Compile air temp data for NHD and NHM scales?

[lekoenig]

For the downscaling experiments we need to compile meteorological drivers, including daily air temperature, at the NHD-resolution using gridmet data. For the NHM scale, air temperature (seg_tave_air) is the daily average air temperature from PRMS-SNTemp, but gridmet only contains daily min/max values. Using either the daily min or max results in a systemic bias in the raw air temperature values between the NHD and NHM scales, as shown in the plots below.

In #19, we used the average of the daily min and max temperatures from gridmet to make the values more comparable between NHD and NHM. However, given the importance of air temperature for water temperature predictions, we may want to take a further step of compiling temp data in the same way for the NHM-scale (i.e., use the data aggregated from gridmet rather than using the SNTemp values). If we do that, we'll need to rename the variables so that the names are shared between the NHD and NHM scales but differ from the naming used in PRMS/SNTemp (e.g. use seg_airtemp_gridmet).