Kramers-Kronig relations for limited data point

  • Context: Graduate 
  • Thread starter Thread starter physengineer
  • Start date Start date
  • Tags Tags
    Data Point Relations
Click For Summary
SUMMARY

The discussion focuses on applying Kramers-Kronig relations to measure complex-optical conductivity when only the imaginary part is available for a limited wavelength range of 1030 nm to 2300 nm. It is established that while Kramers-Kronig relations require integration over the entire spectrum, one can extrapolate the missing real part by fitting a curve to the available data. Additionally, the discussion highlights the importance of minimizing edge effects through the application of window functions, acknowledging the practical limitations of data acquisition in optical measurements.

PREREQUISITES
  • Understanding of Kramers-Kronig relations
  • Familiarity with complex-optical conductivity
  • Knowledge of curve fitting techniques
  • Experience with window functions in data analysis
NEXT STEPS
  • Research methods for extrapolating data in optical measurements
  • Learn about different curve fitting techniques and their applications
  • Study the impact of edge effects in signal processing
  • Explore various window functions and their effectiveness in minimizing edge effects
USEFUL FOR

Researchers and engineers in materials science, optical engineering, and anyone involved in the analysis of complex-optical conductivity using limited data sets.

physengineer
Messages
19
Reaction score
0
Hello,

I need to measure the complex-optical conductivity of some materials. The problem is that I can only measure the imaginary part of the complex conductivity only for limited wavelengths between 1030 nm and 2300 nm.

From Kramers-Kronig relations, we know that the real and imaginary parts of the optical conductivity are related but the required integrations are from -\infty to
+\infty.

Is there any way to still be able to use Kramers-Kronig relations when I only know the imaginary part for just an interval rather than the whole spectrum.

I would appreciate any help in this regard!
 
Physics news on Phys.org
If you have enough data, you may be able to reasonably extrapolate to find the other values. Fit a curve to your data and use the curve for values you don't have.

Also, in practice, we can never acquire an infinite number of points in order to satisfy certain relations such as Kramers-Kronig or Fourier Transforms, and thus edge effects are inevitable. We can minimize edge effects by applying http://en.wikipedia.org/wiki/Window_function" .
 
Last edited by a moderator:

Similar threads

  • · Replies 3 ·
Replies
3
Views
4K
  • · Replies 5 ·
Replies
5
Views
2K
  • · Replies 1 ·
Replies
1
Views
3K
  • · Replies 4 ·
Replies
4
Views
4K
  • · Replies 6 ·
Replies
6
Views
18K
  • · Replies 8 ·
Replies
8
Views
13K
  • · Replies 0 ·
Replies
0
Views
3K
Replies
3
Views
4K
  • · Replies 8 ·
Replies
8
Views
5K
  • · Replies 7 ·
Replies
7
Views
4K