ISSI Team “Sunspot re-calibration”: 2nd meeting (Aug. 19 – 23, 2019)
Brief report of the second meeting of the ISSI Team “Re-calibration of the sunspot number”
The ISSI team “Recalibration of the Sunspot Number” had its second meeting on August 19-23, 2019. This was the occasion to present and discuss the multiple progresses accomplished since the first meeting in January 2018. The sessions were articulated around four main themes: validation of data processing methods, recovery and correction of original sunspot data, benchmarking of new sunspot number series and organisational schemes for future official releases of the upgraded sunspot series. The very open and long discussions led to a final list of common actions needed for assembling a new version of the reference sunspot time series.
Next to the core team members, several young scientists were invited, which indicated the interest of a new generation of researchers who are very active on this topic. As for earlier sunspot number workshops, remote speakers also participated in the sessions. The investigations addressed both the international sunspot number and the group sunspot numbers, which remain neighboring but distinct sunspot-based measures of solar activity.
While intrinsic limitations were identified in all existing statistical re-construction methods, three new approaches were introduced in the framework of this ISSI team. They involve techniques like tied-ranking (T. Dudodk de Wit), a two-component error model (S.Mathieu) as well as an extension of the active-day fraction method (Usoskin et al. 2016, Willamo et al. 2018) by separating different phases of the solar cycle (A. Munoz-Jaramillo). Those innovative approaches still need futher testing and full validation over coming months. They could eliminate the need for inter-comparisons between parallel observers, and also the resulting accumulation of errors backwards in time. They may thus particularly help calibrating the sparse early sunspot observations made before the 19th century.
In this respect, an entirely new vizualization of the temporal density and multual overlap of past observations was published by A. MuÃ±oz-Jaramillo and J. Vaquero (2019). By synthesizing all mutual links between data sets from all past observers, those multi-century timelines reveal a few major data gaps that divide the chronological sunspot record into several loose segments. Those missing links will pose the main challenge for any future sunspot or group number reconstruction.
This panoramic view confirmed the prime importance of recovering additional “lost” data. In this respect, another highlight of this meeting was the presentation of newly recovered and corrected data from the earliest observers in the 17th and 18th century (V.Carrasco, J.Vaquero, M-C. Gallego, R. Arlt), including thus-far unknown observations from Japanese observers (H. Hayakawa). Moreover, the construction of a full digital archive of the sunspot numbers collected by Rudolph Wolf and his successors at the Observatory of Zurich is now at an advanced stage (F.Clette, L.Lefavre, T.Friedli). In this context, missing Zurich source-data tables for the interval 1945 to 1980 were entirely recovered in several archives in Locarno, Zurich and Brussels, closing many years of fruitless searches (F.Clette). After completion of the full digitization of those resurrected Zurich archives, a full end-to-end reconstruction of the sunspot number series during the main Zurich era 1849-1980 will finally be possible.
At this stage, the team agreed that the most mature re-construction method is the improved “backbone observer” method implemented for the group number series (Chatzistergos et al. 2017). In addition, the diurnal geomagnetic oscillation (rY index), which is induced by the EUV solar irradiance variations, was adopted as one of the key benchmarking references. Usable only in the form of yearly means, it gives a reliable validation reference back to 1840 (Svalgaard 2016, Hudson and Svalgaard 2019). This rY index indicates that all reconstructions giving low sunspot numbers before the 20th century, by as much as 40% like the original group sunspot number by Hoyt & Schatten (1998), are incompatible with this geomagnetic reference. It will thus be used for narrowing down the range of valid reconstructions.
By combining those advances in data recovery and statistical methods, this team work is paving the way towards more robust and much more documented versions of the sunspot number and group number series, giving an accurate long-term record of past solar activity.
References and recent publications by the team members:
- Chatzistergos, T., Usoskin, I.G., Kovaltsov, G.A., Krivova, N.A., and Solanki, S.K.: 2017, Astronomy and Astrophysics 602, A69.
- Hoyt, D.V., and Schatten, K.H.: 1998, Solar Physics 181, 491.
- Hudson, H., and Svalgaard, L.: 2019, Physics Today 72, 10.
- Karachik, N.V., Pevtsov, A.A., and Nagovitsyn, Y.A.: 2019, Monthly Notices of the Royal Astronomical Society 488, 3804.
- Munoz-Jaramillo, A., and Vaquero, J.M.: 2019, Nature Astronomy 3, 205.
Svalgaard, L.: 2016, Solar Physics 291, 2981.
- Usoskin, I.G., Kovaltsov, G.A., Lockwood, M., Mursula, K., Owens, M., and Solanki, S.K.: 2016, Solar Physics 291, 2685.
- Willamo, T., Usoskin, I.G., and Kovaltsov, G.A.: 2018, Solar Physics 293, 69.
Agenda, with links to presentations
Monday 19 AM, Part 1: Introduction and progress summaries
Start of the meeting at 8:30 AM
- F. Clette: Welcome, administration, meeting goals & ground rules
- F. Clette: List of key questions for discussions
- L. Van Driel: Expectations after the Jan. 2018 ISSI-1 meeting?
- “Triad” status:
- Kopp: Summary: Sunspot Number Corrections (Lockwood / Kopp / Clette Triad)
- Friedli: Sunspot Number: Wolf-Wolfer transition, 19th century
- Usoskin: Group number: ADF method
- Cliver: Comparison of two backbone-based series: Svalgaard & Schatten (2016) & Chatzistergos et al. (2017)
- Hayakawa: Early data and GN database
- Lefèvre: Methods, errors, best practices
- F. Clette: Common data and software repository. A premature idea?
- Pesnell: Which scheme for future re-calibrations and official release?
- Discussion: refining the agenda and goals
Monday 19 PM, Part 2: Fundamentals first: context and errors
- The re-calibration in a broader context:
Tuesday 20 AM, Part 2 (cont): Fundamentals first: context and errors
- Sources of errors and biases:
- Dudok de Wit: Short-term errors in spot/group observations
- Karachik: The effect of telescope aperture, scattered light, and human vision on early measurements of sunspot and group numbers
- Mathieu: Goals and progress of the Val-U-Sun project
- Cliver: Evolution of the Definition of a Sunspot Group: Impact on group & sunspot number time series
Part 3: Issues in methods
- Methods without observer inter-calibration:
- Usoskin: ADF method: The effect of the reference period
- Lefèvre: Issues in the ADF method
- Dudok de Wit: Merging sunspot records by tied-ranking
- Muñoz-Jaramillo (teleconf): Improvements on the estimation of observer detection threshold using the Active Day Fraction
Tuesday 20 PM, Part 3 (cont): Issues in methods
- Methods with inter-calibration of observer pairs:
- Chatzistergos: New tests of the PDF (probability distribution function) method
- Kopp: Reconstruction methods
- Friedli: Bivariate reduction of sunspot numbers: Is there an Advantage over univariate approaches?
- Svalgaard (teleconf): New Backbone Reconstructions of Sunspot Group Numbers
- T. Dudok de Wit: Suggestions for building a composite group number record
- General discussion
Wednesday 21 AM, Part 4: Open issues in data: Early historical data
- V. Carrasco: A revised collection of sunspot group numbers: 17th century and Maunder Minimum
- H. Hayakawa: A brief Overview for the Sunspot Observations in Paris Observatory
- R. Arlt: New digitized data from Horrebow and Flaugergues
- H. Hayakawa: Are these Sunspot Drawings original? A Case Study in the East Asian Sunspot Drawings and Its Insight
- H. Hayakawa: Key References for the Dalton Minimum: Sunspot Observations by Derfflinger (1802 — 1824) and his contemporaries
- V.Carrasco: A revised collection of sunspot group numbers: 18th and 19th centuries
Wednesday 21 PM, Part 4 (cont): Open issues in data: Modern sunspot data
- T. Friedli: The construction of the Wolf series
- T. Friedli: The Arlt data set: recalibrating Schwabe and Wolf
- T. Friedli: Wolf – Wolfer transition, 19th century
- T. Friedli: Wolf – Wolfer: The Frenkel data set
Thursday 22 AM, Part 4 (cont): Open issues in data: Modern sunspot data
- F. Clette: Construction of the Sunspot Number database: Mittheilungen and recovery of lost Zürich source data.
- V. Carrasco: A revised collection of sunspot group numbers: long series in the 20th century
- H.Hayakawa, F.Clette: Hisako Koyama’s Sunspot Observations and their Digitization
- J. Vaquero: A revised collection of sunspot group numbers: from version 1.2 to 1.3 (synthesis of new changes).
Thursday 22 PM, Part 5: Steps to a Version-3 release and beyond
- Discussion: Scheme for future re-calibration and releases
- Discussion: Where do we stand now? Next actions and priority ranking (cont.):
Friday 23 AM, Part 5 (cont): Action plan, organizing future updates, publications
- Discussion: Where do we stand now? Next actions and priority ranking (cont.):
- Meeting Summaries (L. VanDriel, D. Pesnell, G. Kopp): general wishes and recommendations.