Definition of DNI and DHI inputs values to SAM PVWatts5

I am using PVLIB to generate some 1-minute DNI and DHI input for use in the SAM PVWatts5 module. Normally, I just pull solar resource data right from NREL using the NSRDB API or using MIDC API, but I can only get 30 min resolution on DNI and DHI for NSRDB and MIDC has limited coverage, so that's why I'm trying PVLIB.

My question is: Do the DNI and DHI values used by PVWatts5 in SAM represent the plane of array irradiance, assuming the array's normal vector always points toward the sun? For DNI this seems an obvious yes by the definition of DNI, but it's not so clear with DHI...the PVLIB algorithms require a tracking angle to determine DHI.

Many thanks!
Mike

Actually, I just found the PV Model Technical Reference, which states the following (pdf pages 13-15):
URL: www.nrel.gov/docs/fy15osti/64102.pdf

DNI: The portion of the solar irradiance that reaches a surface normal to the sun in a direct line from the solar disk (typically assuming a measurement device with a 5? field of view), called beam normal or direct normal irradiance.

DHI: The solar irradiance on a horizontal surface from the sky excluding the solar disc.

So, looks like DNI is calculated as expected (pointing towards sun), while DHI assumes horizontal surface (which, upon reflection, makes sense from the name as well...). SAM uses the Isotropic, HDKR or Perez...I'm going with HDKR for now.

Someone please correct me if I misinterpreted the document.