The file contains one spectrum at each power from 5 to 100 µW. I fitted two Lorentzian bands per spectrum, each with a local linear baseline. The table of fitted peak intensities, positions, and widths is in [lorentzian_peak_fits.csv](<BENCH>/raw/commercial/luna/bench/t-pl/run-21/lorentzian_peak_fits.csv).

The plots show the fitted values and their allometric power-law curves:

- [Peak intensity vs power](<BENCH>/raw/commercial/luna/bench/t-pl/run-21/peak_intensity_vs_power.png)
- [Peak position vs power](<BENCH>/raw/commercial/luna/bench/t-pl/run-21/peak_position_vs_power.png)

For intensity versus power, the fitted exponents are \(b=0.9233 \pm 0.0121\) for Peak 1 and \(b=0.8849 \pm 0.0333\) for Peak 2. For position versus power, they are \(b=0.00447 \pm 0.00008\) and \(b=0.00194 \pm 0.00041\), respectively. Uncertainties are standard errors from the log–log fits.

RESULT: b = 0.9233 ± 0.0121