Inconsistent default settings for _prep_inputs_solar_pos in prepare_inputs and prepare_inputs_from_poa
See original GitHub issueHi there,
I find that _prep_inputs_solar_pos
method has been both called in prepare_inputs
and prepare_inputs_from_poa
. However, the former takes an additional argument, press_temp that contains temperature pulled from the weather data provided outside. For the default nrel_numpy
algorithm, I further checked its input requirement is avg. yearly air temperature in degrees C rather than the instantaneous temperature provided in weather. Hence I would like to ask if the following codes in prepare_inputs
are redundant at least for the default ‘nrel_numpy’ algorithm?
# build kwargs for solar position calculation
try:
press_temp = _build_kwargs(['pressure', 'temp_air'], weather)
press_temp['temperature'] = press_temp.pop('temp_air')
except KeyError:
pass
And thereby we change self._prep_inputs_solar_pos(press_temp)
to self._prep_inputs_solar_pos()
in prepare_inputs
?
Meanwhile, does the temperature really matter? How much uncertainty will it cause in the calculation of the sun’s position? Should we provide avg. local temperature data if for a global modelling purpose?
Any help would be appreciated!
Issue Analytics
- State:
- Created 3 years ago
- Comments:19 (16 by maintainers)
Top GitHub Comments
Looks to me like Michalsky 1988 is the one to cite.
It looks like SAM uses that solpos C function. For the refraction calculation (which appears to be different than in SPA) pressure and temperature can be time-dependent, or fixed at defaults (1013mb and 15C, the docstring says 10C but that looks to be a typo).
I don’t have a preference for any of the solar position algorithms. A drawback of SOLPOS is that it does not appear to be published other than as code.