[RegCNET] about cld

[XUNQIANG BI]

Hi, Costas:

On Tue, 19 Jun 2007, Costas Douvis wrote:

> Dear all
>
> I am in doubt about the cld variable of the RAD output files. It is the
> cloud fractional cover but it never takes values larger than 0.8
>
> Is this really the largest value that the model accepts or the 0.8
> corresponds to 100% cloud cover, 0.4 corresponds to 50%, 0.2 corresponds
> to 25% and so on?

In RegCM, clouds that form in individual layers are assumed to be
randomly overlapped in the vertical. So the total cloud for the whole
column would be

  total_cloud(i,j) = 1 - (1-cld(i,j,1))*(1-cld(i,j,2)* ......
                        *(1-cld(i,j,17)*(1-cld(i,j,18)

suppose at one point, cloud for every layer is 0.8, then
the total cloud at this point would be
  1 - 0.2**18 = .999999999999737856

You see, very close to 100% :)
>
Regards,
Xunqiang Bi