From Martin et al:
Table 1. Current HadGEM2 configurations. Configuration Processes included
HadGEM2-A Troposphere, Land Surface & Hydrology, Aerosols
HadGEM2-O Ocean and sea–ice HadGEM2-AO Troposphere, Land Surface & Hydrology, Aerosols, Ocean & Sea-ice
HadGEM2-CC Troposphere, Land Surface & Hydrology, Aerosols, Ocean & Sea-ice, Terrestrial Carbon Cycle, Ocean Biogeochemistry
HadGEM2-CCS Troposphere, Land Surface & Hydrology, Aerosols, Ocean & Sea-ice, Terrestrial Carbon Cycle, Ocean Biogeochemistry, Stratosphere
HadGEM2-ES Troposphere, Land Surface & Hydrology, Aerosols, Ocean & Sea-ice, Terrestrial Carbon Cycle, Ocean Biogeochemistry, Chemistry
"Lakes in the Met Office Surface Exchange Scheme version 2 (MOSESII; Essery et al., 2003) are not modelled interactively, but have a fixed extent"
From Collins et al:
"The land surface scheme in HadGEM2 is MOSES II (Essery et al., 2003)"
"Lakes in MOSES II do not vary with climate, so the areas remain constant. In HadGEM1 evaporation from lakes was therefore a net source of water into the climate system. In HadGEM2 this evaporation now depletes the soil moisture. It would have been desirable to take this from the surface grid squares containing the lake, but this often led to unacceptably low soil moisture levels. As a compromise, the global lake evaporation flux is calculated and removed evenly from the deep soil moisture over the whole land surface (providing the soil moisture content is greater than the wilting point in the grid box). This water conservation is necessary to diagnose trends in sea level. Precipitation falling onto a grid square adds to the soil moisture regardless of the lake fraction in that grid square, thus always conserving water."