It’s not even possible to fit all of the data points by tweaking Cd, so Zwift is using a different formula to compute frontal area. Using the Cycling Power Labs formula and classical physics equations produces significantly different numbers (higher across the board) than found from your simulations. This is a simple analysis - nothing more than a reality check. I neglected the 16 m elevation gain at 1% since I figure that makes a 0.01 watt difference average over the 16 km course. I used equations to include aerodynamic drag and rolling resistance in a steady state.

I guess you’ll going to land on nowhere with Cycling Power Lab formula. To start the thread, I’m going to paste some messages from an older topic. But when I need something directly related to my figure or my equipment or a preferred route? I hope this thread will be here to register data from your tests. Sure I’m aware of Zwift Insider amazing work on the field. A Strava segment may be an alternative, but then you’re going to skew up your profile adding rides generated by the simulator and NOT by you. After a lot of mumbling, IMHO, the easiest way is to take screenshots on know points, like a finish line, and manually calculate times. Anyway, else in workout mode on a given route, there is no guarantee your avatar will start precisely on the same point. Ride a TT bike or enter workout mode or both. You also would hopefully clear the field of drafting effect. I’ll not enter in details of the actual set-up unless requested. Such a device is often known as ANT+ simulator. That’s when you’ll need a way to automagically submit power and cadence to the system. Maybe you want to run some experiments on Zwift.

Ant simulator download