Data assimilation not working properly

[der-knight]

Hello lis community

I am trying to assimilate smops ascat soil moisture in noah mp. The noah mp open loop has shown good results when compared with ground data but the data assimilation is not working properly.
I tried changing the perturbations but there was minimal change in data assimilated output.
I believe there is a major mistake since the data assimilated output is supposed to be between open loop and observations but in my case the output becomes a straight line away from both data points.

I am attaching the :
cdf and enrst nc files
da config file
lvt comparison image
attributes files

DA-lisf.zip