modification of cloud cover calculation and adding namelist options to Thompson MP

[RuiyuSun]

In the gfsv17 based Thompson MP experiments, there is significantly less cloud cover (low and high) compared with CERES data. The radiative flux biases (dsw at surface and OLR) are larger in the Thompson MP experiments than in the gfsv17 control. The purpose of the work is to increase the cloud cover and reduce the radiative flux bias at the surface and TOA.