Extinction studies


9/?/05
Point-to-point comparison between Av derived from JHK colors (Tracy) and SED fitting (Shih-Ping)  Plot
10/13/05
Experiments for spectral type fitting

    Sample : 100 stars with 10%  random Gaussian noise to stellar SED models for 14 main sequnce types
    Using my SED fitting program to compare the input and output spectral types and Av  
  1. Distribution of  output spectral types   spec_histo.ps
  2. Av vs. Output Spectral typpes spec_Av.ps
  3. Histogram of output Av for each input spectral types spec_Av_histo.ps
  4. The output Av distribution centered at the input Av value, and the distribution is gaussian-like.
  5. Histogram of output Av for each input spectral types,  but using only specific spectral type  for the SED fitting spec_Av_histo2.ps
  6. This show you fitting a single spectral or the average SED introduce a systematic error. It is not obvious that the systematics will be canceled out by adding more stars.
  7. Histogram of output Av for sample with output spectral types of K5,M0, and M2 output_K5-M2.ps
  8. Note: the input sample contains the same number of stars for each spectral types. In reality the input K5-M2 component should be much bigger than the input B0-K0 component
What we learn:
  1. KM stars can be separated from everything else
  2. Fitting spectral types does NOT give you reliable spectral type information for O-G stars
  3. Av from fitting spectral types show normal distribution centered at the input Av value
  4. Av from fitting specific spectral type or the average spectral type introduce systematic shift
Conclusion(?) : 
  1. Fitting spectral types reduce systematic errors
  2. Since KM stars can be more reliably determined, we can use them to caculate extinction laws and intrinsic standard deviataion  of Av