I treated the CSV’s power columns as the merged spectra and fitted the dominant band in each spectrum with a Lorentzian plus linear background. The listed intensity is the fitted Lorentzian height; \(x_c\) is its center. The allometric model was \(y=aP^b\), with uncertainty reported as the standard error of the log–log fit.

| Power (µW) | Intensity (counts) | \(x_c\) (eV) |
|---:|---:|---:|
| 5 | 128.9 | 1.31679 |
| 10 | 236.2 | 1.32190 |
| 15 | 334.6 | 1.32426 |
| 20 | 425.7 | 1.32604 |
| 25 | 517.3 | 1.32747 |
| 30 | 621.2 | 1.32843 |
| 35 | 695.8 | 1.32908 |
| 40 | 824.7 | 1.33043 |
| 45 | 888.9 | 1.33066 |
| 50 | 989.1 | 1.33175 |
| 60 | 1189.9 | 1.33227 |
| 70 | 1395.0 | 1.33366 |
| 80 | 1585.6 | 1.33382 |
| 90 | 1817.8 | 1.33451 |
| 100 | 2099.1 | 1.33514 |

The intensity fit gives \(b=0.9224\pm0.0124\). For peak position, the fit gives \(b=0.004531\pm0.000085\).

- [Peak intensity vs power](<<BENCH>/raw/commercial/luna-high/bench/t-pl/run-26/peak_intensity_vs_power.png>)
- [Peak position vs power](<<BENCH>/raw/commercial/luna-high/bench/t-pl/run-26/peak_position_vs_power.png>)
- [Fitted peak values and uncertainties (CSV)](<<BENCH>/raw/commercial/luna-high/bench/t-pl/run-26/lorentzian_peak_results.csv>)

RESULT: b = 0.9224 ± 0.0124