The AAVSO Solar Section has been actively improving its data collection and analysis methods over the past year. We've gotten critically important feedback from the global solar research community on the importance of amateur observations of sunspots, and the AAVSO Solar Section is building new and stronger relationships with the professional community. We've never been in a better position to support both the solar observers and the researchers that critically depend upon your data, and we're looking to make even more improvements in the near future.
Observer k-coefficients recalculated
As of February 2013, we are pleased to report that we have computed new solar observer "k"-coefficients for 60 AAVSO active solar observers, including 24 new observers previously without k-coefficients. This increases the reliability and historical consistency of the American Relative Sunspot Number, and is a key part of the maintenance that the AAVSO Solar program does for this important measure of solar activity.
What are the k-coefficients?
The k-coefficient or "k-factor" is a scaling factor used to assess how different observers count sunspots relative to the group standard. An ensemble of observers will each typically observe a different number of sunspots on the Sun, due to differences in observing equipment, site parameters, methodology, and perception. Although each observer is different, the individual observer offsets and hence the overall statistical properties of the ensemble remain very consistent, and it is possible to quantitatively parameterize these observer differences with individual constant offsets, k.
Why do we use them?
Use of the k-factor is to some extent based solely in the history of sunspot counting, but for continuity of the historical record, it remains in use today. The principle of using site- or observer-coefficients began with Rudolf Wolf in the mid-19th Century, when he noticed that his sunspot counts would consistently differ when he observed with an observatory-quality 80mm solar telescope, and a smaller 37mm travelling scope. It was continued with the Sunspot Number, Ra, and both the equation for Ra and the method for computing k-factors were codified very early on in the American program. The AAVSO modified this procedure slightly in 1951, with the American Relative Sunspot Number, Ra', generated using a weighted average of observers where observers with k-factors had heavier weights than those without.
Why do we recalculate observers' k-coefficients?
Observers' equipment, conditions, and ocular perception can change over time, and it's important to periodically reassess observer statistics on a regular basis. However, this had not been done for several years, partly because sunspot activity had been low for that time, and partly because of changes in the AAVSO Sunspot program and its leadership. Because of this, a number of observers had been regular contributors to the sunspot program for years, but were still being weighted lower than other observers with known k-coefficients.
At the end of 2012 we were in an excellent position to re-compute k-factors for the active solar observers, and also to compute k-factors for two dozen "new" regular observers -- those active for less than five years. We had two solid years of prolific observing by several dozen people, and more crucially, we also centralized data collection for the AAVSO solar program. As of mid-2012, all of the observers' sunspot observations are now located in an easily accessible MySQL database, in a format that is easy to access and analyze. Finally, in a moment of good fortune, Bradley Schaefer of Louisiana State University was visiting AAVSO and the Harvard Plate Stacks in late January 2013, and he provided us with both motivation and guidance to move forward with a re-computation of k-factors. Schaefer wrote a critically important paper in 1997 that explained an important defect in the calculation of the American Relative Sunspot Number, along with a proposal to revise the counts. His input at the end greatly assisted us in assessing the final calculations.
In January and early February, AAVSO staff (Sara Beck and Matthew Templeton) and the Solar Section leaders (Rodney Howe, chair, and Kim Hay, sunspot leader) collaborated to create the tools that let us recalculate k for our entire ensemble of observers, and will make it easier to do so on a regular basis in the future. As of the January 2013 Solar Bulletin, we are now using the updated k-factors.
What is left to do?
As mentioned above, Schaefer noted a major problem in the calculation of Ra', namely an artificial inflation in sunspot counts caused solely by the method of creating the k-factors. He suggested an alternative count for sunspots that better mirrors the method by which the k-factors themselves are created. The AAVSO has decided to adopt his revised methodology, and we are in the process of revising the way we calculate -- and publish -- Ra' each month. This must be carefully done for two reasons. First, we need to make clear to the solar research community that this change is being made, and that the new Ra' will be slightly different statistically than before. Second, we need to revise prior data to place it on the same system as the revised count. This is critically important because the long-term record of sunspot counts is what enables the solar research community to assess changes in solar activity. We have to ensure that there are no hidden biases in the historical record, and we'll be proceeding very carefully with revisions. In the meantime we will continue using the current methodology so that the historical record remains on the same system throughout.
What does this mean for you?
All active solar observers now have k-factors, meaning that their observations are all strongly weighted in the calculation of Ra. At this time, only a small number of observers are without k-factors, but given that we now plan to resume regular yearly updates of observer coefficients, those observers will likely get them in January 2014 if they keep observing.
For those of you who are not yet regular solar observers, we need your help! Visual observation of sunspots is still needed to provide the longest and most continuous historical record. Solar activity is known to change on nearly all timescales, from days to centuries, and visual observations are still the longest observational record for the Sun. Other measures of solar activity exist -- including the measurement at 10-centimeter radio waves -- but sunspot counts go back centuries further than any instrumental record. The recent weak solar cycle indicates that our understanding of the Sun is still far from complete. Your observations can help carry forward this historical record, improving our knowledge of historical solar activity and helping researchers correlate the sunspot number with other quantitative measurements.
For more information
- Dances With Wolf's: A Short History of Sunspot Indices by Carl Feehrer
- Schaefer, B., 1997, Automatic Inflation in the AAVSO Sunspot Number, JAAVSO 26, 40
- Sunspot data page at the National Geophysical Data Center [external link]
- United States National Solar Observatory [external link]
- Solar Influences Data Center (SIDC) [external link]