I identified two peaks in each spectrum. Each spectrum was fit over 1.22–1.58 eV with two Lorentzians and a quadratic background. Intensity below is the fitted Lorentzian height above background; \(x_c\) is the fitted center.

| Power (µW) | Peak 1 intensity (a.u.) | Peak 1 \(x_c\) (eV) | Peak 2 intensity (a.u.) | Peak 2 \(x_c\) (eV) |
|---:|---:|---:|---:|---:|
| 5 | 134.0 | 1.317806 | 27.7 | 1.527982 |
| 10 | 249.4 | 1.322795 | 52.2 | 1.527892 |
| 15 | 348.0 | 1.324791 | 79.3 | 1.530543 |
| 20 | 445.5 | 1.326459 | 106.6 | 1.531473 |
| 25 | 544.3 | 1.327783 | 126.4 | 1.531893 |
| 30 | 647.4 | 1.328752 | 134.4 | 1.532349 |
| 35 | 730.0 | 1.329273 | 156.6 | 1.533306 |
| 40 | 861.6 | 1.330518 | 171.8 | 1.533494 |
| 45 | 925.9 | 1.330796 | 191.9 | 1.532855 |
| 50 | 1032.4 | 1.331829 | 213.4 | 1.531647 |
| 60 | 1228.9 | 1.332297 | 261.9 | 1.532602 |
| 70 | 1431.1 | 1.333452 | 273.3 | 1.533094 |
| 80 | 1617.9 | 1.333654 | 291.7 | 1.535832 |
| 90 | 1855.6 | 1.334322 | 317.4 | 1.535606 |
| 100 | 2137.8 | 1.334950 | 325.9 | 1.534866 |

Plots: [peak intensity vs power](peak_intensity_vs_power.png) · [peak position vs power](peak_position_vs_power.png). The full fit results, including widths and fit RMSE, are in [lorentzian_peak_results.csv](lorentzian_peak_results.csv).

Allometric fits \(y=aP^b\) gave intensity exponents \(b=0.9133\pm0.0103\) for Peak 1 and \(b=0.8268\pm0.0210\) for Peak 2. For peak position, they gave \(b=0.004201\pm0.000084\) and \(b=0.001639\pm0.000194\), respectively. Uncertainties are 1σ standard errors of the log–log regression.

RESULT: b = 0.9133 ± 0.0103