Quantifying the impact of the terminator on MIGHTI wind estimates¶
ICON Winds Working Group
27 Jan 2016
First orbit: Green¶
In [2]:
from IPython.display import Image
from IPython.core.display import HTML
Image(url='http://remote2.ece.illinois.edu/~bhardin2/MIGHTI/full_orbit_20160117.png')
Out[2]:
Quantifying terminator effect
- Depends strongly on how you define day/night
- Depends strongly on airglow model used (Are there any models for dawn/dusk airglow?)
- Assume red nighttime error will be resolved with brighter airglow model
- How to define "bad"?
- 68% of profile is okay?
- A single bad point?
- Recall that spec is on precision, not accuracy.
- What I did: Used 10 m/s as error threshold and picked bounds (to isolate from other error sources)
- Record duration as a longitude interval, not time interval
- Longitude/time not a one-to-one mapping. Different altitudes have different longitudes.
- Consider total error (though meridional should be affected less)
- Forward model used nearest-neighbor interpolation.
| First Orbit Green | Second Orbit Green | First Orbit Red | Second Orbit Red | |
|---|---|---|---|---|
| Day-to-Night Terminator Duration | 35.6$^\circ$ | 29.9$^\circ$ | 24.4$^\circ$ | 26.1$^\circ$ |
| Night-to-Day Terminator Duration | 14.8$^\circ$ | 18.0$^\circ$ | 13.0$^\circ$ | 9.0$^\circ$ |
| Percentage of orbit | 14.0% | 13.3% | 10.4% | 9.8% |
Second orbit: Green¶
In [3]:
from IPython.display import Image
from IPython.core.display import HTML
Image(url='http://remote2.ece.illinois.edu/~bhardin2/MIGHTI/full_orbit_20160119.png')
Out[3]:
First orbit: Red¶
In [4]:
from IPython.display import Image
from IPython.core.display import HTML
Image(url='http://remote2.ece.illinois.edu/~bhardin2/MIGHTI/full_orbit_20160118.png')
Out[4]:
Second orbit: Red¶
In [5]:
from IPython.display import Image
from IPython.core.display import HTML
Image(url='http://remote2.ece.illinois.edu/~bhardin2/MIGHTI/full_orbit_20160120.png')
Out[5]:
In [10]:
import subprocess
import shutil
nb_name = 'MIGHTI_Terminator_20160126.ipynb' # There's probably a way to get this programmatically
cmd = ['jupyter', 'nbconvert', '--to=slides', nb_name, \
'--reveal-prefix', '"http://cdn.bootcss.com/reveal.js/3.1.0"',\
'--template','output_toggle']
# The --template command points to a custom template which opens and closes the input code cells with a click
subprocess.call(cmd)
html_name = '%s.slides.html' % (nb_name.split('.')[0])
shutil.copy(html_name, '/home/bhardin2/public_html/slideshows/')
In [ ]: