Anomalous radiance bump at 620 nm in cloudy conditions

Our research group uses a pySAS system with HyperOCR radiometers in the Laurentian Great Lakes. After the 2025 field season, we had our radiometers calibrated and characterized at the University of Tartu. When re-processing the 2025 data, we still observed the anomalous radiance bumps around 620 nm that had previously been brought up in #345. They only seem to appear in the more oligotrophic waters that we sample in and only when the L_sky sensor shows cloud contamination.

We have processed the data with HyperCP using both the class-based and sensor-specific processing regimes, and the feature is visible in both with the sensor-specific processing appearing slightly worse. Here is an example showing the derived reflectance using the class-based approach (CS pySAS), the sensor-specific approach (SS pySAS), and the reflectance from our handheld radiometer (ASD FieldSpec 4) at one discrete station from April 2025.

I have put the raw file, HyperCP config file, and .sip used as inputs to HyperCP, as well as the generated reports for both class-based and sensor-specific runs, in a Drive folder available here:
https://drive.google.com/drive/folders/1uGs-wDUiU2q2sdMpnJAMD0bxZLPdJ0HO?usp=sharing

Thanks,
Karl

Start with the supplied raw file, HyperCP configuration, .sip input, and generated reports, then compare the class-based and sensor-specific results around 620 nm under cloud-contaminated L_sky conditions. Review issue #345 for prior investigation and use the ASD FieldSpec 4 comparison as the reference; done means identifying the processing cause and confirming the anomalous bump is resolved or consistently explained in both regimes.