def test_sunzen_corr(self):
"""Test Sun zenith angle correction.
"""
import datetime as dt
chan = Channel(name="test")
original_value = 10.0
chan.data = original_value * np.ones((2, 11))
lats = np.zeros((2, 11)) # equator
lons = np.array([np.linspace(-90, 90, 11), np.linspace(-90, 90, 11)])
# Equinox, so the Sun is at the equator
time_slot = dt.datetime(2014, 3, 20, 16, 57)
new_ch = chan.sunzen_corr(time_slot, lonlats=(lons, lats), limit=80.0)
# Test minimum after correction, accuracy of three decimals is enough
# self.assertTrue(np.abs(10.000 - np.min(new_ch.data)) < 10**-3)
self.assertAlmostEqual(10.000, np.min(new_ch.data), places=3)
# Test maximum after correction
self.assertAlmostEqual(57.588, np.max(new_ch.data), places=3)
# There should be ten values at zenith angle >= 80 deg, and
# these are all equal
self.assertTrue(np.where(new_ch.data == np.max(new_ch.data))[0].shape[0] == 10)
# All values should be larger than the starting values
self.assertTrue(np.all(new_ch.data > original_value))
# Channel name
self.assertEqual(new_ch.name, chan.name + "_SZC")
# Test channel name in the info dict
self.assertEqual(new_ch.name, chan.info["sun_zen_corrected"])
# Test with several locations and arbitrary data
chan = Channel(name="test2")
chan.data = np.array(
[
[0.0, 67.31614275, 49.96271995, 99.41046645, 29.08660989],
[87.61007584, 79.6683524, 53.20397351, 29.88260374, 62.33623915],
[60.49283004, 54.04267222, 32.72365906, 91.44995651, 32.27232955],
[63.71580638, 69.57673795, 7.63064373, 32.15683105, 9.05786335],
[65.61434337, 33.2317155, 18.77672384, 30.13527574, 23.22572904],
]
)
lons = np.array(
[
[116.28695847, 164.1125604, 40.77223701, -113.54699788, 133.15558442],
[-17.18990601, 75.17472034, 12.81618371, -40.75524952, 40.70898002],
[42.74662341, 164.05671859, -166.58469404, -58.16684483, -144.97963063],
[46.26303645, -167.48682034, 170.28131412, -17.80502488, -63.9031154],
[-107.14829679, -147.66665952, -0.75970554, 77.701768, -130.48677807],
]
)
lats = np.array(
[
[-51.53681682, -83.21762788, 5.91008672, 22.51730385, 66.83356427],
[82.78543163, 23.1529456, -7.16337152, -68.23118425, 28.72194953],
[31.03440852, 70.55322517, -83.61780288, 29.88413938, 25.7214828],
[-19.02517922, -19.20958728, -14.7825735, 22.66967876, 67.6089238],
[45.12202477, 61.79674149, 58.71037615, -62.04350423, 13.06405864],
]
)
time_slot = dt.datetime(1998, 8, 1, 10, 0)
# These are the expected results
results = np.array(
[
[0.0, 387.65821593, 51.74080022, 572.48205988, 138.96586013],
[227.24857818, 105.53045776, 62.24134162, 172.0870564, 64.12902666],
[63.08646652, 311.21934562, 188.44804188, 526.63931022, 185.84893885],
[82.86856236, 400.6764648, 43.9431259, 46.58056343, 36.04457644],
[377.85794388, 191.3738223, 27.55002934, 173.54213642, 133.75164285],
]
)
new_ch = chan.sunzen_corr(time_slot, lonlats=(lons, lats), limit=80.0)
self.assertAlmostEqual(np.max(results - new_ch.data), 0.000, places=3)
def test_sunzen_corr(self):
'''Test Sun zenith angle correction.
'''
import datetime as dt
chan = Channel(name='test')
original_value = 10.
chan.data = original_value * np.ones((2, 11))
lats = np.zeros((2, 11)) # equator
lons = np.array([np.linspace(-90, 90, 11), np.linspace(-90, 90, 11)])
# Equinox, so the Sun is at the equator
time_slot = dt.datetime(2014, 3, 20, 16, 57)
new_ch = chan.sunzen_corr(time_slot, lonlats=(lons, lats), limit=80.)
# Test minimum after correction, accuracy of three decimals is enough
#self.assertTrue(np.abs(10.000 - np.min(new_ch.data)) < 10**-3)
self.assertAlmostEqual(10.000, np.min(new_ch.data), places=3)
# Test maximum after correction
self.assertAlmostEqual(57.588, np.max(new_ch.data), places=3)
# There should be ten values at zenith angle >= 80 deg, and
# these are all equal
self.assertTrue(np.where(new_ch.data == \
np.max(new_ch.data))[0].shape[0] == 10)
# All values should be larger than the starting values
self.assertTrue(np.all(new_ch.data > original_value))
# Channel name
self.assertEqual(new_ch.name, chan.name + '_SZC')
# Test channel name in the info dict
self.assertEqual(new_ch.name, chan.info['sun_zen_corrected'])
# Test with several locations and arbitrary data
chan = Channel(name='test2')
chan.data = np.array(
[[0., 67.31614275, 49.96271995, 99.41046645, 29.08660989],
[87.61007584, 79.6683524, 53.20397351, 29.88260374, 62.33623915],
[60.49283004, 54.04267222, 32.72365906, 91.44995651, 32.27232955],
[63.71580638, 69.57673795, 7.63064373, 32.15683105, 9.05786335],
[65.61434337, 33.2317155, 18.77672384, 30.13527574, 23.22572904]])
lons = np.array([[
116.28695847, 164.1125604, 40.77223701, -113.54699788, 133.15558442
], [-17.18990601, 75.17472034, 12.81618371, -40.75524952, 40.70898002],
[
42.74662341, 164.05671859, -166.58469404,
-58.16684483, -144.97963063
],
[
46.26303645, -167.48682034, 170.28131412,
-17.80502488, -63.9031154
],
[
-107.14829679, -147.66665952, -0.75970554,
77.701768, -130.48677807
]])
lats = np.array([
[-51.53681682, -83.21762788, 5.91008672, 22.51730385, 66.83356427],
[82.78543163, 23.1529456, -7.16337152, -68.23118425, 28.72194953],
[31.03440852, 70.55322517, -83.61780288, 29.88413938, 25.7214828],
[-19.02517922, -19.20958728, -14.7825735, 22.66967876, 67.6089238],
[45.12202477, 61.79674149, 58.71037615, -62.04350423, 13.06405864]
])
time_slot = dt.datetime(1998, 8, 1, 10, 0)
# These are the expected results
results = np.array([[
0., 387.65821593, 51.74080022, 572.48205988, 138.96586013
], [
227.24857818, 105.53045776, 62.24134162, 172.0870564, 64.12902666
], [
63.08646652, 311.21934562, 188.44804188, 526.63931022, 185.84893885
], [82.86856236, 400.6764648, 43.9431259, 46.58056343, 36.04457644],
[
377.85794388, 191.3738223, 27.55002934,
173.54213642, 133.75164285
]])
new_ch = chan.sunzen_corr(time_slot, lonlats=(lons, lats), limit=80.)
self.assertAlmostEqual(np.max(results - new_ch.data), 0.000, places=3)
