Dear RegCM4-CLM3.5 developers,<BR>Now I am studying the difference of climate effects with BATS1e and CLM3.5 land modle over East Asia. But I found the surface temperature simulated with RegCM4-CLM3.5 was averaged lower (area averaged about 6 degrees, up to 10 degrees over some regions) than NCEP reanalysis data or station observation data. By contrast, the RegCM4-BATS model simulates the similar distribution and value of the surface temperature. Why the bias is so high? Is CLM3.5 model not completely coupled with RegCM4 model? My spin-up time is one year. Please kindly give me some suggestions.&nbsp;Three figures of temperature distributions in apr2006 are attached. You can see the difference among them. Thanks very much in advance. <BR>Su&nbsp;&nbsp;&nbsp; <BR>&nbsp;