Question regarding "GCOMW_AMSR2L3SND" reader

[yhkwon8329030]

Dear LISF team,

I am trying to use the AMSR2 snow depth data assimilation (DA) functionality in LIS.

While reviewing the source code, I noticed the following lines (184–185 and 193–194):

                GCOMW_AMSR2L3SND_struc(n)%sndobs(c,r) = &
                     sndobs_A(c+(r-1)*LIS_rc%obs_lnc(k))/100.0

I am wondering why the AMSR2 snow depth observations are divided by 100.0 at this point. In contrast, the bias-corrected observations do not appear to be divided by 100.0 later in the subroutine.

Could you please explain the reason for this difference? Specifically, I would like to understand whether the division is intended for unit conversion or serves another purpose.

Thank you very much for your help.

Best regards,
Yonghwan Kwon

[dmocko]

Hi @yhkwon8329030

What is the exact dataset you are using with this reader?

The GCOMW_AMSR2_SND data that I see on Discover has a scale_factor of 0.1, which doesn't match the division by 100.0 in the code.

Other than that, I'm not sure why the division is done at this point in the code, but not later in the subroutine.

We will need to take a closer look.

[yhkwon8329030]

Thank you for your answer.

I am using the GCOM-W AMSR2 Level-3 Daily Global Snow Depth product from JAXA.

I noticed that a scale factor of 0.1 is applied when the data are read, and later the values are divided by 100.0. I assume there is a reason for this implementation, but I have not been able to understand why these two scaling steps are used.

If you could find out the reason behind these code lines, I would greatly appreciate it.