In 1989, a SiB (Simple Biosphere) model was developed as a land-surface scheme for JMA's operational NWP model (JMA-GSM). However, the SiB scheme has not been modified substantially after that. Recently, problems considered to be relevant to the land-surface scheme were pointed out as follows:

* Overestimate of thaw.

* Warming bias in a lower troposphere on ice sheet area in summer.

In order to solve such problems, we have developed a new SiB, which treats soil and snow processes more accurately. Major changes are as follows:

* A conventional force restore method for predicting soil temperature is abandoned and thermal conduction among multiplied soil layers is explicitly calculated.

* Phase change of soil water is considered.

* Multi-layered snow model is introduced and phase change of snow water is considered.

* Snow cover is classified into two categories, "partial snow cover" and "full snow cover".

* A sophisticated snow process is introduced. (Aging of snow albedo, temporal changes of snow density and thermal conduction, including water in snow layers and so on.)

We carried out preliminary forecast experiments for April 2002 with JMA-GSM. The new SiB predicted a snow cover area at Tibetan Plateau better than the operational one, comparing with the SSM/I observation. However, the overestimate of thaw had still remained in vast extent of taiga.

As a result of various sensitivity experiments, it became clear that the overestimate of thaw is closely related to the estimation of wet-snow albedo. Meanwhile, it was found that the warming bias in a lower troposphere on ice sheet was relevant to a vertical diffusion coefficient in a PBL scheme.

We will improve further the snow processes of the new SiB based on these experiments. We are planning to put it into operation within a next few years.