Evaluation of Solar Radiation used in NLDAS

A phase error in the EDAS/LDAS solar radiation in the "real-time" forcing data archive has been observed (see figure below).   This reminds us a previous error in the software that interpolates the EDAS field into the LDAS domain. The mission here is to verify whether and when such error has been fixed in
I) the "real-time" EDAS/LDAS1 forcing data; and
II) the "retrospective" EDAS/LDAS2 forcing data;

and also to evaluate the solar radiation forcing in LDAS against SURFRAD observations.

Results:

• the "real-time" EDAS/LDAS1 phase error was fixed in 2000 03 (click here);
  
• the "retrospective" EDAS/LDAS2 does not show such a phase error from the start;
• in general, GOES is in better agreement with observations; while EDAS tends to overestimate the downward shortwave fluxes;
• in winter, GOES sometimes underestimated shortwave fluxes due to miss-identify clouds and ground snow (e.g., 2000 01).