A guest post by Michael Ronayne
Note: Mike has created a movie (solar_cycle_23-24_sunspots.gif large (2.4MB) animated GIF) that shows how the cycle 23 forecast has progressed through time. Given that NASA’s David Hathaway recently commented on SpaceWeather that we are still seeing Cycle 23 spots, this seemed like a good time to post Mike’s effort.
The Space Weather Prediction Center (SWPC) at http://www.swpc.noaa.gov/ issues weekly reports on solar activity know as Preliminary Report and Forecast (PRF) of Solar Geophysical Data or “The Weekly”. Generally on the week following the end of the month a monthly summary is issued which includes graphics for the past month.
In the summary is the “ISES Solar Cycle Sunspot Number Progression” graphic which shows past, present and predicted average sunspot numbers by month. SWPC maintains a compressed archive of all weekly PRD reports in PDF format since 1996 which is available here.
The most current graphic is always here.
All of “The Weekly” reports were inspected to identify the monthly summaries and determine the quality of the “ISES Solar Cycle Sunspot Number Progression” graphic contained therein. It was determined that the graphs prior to April 30, 2003 were in a significantly different format, had quality control problems and skipped months, therefore only graphs from April 30, 2003 to present were used.
Using Adobe Acrobat Professional the “ISES Solar Cycle Sunspot Number Progression” graphics was extracted from each of “The Weekly” PDF reports as oversized TIFF graphics to preserve resolution. The standard publication size for the graphic was 720×550 pixels but the aspect ratio for some of the graphs was not preserved within the PDF document. When the oversized TIFF graphic were resized to 720×550 without preserving the aspect ratio within the PDF the original 720×550 graphic was recovered in all cases. The 720×550 TIFF graphic was then converted to a GIF graphic for use in the animation sequence.
While extracting the “ISES Solar Cycle Sunspot Number Progression” graphs it was found that January 31, 2008 monthly summary had not been generated, a fact which SWPC confirmed in response to an Email inquiry. The February 29, 2008 graphic was hand edited at the pixel level to recreate the missing month and is identified in the animation sequence “proxy200801.gif”. The remaining graphics are all identified by the PRF document number.
The Advanced GIF Animator program was used to create the animation sequence. With the exception of January 31, 2008 all of the frames are prefixed by PRF9999 when 9999 is the document number of the original PDF report from which the graphic was extracted.
When the animated frames were inspected in sequence it was found that there was a discontinuity between July 31, 2006 (PRF1510), August 31, 2004 (PRF1514) and the September 30, 2004 (PRF1520) frames. The causes of the discontinuities were:
- Data was retroactively changed on the August 31, 2004 frame.
- The August 31, 2004 data point was not plotted on the August 31, 2004 frame.
These three frames were not altered or correct in anyway and are displayed as published. This technique is very good at identifying data discontinuity problems.
Excluding the problems noted above the reconstructed graphic went very well and there was no discernible flicker between frames indicating that the PDF extraction process was near prefect. With the exception of the problem about August 31, 2004 and the missing monthly summary for January 31, 2008 the SWPC product has been amazingly consistent since April 30, 2003.