I found the MCIP output meteorological data are 25 time steps and begin from 00:00.
It needs the previous day wrfout 23:00:00 file when run MCIP, but for reinitial day in WRF (mine is every five days reinitialization), it reminded that it from different domain with different simulation start date.
Could anyone please give me suggestion to solve this problem?
*** SUBROUTINE: CHKWRFHDR
*** WRF FILES DO NOT SEEM TO BE FROM SAME DOMAIN
*** VARIABLE = SIMULATION_START_DATE
*** FIRST FILE = /20170730/wrfout_d01_2017-07-30_23:00:00
*** VALUE IN FIRST FILE = 2017-07-26-00:00:00
*** NEW FILE = /20170731/wrfout_d01_2017-07-31_00:00:00
*** VALUE IN NEW FILE = 2017-07-31-00:00:00
The 23 UTC file is needed when there is a continuation of the WRF simulation so that the precipitation can be correctly incremented. In WRF, the precipitation accumulates continuously from the beginning of the simulation, so that earlier file is needed to isolate the amount of precipitation over the user-defined time increment (usually one hour). When WRF is reinitialized, that earlier file is not needed because the “bucket” begins at zero.
Try again without the earlier file. You will not be able to process 00 UTC from the initial WRF run because a number of other key variables are set to zero at that time, too.
That is correct. You cannot use the initial time from WRF for CMAQ. Most of the time, if we are using reinitialized run segments, we have some sort of overlap period (nominally 12 hours, but it does not need to be that length) that we “throw away”; in your case, the new segment would not pick up at 00 UTC, but it would instead by at 12 UTC (12 hours earlier).
Thank you for this explanation.
So if 12 hours earlier, I should set reinitialized start time 2017-06-30_12:00:00 instead of 2017-07-01_00:00:00 (my current one) to the end time 2017-07-06_00:00:00 (every five days reinitialization), then I can use 2017-06-30_23:00:00 for MCIP. right?