Sigma(r) Forecast vs Observation Time

This script estimates the uncertainty on the tensor-to-scalar ratio, ( \sigma(r) ), as a function e.g. of total observation time for a CMB experiment like Simons Observatory.

Features

• Converts detector-level NETs and observing time into per-band noise levels (in μK·arcmin).
• Models noise power spectra including low-ℓ atmospheric-like excess (( 1/\ell^\alpha )).
• Combines frequency bands using a mixing matrix with realistic foreground components (CMB, dust, synchrotron).
• Computes ( \sigma(r) ) via a Fisher matrix approach, marginalizing over ( A_{\text{lens}} ).
• Allows tuning of:
  • Observation duration
  • Delensing efficiency
  • Detector configuration and noise models

Requirements

• Python 3.x
• camb
• fgbuster
• numpy, matplotlib, scipy

Install dependencies:

pip install camb fgbuster numpy matplotlib scipy

Usage

To run the forecast:

python fisher_sigma_r.py

Output

A plot of ( \sigma(r) ) versus total observation time is displayed, with a logarithmic y-axis.

Author

Developed by SciPol project, CNRS / APC

For questions, contact: [email protected]