Sediment Heat Exchange Coefficient and Sediment Temperature (CBHE and TSED) Extremely Sensitive to Simulated Bottom Temperature

I am applying CE-QUAL-W2 to a Lake in Virginia. Things are going quite well and the model is doing a great job. The only problem is the simulation of the bottom temperature that is extremely sensitive to the sediment temperature and the water-sediment heat exchange coefficient. The observed temperature profile data in this lake (with depth varying from 40 to 45m) is quite detailed and cover all the depth profile up to the bottom of the lake. The observed data spans the period of April to November (on the average 2-3 profiles a month) – Observed bottom temperature varies for 5 oC in April to 14.5 oC in August- September Because of this quite detailed observed data (temperature measured close to the bottom of the lake) and the seasonal variation in observed lake bottom temperature (from 5 to 14.5 oC), it has been very difficult to match the bottom temperature profile (hypolimnion) using a constant sediment temperature and water-sediment heat exchange coefficient. Ideally it would have been extremely useful to input monthly sediment temperature values. I tried numerous sensitivity analyses, refinements to the surface and vertical segmentations, and used an extremely low time-step but the problem persisted. Is it possible to turn-off completely the water-sediment heat exchange? Any other recommendations are appreciated. Thank you for all your help. Djamel