1. Zeta Cassiopeiids (ZCS-444)
Analysis Method and Automation

2. IAU working list
Observations / analysis must “strongly suggest” a shower
Discoverer of probable new shower proposes a unique name to the Meteor Data Centre (MDC)
Discoverer presents claim to IMO
Accepted streams reported via paper (WGN)
ESTABLISHED if subsequent research establishes validity

3. Things that concern my enquiring mind *
“Observations / analysis must “strongly suggest” a shower”
How much evidence is sufficient to “strongly suggest”?
Is there a minimum standard for this evidence?
Is there criteria for acceptance ?
“ESTABLISHED if subsequent research establishes validity”
Who does this research and when?
Are there standard tools for evaluation?
It’s a long list (957 showers listed, only 112 established)!
* Must get out more

4. An analysis by Paul Roggermans

5. Approach followed by Paul Roggermans
Approach based on orbital (dis)-similarity:
Find candidates meteors
Determine a representative “Master Orbit”
Allocate observations to D criteria bands (using master orbit)
Analyse
Iterative calculation of median orbital elements for orbits below D threshold
Filter by velocity and radiant position
Uses DD, DSH and DH
Analysis by criteria band

6. Candidate Zeta Cassiopeiids
Search entire EDMOND / SONOTACO / CAMS data for orbits similar to the ZCS reference orbit:
Time interval: 97° < λʘ < 125°
Radiant area: 350° < αg < 22° and +41° < δg < +60°
Velocity: 51 km/s < vg < 62 km/s
3,149 candidate orbits selected

7. Determine the “reference orbit”
Initial value:
Calculate median orbital elements of the 3149 candidate orbits
Pass 1:
Use median to calculate new D Criterion
Calculate median orbit for all orbits with DD < 0.04
Pass 2:
Use new median to recalculate D Criterion
Re-calculate median orbit for all orbits with DD < 0.04

8. Discrimination Bands
Characteristics are then plotted according to discrimination bands:
DSH DD DH
Low
<= 0.25
<= 0.105
Medium-Low
<= 0.20
<= 0.08
Medium High
<= 0,15
<= 0.06
High
<= 0.10
<= 0.04
Very high
<= 0.05
<= 0.02

9. Calculation performed in excel
Many wide (massive) worksheets
Lots of computed columns
Iterative steps performed manually
Laborious and high risk of mistakes

10. A New Approach
SAS Analytical software is good at processing this sort of data
CSV data can be imported easily
Data can be manipulated / filtered:
Sequentially (SAS DATA step)
Using SQL query language (SAS PROC SQL statement)
SAS Studio enables workflows to be created simply

11. SAS Studio Workflow

12. Workflow Part 1

13. Workflow Part 2

14. Outputs

15. Example DATA Step

16. Comments on the Analysis Approach
How to calculate median or mean angles?
Add 360° (i.e mean of 355° and 10° becomes mean of 355° and 370°):
Calculate on the basis of a resultant vector?
Median of individual elements results solution that does not meet laws of celestial mechanics:
T. J., Rudawska, R., & Pretka-Ziomek, H. (2006). Calculation of the mean orbit of a meteoroid stream. Monthly Notices of the Royal Astronomical Society, 371(3),

17. Comments on SAS vs R Comparable performance (but R has the edge)
Both are freely available (SAS University edition)
Both import CSV easily - but there are still pitfalls
Neither are exactly “mainstream” but SAS less so
SAS has good graphics capability but R is better
SAS makes lighter work of tabular data than R (once imported)
SAS’s ability to combine sequential processing& SQL language is powerful
SAS enables complex workflows to be built (simple steps / reusable elements)
SAS supports sophisticated macro processing

18. But…
You just got us all to love R and now you want to introduce us to another ****** language!
Woolfram Mathematica…? anyone ?