I’m actually building something that tackles exactly this problem to some degree.
Right now the engine runs on the manufacturer PAR maps I’ve gathered, blended with an attenuation model for depth. It can’t account for every nuance (glass refraction, rock shadowing, surface agitation, etc.), but it’s surprisingly close for ball-parking how many lights you’ll need and where to hang them.
Im short on real world PAR readings to keep refining the model, so if anyone has data I’d love to incorporate it.
Let me know what breaks or what features you’d like next:
PAR Distribution Calculator
Cheers