The effect of a new snow and sea ice albedo scheme on regional climate model simulations

Abstract

The HIRHAM snow and sea ice albedo scheme and several other existing snow and sea ice albedo parameterizations forced with observed input parameters are compared with observed albedo. For snow on land in non‐forested areas, the original linear temperature‐dependent snow albedo is suggested to be replaced with a polynomial temperature‐dependent scheme. For sea ice albedo none of the evaluated schemes manage to simulate the annual cycle successfully. A suggestion of a new sea ice albedo including the effects of melt ponds, snow on the sea ice and the surface temperature is presented. Simulations with original and new snow and sea ice albedo are performed in the regional atmospheric model HIRHAM and the results are compared. Compared with ERA40 the control simulation with original surface albedo reveals a warm bias during spring in the Arctic. Changing the surface albedo, the biggest differences are found in the same period. Model simulations with old and new surface albedo in HIRHAM clearly reveal that the new albedo scheme is superior to the currently implemented scheme in reproducing the ERA40 temperature climatology. In these experiments the new snow albedo scheme has less impact than the new sea ice albedo. This is probably because areas with changed snow albedo have smaller extent than areas with sea ice in the model setup and are more constraint by the lateral boundaries.