Initial condition update

[Monica]

Hi all,
When we run MPAS-A or WRF, we need to update the initial conditon or boundary condition, such as every 6 hours. However, the problem is that when we want to predict future weather, we cannot obtain updated initial field data (because cannot obtaine future data now). For example, it is 00:00 on November 7, 2022, and I need to forecast the weather at 12:00 on November 10. But I can't get the observation data at 06:00, 12:00 and 18:00 on the 7th. I don't know where my understanding goes wrong.
In addition, I also want to know how to calculate MPAS-A. My understanding is that from 00:00 on the 7th to 06:00 on the 7th, and then from 6 hours to 12:00 on the 7th with the updated initial field integral, and so on. Is my understanding correct?
Best,
Monica

 

[mgduda]

For global simulations, only initial conditions are needed (if we ignore any updates to forcing fields), and ultimately, it's these global simulations (often run by operational NWP centers) that are used to produce boundary conditions for regional simulations that run into the future. In other words, the boundary conditions do not come directly from future observations, but from the output of other models.

In addition, I also want to know how to calculate MPAS-A. My understanding is that from 00:00 on the 7th to 06:00 on the 7th, and then from 6 hours to 12:00 on the 7th with the updated initial field integral, and so on. Is my understanding correct?

It may be that I'm not completely understanding the question, but in MPAS-A we integrate the initial state forward in time (subject to boundary conditions in the case of a regional simulation) using an integration step size that is specified by the namelist option 'config_dt'. This integration time step is typically much smaller than 6 hours, which is the default interval between model output files for the "output" stream.