Difference between gfdda_interval_m and fgdt

[mmi]

Hi,

In the namelist.input, what is the difference between gfdda_interval_m [time interval between analysis times] and fgdt [calculation frequency for analysis nudging]? For instance, what happens if gfdda_interval_m = 360 (i.e., 6 h) and fgdt=0 (i.e., every time step)?

Thanks.

 

[Ming Chen]

gfdda_interval_m is the interval (in min) between analysis times

fgdt is the calculation frequency for grid-nudging

 

[mmi]

Thanks Ming. If my analyses are available every 6 hours, and the calculation frequency for grid-nudging is 1 hour, what is the data used for nudging? Is it the analysis data interpolated at those times?

Thanks.

 

[Ming Chen]

In this case grid-nudging should be done at 6-h interval. If you set the interval to be 1 hour, the analysis data will be interpolated, which doesn't really make sense.

 

[mmi]

Thanks Ming.
It's strange that the WRF guide recommends setting fgdt=0 (every time step).

 

[Ming Chen]

I apologize for misunderstanding your previous post. It is recommended to conduct nudging every time step, and the analysis data at each time step is interpolated from the 6-h interval reanalysis data.

 

[mmi]

Thank you Ming. It's clear now.