Best ESPC-D-V02 'Nowcast' Product

[mark]

I am trying to determine the ESPC-d products (i.e. _ssh.nc, _t3z.nc, etc.) from /datasets/ESPC-D-V02/data/forecasts that would be the forecast dataset to use that most accurately represents a 'nowcast' product if I am retrieving data at roughly 04:00 AM MDT/MST (06:00 AM EDT/EST, 10:00 UTC).  For example, at present, I am retrieving the /datasets/ESPC-D-V02/data/forecasts/US058GCOM-OPSnce.espc-d-031-hycom_fcst_glby008_2025082512_t0012_t3z.nc, but not sure if that is the best representation of conditions for the morning of 20250826.  Would a forecast product farther into the future be a more accurate 'prediction' of the conditions (e.g. ocean temperature at the surface and depth, SSH and currents)? 

I also see products such as /datasets/ESPC-D-V02/data/daily_netcdf/2025/US058GCOM-OPSnce.espc-d-031-h‎ycom_fcst_glby008_2025082512_M0000_t3z.nc, but they post roughly 8-12 hours after the forecast products and not sure how much they differ from the t0012 product.  Can you please provide recommendations on what the best dataset would be to represent a nowcast product at the time I am downloading (early morning on the date after the dataset date)?

Also, is there documentation somewhere that tells how far in the past model input data from based on the forecast time? For example, for the model run 20250825 t0012, are the input data from 12Z 20250824 through 12Z 20250825?

 

Thanks,

 

Mark

[Alan Wallcraft]

The analysis time is the date ending in 12 (e.g. 2025082512 for 12Zulu Aug 25th 2025), and the forecast length is "tau" (e.g. t0012 for 00Zulu on the 26th).

The daily_netcdf files are the daily means, and the start time of the mean is the analysis time (e.g. 2025082512) and its nominal time is 12 hours later (00Zulu on the 26th).

Alan.