Transcript An Assessment of the NRLMSISE-00 Density Thermosphere

An assessment of the NRLMSISE-00 density
thermosphere description in presence of
space weather events
C. Lathuillère and M. Menvielle
•
•
•
•
The data and the method
Statistical analysis for year 2004
8 larger events of 2004
Conclusion
The data
to the
sun
STAR atmosphere density
along CHAMP trajectory at
about 400 km altitude
Inclination:87°
STAR/CHAMP densities – May 3rd, 2003
… between -50° and 50° in
latitude, with a 1° sampling
rate with respect to the
latitude
daytime and night time orbit
sections are considered
separately
Densities (10-15 g.cm-3)
Densities (10-15 g.cm-3)
STAR/CHAMP densities – May 3rd, 2003
~5:30 pm LT at equator
~5:30 am LT at equator
The method:
10 LT
Running SVD analysis over 15 consecutive orbits (about 1 day)
TOTAL MASS DENSITY at the satellite altitude (10-15 g/cm3)
22 LT
The projection on the first component accounts for large scale variations: spatial variations
are captured by the first principal component, and time variations are captured by the
associated projection coefficient: C1.
Residuals account for smaller scales, as tides and gravity waves
Comparison with NRLMSISE-00 model
CHAMP data
NRLMSIS:
with MgII proxy and ap
NRLMSIS_quiet:
with MgII proxy and Ap=4
Normalized coefficient:
C1 CHAMP / C1 NRLMSIS_quiet
C1 NRLMSIS/ C1 NRLMSIS_quiet
MgII proxy: The composite MgII index (Viereck et al, 2004) is used as a proxy for solar EUV instead of F10.7
NRLMSIS_quiet is used as a reference, that accounts for LT,
seasonal variations, solar activity…
- Statistical analysis: binning of normalized C1 coefficients as a function of ap
- Analysis of the height larger events
Night time
Density perturbation
Day time
ap index
x : CHAMP data
x : NRLMSIS model
ap index
correlation coefficient
between CHAMP data
and ap
Binning is done using the value of ap from the previous 3 hour interval
x Day time
+ Night time
x Day time
+ Night time
Linear fit
y =1.9 x – 0.9
x Day time
+ Night time
Day and Night time
with ap <100
x Day time
+ Night time
Day and Night
with ap <100
Quadratic fit
y =1.92 x2 – 3 x + 2.14
Density perturbation: CHAMP data and NRLMSIS model
Day time
Night time
Day of 2004
3 days
ap magnetic index
- the relative density increase is greatly underestimated by the model
- the model seems to correctly represent the shape of the perturbation
Density perturbation: CHAMP data and NRLMSIS model
normalized to the amplitude of the data perturbation
Normalization factor mean value = 2.3
3 days
The perturbations appear later in the data than in the model and this
timing discrepancy is slightly larger during night ( 3-4.5 hours) than
during day time.
Conclusion
Altitudes about 400km. Latitudes between 50S and 50N
• NRLMSISE-00 correctly estimates the main features of the
thermosphere density response to geomagnetic activity:
- the morphology of UT variations
- the larger relative increase during night than during day time
But it underestimates :
• the amplitude of the density response (by about a factor 2)
• and its phase lag (up to 4.5 hours)