DO Increasing when lake is frozen

Hello,

I am running a model where there is ice covering the lake from the months of October to May. The dissolved oxygen results for the months of June to September look as expected however the months with Ice coverage have a spike of dissolved oxygen when the opposite is expected. How does EFDC+ Generate the DO? Is there a way to locate what is causing this spike in DO? I am using the Heat Coupled Ice Model with the following COD and DO Parameters:

image

Thank you for any help.