fig. 4.14

Again another part of the data set was apparently used. Which GOES satellite data are the data from. I don't understand the y-scale when compared with the scaled, used in fig. 4.13. A=+1 and Eq. 4.10??? I don't even dare to ask for a correlation coefficient.

Fig. 4.15


... the Earth orbit - period missing 4.5.2.2 The values given in Eqs. 4.4 and 4.5, are not a same, as these given in specification. I don't see what Eq. 4.6 has to do with the size event. What is stated in the text is only valid for low energies. I don't understand the 1 curve that is sometimes smaller and sometimes larger than the model curve. 0.01 W>100 while in text W100

Fig.6.6 Fig.6.12 6.13 Fig. 6.16, 6.18 Fig 6.16 Fig. 6.18 Fig. 6.19 6.23

content a very small information. Period? For any period, for any mood, for any scientist..... Eq.4.5 is corrected for You: instead of 8.9 is written 9 Nothing. Eq. 4.6 needed to determine the value of and therefore the particle fluences or peak fluxes for energies, different from 30 MeV. ??? In the calculation that curve we used the C and only, but not . Therefore the curve is correct only for region E30 MeV. 0.001 is right OK! But, I have never seen the experimental data exactly W=100 or 50, or 10. Always they are not exactly equal to that numbers. ( for example 100.001, 99.995 or something else). Legend says ­ active years W<100. Active years are always W>40.

Legend says W<100, text says 40W<100 The same 6 events.... Something is wrong with the formula.....

From caption: "The same as on the Fig. 6.16 6 annual fluence energetic spectra"
The formula is replaced to explanation in text.