Source code for bsrn.qc.closure
"""
BSRN Level 3 inter-comparison checks: GHI–BNI–DHI closure.
"""
import numpy as np
import pandas as pd
[docs]
def closure_low_sza_test(ghi, bni, dhi, zenith):
r"""
Check consistency between GHI, BNI, and DHI for low solar zenith angles ($Z \le 75^\circ$) [1]_.
Parameters
----------
ghi : numeric or Series
Global horizontal irradiance ($G_h$). [W/m^2]
bni : numeric or Series
Beam normal irradiance ($B_n$). [W/m^2]
dhi : numeric or Series
Diffuse horizontal irradiance ($D_h$). [W/m^2]
zenith : numeric or Series
Solar zenith angle ($Z$). [degrees]
Returns
-------
flags : Series or ndarray
Boolean flags (True = Pass). [bool]
References
----------
.. [1] Long, C. N., & Shi, Y. (2008). An automated quality assessment
and control algorithm for surface radiation measurements. The Open
Atmospheric Science Journal, 2(1), 23-37.
"""
mu0 = np.cos(np.radians(zenith))
# GHI reconstructed from BNI and DHI
ghi_calc = bni * mu0 + dhi
ghi_calc_safe = np.where(ghi_calc > 0, ghi_calc, np.nan)
# Condition: |GHI / (DNI * cos(SZA) + DIF) - 1| <= 0.08
diff_ratio = np.abs(ghi / ghi_calc_safe - 1)
# Domain: Z <= 75 and GHI > 50
in_domain = (zenith <= 75) & (ghi > 50)
condition_met = diff_ratio <= 0.08
if hasattr(in_domain, 'iloc'):
return (~in_domain) | condition_met
else:
return (not in_domain) or condition_met
[docs]
def closure_high_sza_test(ghi, bni, dhi, zenith):
r"""
Check consistency between GHI, BNI, and DHI for high solar zenith angles ($Z > 75^\circ$) [1]_.
Parameters
----------
ghi : numeric or Series
Global horizontal irradiance ($G_h$). [W/m^2]
bni : numeric or Series
Beam normal irradiance ($B_n$). [W/m^2]
dhi : numeric or Series
Diffuse horizontal irradiance ($D_h$). [W/m^2]
zenith : numeric or Series
Solar zenith angle ($Z$). [degrees]
Returns
-------
flags : Series or ndarray
Boolean flags (True = Pass). [bool]
References
----------
.. [1] Long, C. N., & Shi, Y. (2008). An automated quality assessment
and control algorithm for surface radiation measurements. The Open
Atmospheric Science Journal, 2(1), 23-37.
"""
mu0 = np.cos(np.radians(zenith))
# GHI reconstructed from BNI and DHI
ghi_calc = bni * mu0 + dhi
ghi_calc_safe = np.where(ghi_calc > 0, ghi_calc, np.nan)
# Condition: |GHI / (DNI * cos(SZA) + DIF) - 1| <= 0.15
diff_ratio = np.abs(ghi / ghi_calc_safe - 1)
# Domain: Z > 75 and GHI > 50
in_domain = (zenith > 75) & (ghi > 50)
condition_met = diff_ratio <= 0.15
if hasattr(in_domain, 'iloc'):
return (~in_domain) | condition_met
else:
return (not in_domain) or condition_met
