jinki
Level 3 Rank
Posts: 123
|
Post by jinki on Mar 26, 2010 21:31:34 GMT
1057 has a darkness ratio of 77%. The highest for SC24 by a fair margin.
|
|
|
|
Post by Bob k6tr on Mar 27, 2010 3:50:46 GMT
No assumptions ? Well presumptions then. Let's take a look. arxiv.org/PS_cache/arxiv/pdf/1003/1003.4281v1.pdf On Page 2 Figure 2 You say that two Epochs were selected 1951-1988 and 1996-2009. The latter was selected because it represents the last complete Solar Cycle and the most recent data which is logical. But the selection of the start and stop points for the first epoch is to say least curious. You state the parameters were selected because it was "surprisingly precise and stable". But what immediately jumps out at the casual observer is the overall intensity of Solar Activity during this period. Cycle 19, 1954 to 64, was the highest recorded Sunspot Cycle ever. Cycle 21 the second highest recorded. And Cycle 22, of which you only use the early years, is the 4th most intense ever recorded. Of the 3.5 Cycle that are encompassed in the first epoch only Cycle 20 compares favorably with Cycle 23 in terms of intensity. Comparing the SSN/SFI ratio between those two cycles alone would give more accurate results. There is a typo [repeated] in the paper. The interval [as indicated on the Figure is 1951-1990, not 1951-1988. The 1951 is just the beginning of the Japanese data. The 1990 and 1996 points were selected not based on solar cycles, but simply to provide some separation of the data into two classes. There did not seem to be a sharp discontinuity so we were not sure where in the 1991-1995 interval to place the separation year. Better to have a gap there so we don't blend the two populations. In addition, both stations moved during that period, the Canadian in 1991 and the Japanese one in 1994. You can see from Figure 3 that cycle 19 and cycle 20 behave identically in spite of their very different sizes. OK fair enough. From Figure 3 it is pretty easy to extract the data from Cycle 20 and compare it to Cycle 23. Have you or anyone you know attempted to define an equation for the relationship of SF to SSN since the departure commenced ? That is from the start of Cycle 23 onward ? |
|
|
|
Post by lsvalgaard on Mar 27, 2010 5:37:35 GMT
There is a typo [repeated] in the paper. The interval [as indicated on the Figure is 1951-1990, not 1951-1988. The 1951 is just the beginning of the Japanese data. The 1990 and 1996 points were selected not based on solar cycles, but simply to provide some separation of the data into two classes. There did not seem to be a sharp discontinuity so we were not sure where in the 1991-1995 interval to place the separation year. Better to have a gap there so we don't blend the two populations. In addition, both stations moved during that period, the Canadian in 1991 and the Japanese one in 1994. You can see from Figure 3 that cycle 19 and cycle 20 behave identically in spite of their very different sizes. OK fair enough. From Figure 3 it is pretty easy to extract the data from Cycle 20 and compare it to Cycle 23. Have you or anyone you know attempted to define an equation for the relationship of SF to SSN since the departure commenced ? That is from the start of Cycle 23 onward ? For 1951-1990 SSN = -0.00106 SF^2 + 1.40438 SF - 80.25 ; r^2 = 0.9928 For 1996-2010 SSN = -0.00226 SF^2 + 1.53925 SF - 92.33 ; r^2 = 0.9908 For very low sunspot numbers [ e.g. less than 10] the formulae are less reliable. If you use the 1951-1990 formula and calculate a synthetic SSN [on a yearly basis] for each year 1951-2010 and divide the actual observed international SSN by the calculated SSN you get this:  |
|
|
|
Post by lsvalgaard on Mar 27, 2010 6:48:23 GMT
1057 has a darkness ratio of 77%. The highest for SC24 by a fair margin. Don't forget that L&P measure in the infrared at 1.5 micrometer. Contrast is different in IR. |
|
radun
Level 3 Rank
Posts: 152
|
Post by radun on Mar 27, 2010 15:19:06 GMT
But then infrared would not make SS invisible, unless effect is across the visible range.
|
|
|
|
Post by Bob k6tr on Mar 27, 2010 16:23:11 GMT
OK fair enough. From Figure 3 it is pretty easy to extract the data from Cycle 20 and compare it to Cycle 23. Have you or anyone you know attempted to define an equation for the relationship of SF to SSN since the departure commenced ? That is from the start of Cycle 23 onward ? For 1951-1990 SSN = -0.00106 SF^2 + 1.40438 SF - 80.25 ; r^2 = 0.9928 For 1996-2010 SSN = -0.00226 SF^2 + 1.53925 SF - 92.33 ; r^2 = 0.9908 For very low sunspot numbers [ e.g. less than 10] the formulae are less reliable. If you use the 1951-1990 formula and calculate a synthetic SSN [on a yearly basis] for each year 1951-2010 and divide the actual observed international SSN by the calculated SSN you get this: Interesting What I have been using is : SF = 63.7 +0.73R + 0.0009R 2This formula comes from the New Shortwave Propagation Handbook. (Which isn't so new anymore, It was published in 1995) Here is the paragraph devoted to this subject as it appears in Chapter 2 Solar flux measurements are more consistent, considerably less variable, and more objectively determined than are the counts derived from telescopic viewing of sunspots. There is a very close correlation between solar flux and sunspot numbers. Figure 2.9 shows the progress of Sunspot Cycle 22 between 1986 and 1993 in both sunspots and solar flux units. Figure 2.10 shows the relationship between sunspots and solar flux based on a long period of measurements. Based on these data the relationship between solar flux (SF) and the daily sunspot count (R) is approximately linear :
SF = 73.4 + 0.62R ............................................ (3)
A better approximation to the relationship between these two parameters, and one that is essentially linear, was derived by Stewart and Leftin9;
SF = 63.7 + 0.73R = 0.0009R2 ............................ (4)The second equal sign between 0.73R and 0.0009R 2 appears to be a typo. It looks like someone didn't press the shift key hard enough. Reference 9 is : Stewart, F. G. and M. Lefkin, "Relationship between Ottawa 10.7 cm Solar Radio Noise Flux and Zurich Sunspot Number," Telecommunications Journal, Vol. 39, No III, 1972 BTW I have always thought the way Stewart and Lefkin resolved this equation for SFI instead of Sunspot Number was backwards. |
|
|
|
Post by lsvalgaard on Mar 27, 2010 18:35:21 GMT
For 1951-1990 SSN = -0.00106 SF^2 + 1.40438 SF - 80.25 ; r^2 = 0.9928 For 1996-2010 SSN = -0.00226 SF^2 + 1.53925 SF - 92.33 ; r^2 = 0.9908 For very low sunspot numbers [ e.g. less than 10] the formulae are less reliable. If you use the 1951-1990 formula and calculate a synthetic SSN [on a yearly basis] for each year 1951-2010 and divide the actual observed international SSN by the calculated SSN you get this: Interesting What I have been using is : SF = 63.7 +0.73R + 0.0009R 2This formula comes from the New Shortwave Propagation Handbook. (Which isn't so new anymore, It was published in 1995) Here is the paragraph devoted to this subject as it appears in Chapter 2 Solar flux measurements are more consistent, considerably less variable, and more objectively determined than are the counts derived from telescopic viewing of sunspots. There is a very close correlation between solar flux and sunspot numbers. Figure 2.9 shows the progress of Sunspot Cycle 22 between 1986 and 1993 in both sunspots and solar flux units. Figure 2.10 shows the relationship between sunspots and solar flux based on a long period of measurements. Based on these data the relationship between solar flux (SF) and the daily sunspot count (R) is approximately linear :
SF = 73.4 + 0.62R ............................................ (3)
A better approximation to the relationship between these two parameters, and one that is essentially linear, was derived by Stewart and Leftin9;
SF = 63.7 + 0.73R = 0.0009R2 ............................ (4)The second equal sign between 0.73R and 0.0009R 2 appears to be a typo. It looks like someone didn't press the shift key hard enough. Reference 9 is : Stewart, F. G. and M. Lefkin, "Relationship between Ottawa 10.7 cm Solar Radio Noise Flux and Zurich Sunspot Number," Telecommunications Journal, Vol. 39, No III, 1972 BTW I have always thought the way Stewart and Lefkin resolved this equation for SFI instead of Sunspot Number was backwards. Note that I calculate SSN from SF, not SF from SSN [or R as you call it]. If you want to calculate SF from SSN, my formula is: SF = 59.76 + 0.7892 SSN + 0.00058 SSN^2 But in any case these formulae don't work well for SSN less that 10, because of the way the SSN is defined: 1 spot = SNN 11, 2 spots in one group, SSN = 12, but in two groups SSN = 22, etc A formula that works better for small SSN is SF = -0.0000114 SSN^3 + 0.0038145 SSN^2 + 0.54394 SSN + 63.63 |
|
|
|
Post by Bob k6tr on Mar 28, 2010 6:55:16 GMT
Interesting What I have been using is : SF = 63.7 +0.73R + 0.0009R 2This formula comes from the New Shortwave Propagation Handbook. (Which isn't so new anymore, It was published in 1995) Here is the paragraph devoted to this subject as it appears in Chapter 2 Solar flux measurements are more consistent, considerably less variable, and more objectively determined than are the counts derived from telescopic viewing of sunspots. There is a very close correlation between solar flux and sunspot numbers. Figure 2.9 shows the progress of Sunspot Cycle 22 between 1986 and 1993 in both sunspots and solar flux units. Figure 2.10 shows the relationship between sunspots and solar flux based on a long period of measurements. Based on these data the relationship between solar flux (SF) and the daily sunspot count (R) is approximately linear :
SF = 73.4 + 0.62R ............................................ (3)
A better approximation to the relationship between these two parameters, and one that is essentially linear, was derived by Stewart and Leftin9;
SF = 63.7 + 0.73R = 0.0009R2 ............................ (4)The second equal sign between 0.73R and 0.0009R 2 appears to be a typo. It looks like someone didn't press the shift key hard enough. Reference 9 is : Stewart, F. G. and M. Lefkin, "Relationship between Ottawa 10.7 cm Solar Radio Noise Flux and Zurich Sunspot Number," Telecommunications Journal, Vol. 39, No III, 1972 BTW I have always thought the way Stewart and Lefkin resolved this equation for SFI instead of Sunspot Number was backwards. Note that I calculate SSN from SF, not SF from SSN [or R as you call it]. If you want to calculate SF from SSN, my formula is: SF = 59.76 + 0.7892 SSN + 0.00058 SSN^2 But in any case these formulae don't work well for SSN less that 10, because of the way the SSN is defined: 1 spot = SNN 11, 2 spots in one group, SSN = 12, but in two groups SSN = 22, etc A formula that works better for small SSN is SF = -0.0000114 SSN^3 + 0.0038145 SSN^2 + 0.54394 SSN + 63.63 Thanks for the response Leif. Much appreciated One thing I would like to address is "Note that I calculate SSN from SF, not SF from SSN [or R as you call it]." Actually it is George Jacobs the principle author of the New Shortwave Propagation Handbook that makes the reference. According to SIDC/NOAA R is the Daily Sunspot Number. Also according to SIDC R i is the International Sunspot Number. It is the 13 month running average more commonly referred to as SSN. Both of these designations are recognized by both NOAA ad SIDC. sidc.oma.be/educational/classification.php#Glosswww.swpc.noaa.gov/info/glossary.htmlYou would be surprised at those who mislabel their work. Check out this Solar Terrestrial Dispatch MUF Chart.  STD makes the most powerful Propagation Software Package for personal use. Checkout the parameter bar at the bottom. SSN ? ? ? That is actually Daily Sunspot Number and should be labelled R. Or how about Jan Alvestad's page www.solen.info/solar/Scroll down to Monthly Data and check out the header for the middle column. International sunspot number ? ? ? Those are the Monthly Mean Sunspot Numbers used by SIDC to compute the International Sunspot Number which is R i/SSN. To my knowledge the Monthly Mean Number has never been given a codified designation although many writers informally use R m. Unfortunately David Hathaway likes to use that designation for R Max. While those who are around this subject all the time can divine what a writer is saying from context the casual reader is left baffled and confused. As I see it there are some loose ends that have to be tied up on this matter. Returning to the question at hand the Solar Flux Measurement is a Daily Measurement. If one is defining a conversation equation wouldn't it make sense for the sake of simplicity to have that equation render its product as a Daily Sunspot Number ? Sorry to be a stick-in-the-mud on this matter. |
|
|
|
Post by lsvalgaard on Mar 28, 2010 8:56:07 GMT
Note that I calculate SSN from SF, not SF from SSN [or R as you call it]. Sorry to be a stick-in-the-mud on this matter. I actually prefer and use R myself. As long as there is no confusion, any designation will do. My formula originally read: or 1951-1990 R = -0.00106 SF^2 + 1.40438 SF - 80.25 ; R^2 = 0.9928 Where the two Rs have different meaning. I changed the last R to 'r', but still there could be confusion, so decided to use SSN for clarity. Also because we used SSN in arxiv.org/PS_cache/arxiv/pdf/1003/1003.4281v1.pdfI have no problem with a stick-in-the-mud, as long as one doesn't stumble over it. SSN is also used on solarcycle24.com/ See the little frame with current activity  |
|
|
|
Post by George Kominiak on Mar 28, 2010 14:30:26 GMT
Hey Guys,
Has anyone heard whether Bill Livingston has had any luck getting recent time on the telescope??
G.
|
|
|
|
Post by Bob k6tr on Mar 28, 2010 16:39:20 GMT
Note that I calculate SSN from SF, not SF from SSN [or R as you call it]. Sorry to be a stick-in-the-mud on this matter. I actually prefer and use R myself. As long as there is no confusion, any designation will do. My formula originally read: or 1951-1990 R = -0.00106 SF^2 + 1.40438 SF - 80.25 ; R^2 = 0.9928 Where the two Rs have different meaning. I changed the last R to 'r', but still there could be confusion, so decided to use SSN for clarity. The way SWPC has it defined R - Daily Spot Number R i - International Spot Number = Smooth Spot Number (SSN) It also is defined as the Wolff Number by SIDC (but we have enough confusion as it is  ) R i being a running 13 month average is actually tabulated until 6 months after it happens. Therefore using SSN when one means a projected calculation forward of SSN starts down the slippery slope of ambiguity. SSN is fine when looking at Solar Activity in a historical context but not very good in realtime applications. Leif at the end of the month SIDC applies a correction factor of .5 to .7 to the Monthly Mean Total. Can you refresh my memory and tell me what the purpose of the correction factor is ? Is it to de-weight the 10 X the number of Spot Groups factor in the daily number ? We've said our peace on that matter Leif a long time ago. I will talk to Paul about correcting the SSN designation in the Solar Terrestrial Widget. |
|
|
|
Post by Bob k6tr on Mar 28, 2010 16:46:42 GMT
Hey Guys, Has anyone heard whether Bill Livingston has had any luck getting recent time on the telescope?? G. George, Leif posts in this thread communications that he receives from Bill on this matter as soon he gets them but makes it a point not to bug Bill. |
|
|
|
Post by lsvalgaard on Mar 28, 2010 17:07:56 GMT
Leif at the end of the month SIDC applies a correction factor of .5 to .7 to the Monthly Mean Total. Can you refresh my memory and tell me what the purpose of the correction factor is ? Is it to de-weight the 10 X the number of Spot Groups factor in the daily number ? About Ri, Rz, etc: I'm working on calibration of sunspot numbers. For this I often have to refer to and graph sunspot numbers covering centuries. It is too cumbersome to have to deal with Wolf Numbers, Zurich Numbers, International Numbers, etc. The generic term SSN [and GSN for the Group Sunspot Number] is much to be preferred, where it is understood that it is composed of several parts [Rw,Rz,Ri,Ra American sunspot number, SWPC, etc]. Also many amateur associations [British, German, Japanese, see: www.vds-sonne.de/gem/res/results.html ] make their own. So a generic term SSN is most welcome. Now, about the fudge factor 0.5-0.7: Wolf's successor Wolfer disagreed with Wolf about how to count spots. Wolf advocated only to count 'real' spots, not pores, specks, and smudges. Wolfer pointed out that it was hard to agree on when a spot was a 'real' spot [depending on telescope, person, seeing, etc], so insisted on counting EVERY spot, pore, speck, etc. In this way he, of course, got a higher count than Wolf. Wolfer introduced a factor 0.6 to reduce his counts to be compatible with Wolf's. The problem is that that 0.6 may not be constant, hence the 0.5-0.7 range. It is also applied to the wrong part of the equation: SSN = k (10*G + S). Better would have been SSN = 10*G + k*S. This is too late to change now. |
|
|
|
Post by Bob k6tr on Mar 28, 2010 17:32:19 GMT
Leif at the end of the month SIDC applies a correction factor of .5 to .7 to the Monthly Mean Total. Can you refresh my memory and tell me what the purpose of the correction factor is ? Is it to de-weight the 10 X the number of Spot Groups factor in the daily number ? About Ri, Rz, etc: I'm working on calibration of sunspot numbers. For this I often have to refer to and graph sunspot numbers covering centuries. It is too cumbersome to have to deal with Wolf Numbers, Zurich Numbers, International Numbers, etc. The generic term SSN [and GSN for the Group Sunspot Number] is much to be preferred, where it is understood that it is composed of several parts [Rw,Rz,Ri,Ra American sunspot number, SWPC, etc]. Also many amateur associations [British, German, Japanese, see: www.vds-sonne.de/gem/res/results.html ] make their own. So a generic term SSN is most welcome. Now, about the fudge factor 0.5-0.7: Wolf's successor Wolfer disagreed with Wolf about how to count spots. Wolf advocated only to count 'real' spots, not pores, specks, and smudges. Wolfer pointed out that it was hard to agree on when a spot was a 'real' spot [depending on telescope, person, seeing, etc], so insisted on counting EVERY spot, pore, speck, etc. In this way he, of course, got a higher count than Wolf. Wolfer introduced a factor 0.6 to reduce his counts to be compatible with Wolf's. The problem is that that 0.6 may not be constant, hence the 0.5-0.7 range. It is also applied to the wrong part of the equation: SSN = k (10*G + S). Better would have been SSN = 10*G + k*S. This is too late to change now. Isn't there a form of this equation that does have such a constant to correct for different Observatories ? Hey if we keep this up maybe every individual observer can get his own Personalized Correction Factor Number. Call it the Calibrated Eyeball Number ! ;D |
|
|
|
Post by lsvalgaard on Mar 28, 2010 17:46:14 GMT
About Ri, Rz, etc: I'm working on calibration of sunspot numbers. For this I often have to refer to and graph sunspot numbers covering centuries. It is too cumbersome to have to deal with Wolf Numbers, Zurich Numbers, International Numbers, etc. The generic term SSN [and GSN for the Group Sunspot Number] is much to be preferred, where it is understood that it is composed of several parts [Rw,Rz,Ri,Ra American sunspot number, SWPC, etc]. Also many amateur associations [British, German, Japanese, see: www.vds-sonne.de/gem/res/results.html ] make their own. So a generic term SSN is most welcome. Now, about the fudge factor 0.5-0.7: Wolf's successor Wolfer disagreed with Wolf about how to count spots. Wolf advocated only to count 'real' spots, not pores, specks, and smudges. Wolfer pointed out that it was hard to agree on when a spot was a 'real' spot [depending on telescope, person, seeing, etc], so insisted on counting EVERY spot, pore, speck, etc. In this way he, of course, got a higher count than Wolf. Wolfer introduced a factor 0.6 to reduce his counts to be compatible with Wolf's. The problem is that that 0.6 may not be constant, hence the 0.5-0.7 range. It is also applied to the wrong part of the equation: SSN = k (10*G + S). Better would have been SSN = 10*G + k*S. This is too late to change now. Isn't there a form of this equation that does have such a constant to correct for different Observatories ? Hey if we keep this up maybe every individual observer can get his own Personalized Correction Factor Number. Call it the Calibrated Eyeball Number ! ;D Every observer already has his own factor. The way Wolf, Wolfer, and Zurich in general did the SSN was to define a primary observer, then a secondary, tertiary, etc observer. If the primary observer had observed on a given day, his number was used [exclusively]. If he did not observe, then the secondary observer's count was used [with the factor for him]. If he didn't observe, then the tertiary, etc. Only with the International number, did Brussels begin to average the counts from many observers [about 60]. It is debatable which method is best. |
|
)