MEGS A & B SURF
Calibrations
Rachel Hock
LASP/CU
8 December 2009
8 December 2009
EVE Working Group Meeting
1
MEGS Calibration Algorithms
 C (i, j,t,TCCD ,tap)

C ′(i, j,t ) = 
− D(i, j,TCCD ,C,tap) ⋅ G(i, j,TCCD ,tap) ⋅ Mask (i, j,C,tap)
∆t


n
1 Ck′ (i, j,t, E beam , filter,α, β )
∑
ISURF (t )
n k=1
RSURF (i, j, E beam , filter,α , β ) =
FSURF (i, j, E beam ,α, β ) ⋅ Aslit ⋅ ∆λ(i, j )
Calculated from
SURF data


R flight (i, j, filter) =
⋅ Aslit ⋅ ∆λ (i, j ) ⋅ ∑ w (α, β ) ⋅ RSURF (i, j, E beam , filter,α , β ) ⋅ f OS (i, j, E beam , filter,α, β )
h⋅c
α ,β

λ (i, j )
I (i, j,t ) =
C ′(i, j,t, filter)
⋅ f degrad (i, j,t, filter) ⋅ f1AU (t )
R flight (i, j, filter)
8 December 2009
EVE Working Group Meeting
Equations used for
data processing
2
Correcting Raw Counts
 C (i, j,t,TCCD ,tap)

C ′(i, j,t ) = 
− D(i, j,TCCD ,C,tap) ⋅ G(i, j,TCCD ,tap) ⋅ Mask (i, j,C,tap)
∆t



  2  2
2 
C
σ
σ
2

  ⋅  C 2 + ∆t 2  + σ D


2
 ∆t   (C ) (∆t ) 
σG 
2 
2
+
σ C ′ = (C ′) ⋅ 
2
2


C
G
(
)


 − D

 ∆t




8 December 2009
EVE Working Group Meeting
3
Dark Correction


Electronic bias: average of 4 “virtual pixel” columns
Thermal dark: measured during thermal vacuum testing &
will be repeated once on orbit
Example of thermal dark for a
MEGS A pixel. Dashed line
shows the expected operating
temperature
8 December 2009
EVE Working Group Meeting
4
Gain Correction


Readout mode gain: normalize to default readout mode; 1
number for each half of CCD; will be recalculated on-orbit
Temperature gain: normalize counts to -85ºC
Gthermal (TCCD ) = a + b(TCCD + 85) + c (TCCD + 85)
2
Temperature Gain for Flight MEGS
Instrument
Flight MEGS A
Flight MEGS B
Rocket MEGS A
Rocket MEGS B
8 December 2009
EVE Working Group Meeting
Readout mode
Left (top half)
Right (top half)
Left (bottom half)
Right (bottom half)
Left (top half)
Right (top half)
a
1.028
1.046
1.068
1.044
0.904
0.842
b
3.363E-03
3.801E-03
3.869E-03
3.285E-03
4.422E-03
2.674E-03
c
3.572E-05
3.832E-05
3.612E-05
3.251E-05
6.526E-05
5.327E-05
Left (bottom half)
Right (bottom half)
Left (top half)
Right (top half)*
Left (bottom half)*
Right (bottom half)
Left (top half)
Right (top half)*
Left (bottom half)*
Right (bottom half)
0.774
0.814
1.007
1.000
1.000
1.003
0.977
1.000
1.000
0.940
9.350E-03
1.046E-02
8.288E-04
0.000E+00
0.000E+00
6.739E-04
1.171E-03
0.000E+00
0.000E+00
3.474E-04
1.409E-04
1.484E-04
8.826E-06
0.000E+00
0.000E+00
7.550E-06
1.705E-05
0.000E+00
0.000E+00
1.235E-05
5
Masking of Invalid Pixels

Exclude certain pixels





Saturated pixels
“Virtual pixel” columns
Cosmic rays
“Bad” pixels identified from flatfield images
Solar data: interpolate spatially & temporally to fill
the missing data
8 December 2009
EVE Working Group Meeting
6
SURF Responsivity
Reduces uncertainties
by a factor of 4-5
n
1 Ck′ (i, j,t, E beam , filter,α, β )
∑
ISURF (t )
n k=1
RSURF (i, j, E beam , filter,α , β ) =
FSURF (i, j, E beam ,α, β ) ⋅ Aslit ⋅ ∆λ(i, j )
σ R2
8 December 2009
SURF
2


n
σ I2SURF 
1  Ck′   σ C2 ′

 2 ∑

 ⋅
2
2 +
2
n k=1  ISURF  (Ck′ ) (ISURF )  σ F

2 
SURF
+
= (RSURF ) ⋅ 
2
2
n


F
( SURF ) 
C′
1



∑I


n

k=1 SURF 


EVE Working Group Meeting
7
SURF Flux

Inconsistencies


Fuzz on/off
Between difference beam energies in the OS data (discussed
later)
8 December 2009
EVE Working Group Meeting
8
Order Sorting Correction
k
1 FSURF (i, j, E beam ,α, β )
⋅ Rk (i, j, filter,α, β )
RSURF (i, j, E beam , filter,α , β ) = ∑ ⋅ k=1
k FSURF (i, j, E beam ,α, β )
k=1
m
 λ
k
FSURF
λ
=
F
( ) SURF  
k
fOS (i, j, E beam , filter,α, β ) =
8 December 2009
R1 (i, j, filter,α , β )
RSURF (i, j, E beam , filter,α , β )
EVE Working Group Meeting
9
Inconsistencies in SURF Fluxes (1)
MEGS A1 from Jan. 2009


For valid OS data, we expect no difference in RSURF at
short wavelengths.
Here, compared to 380 MeV, 331 & 285 MeV are “good”
energies.
8 December 2009
EVE Working Group Meeting
10
OS Results: MEGS A1 (rocket)
8 December 2009
EVE Working Group Meeting
11
OS Results: MEGS A2 (rocket)
8 December 2009
EVE Working Group Meeting
12
Inconsistencies in SURF Fluxes (2)
MEGS A1 from Aug. 2007


For invalid OS data, there are significant differences in
RSURF at short wavelengths.
Here, compared to 380 MeV, both 331 & 229 MeV are off
by up to 15%
8 December 2009
EVE Working Group Meeting
13
OS: MEGS A1 (flight)
8 December 2009
EVE Working Group Meeting
14
8 December 2009
MEGS B from Aug. 2007
MEGS A2 from Aug. 2007
Inconsistencies in SURF Fluxes (3)
EVE Working Group Meeting
15
OS: MEGS A2 (flight)
8 December 2009
EVE Working Group Meeting
16
Inconsistencies in SURF Fluxes (4)


For the Jan. 2009 SURF trip, we performed y-scans to
determine the beam center relative to 380 MeV.
These new values change the SURF flux enough to affect
the order sorting results.
8 December 2009
EVE Working Group Meeting
17
Flight Responsivity
Accounts for differences in
size of SURF beam & Sun


⋅ Aslit ⋅ ∆λ (i, j ) ⋅ ∑ w (α, β ) ⋅ RSURF (i, j, E beam , filter,α , β ) ⋅ f OS (i, j, E beam , filter,α, β )
R flight (i, j, filter) =
h⋅c
α ,β

λ (i, j )
Conversion to irradiance
from photons
σ R2
flight


∑ (w ⋅ RSURF ⋅
2

2
σ
α ,β
= (R flight ) ⋅  λ 2 +
 (λ )


8 December 2009
2
2
 σ2

σ
σ
R
f
SURF
OS

f OS ) ⋅ w 2 +
+
2
2
 (w ) (RSURF ) ( fOS ) 


2

∑ w ⋅ Rall ⋅ f OS 

 α ,β


2
EVE Working Group Meeting
18
Weighting of FOV Maps
Weights for expected on-orbit pointing
(αMEGS=0°,βMEGS=0°)
[°]
-1.0
+1.0 0.0000
+0.5
[°]
0.0
+0.5
0.0000
-1.0 0.0000
EVE Working Group Meeting
+1.0
0.0000
0.0249 0.1455 0.0249
0.0 0.0000 0.1455 0.3180 0.1455
-0.5
8 December 2009
-0.5
0.0000
0.0249 0.1455 0.0249
0.0000
0.0000
19
Rflight for Expected On-orbit Pointing
Flight MEGS A
8 December 2009
Flight MEGS B
EVE Working Group Meeting
20
Comparison to Rocket
Expect flight MEGS B to be more
sensitive at short wavelengths
8 December 2009
… but less sensitive at
longer wavelengths
MEGS A
MEGS B
Wavelength shift
Difference in CCDs
EVE Working Group Meeting
21
Solar Irradiance
I (i, j,t ) =
C ′(i, j,t, filter)
⋅ f degrad (i, j,t, filter) ⋅ f1AU (t )
R flight (i, j, filter)
Now calculated using
solar data
S (i, j, filter) =
What is in data
processing
1
R flight (i, j, filter)
I (i, j,t ) = C ′(i, j,t, filter) ⋅ S (i, j, filter) ⋅ f degrad (i, j,t, filter) ⋅ f1AU (t )

 2
2
2
2
σ
σ
σ
σ
f
R
f
flight
degrad
1AU

σ I2 = (I ) ⋅  C ′ 2 +
2 +
2 +
2
(C ′) (R ) (f
(f 1AU ) 
flight
degrad )

2
8 December 2009
EVE Working Group Meeting
22
April 2008 Solar Spectrum
Expect flight MEGS B to have
smaller uncertainties here by a
factor of ~2.
Expect flight MEGS B to have
higher uncertainties here by a
factor of ~2-3.
8 December 2009
EVE Working Group Meeting
23
Future SURF Calibrations



Continue taking at least 4 minutes of data per FOV point to
reduce uncertainties.
Continue taking temperature gain measurements.
Need to be careful of what SURF fluxes we use:




Avoid fuzz. We are able to make good MEGS B measurements at 183
MeV, which does not need 1 mm fuzz.
MEGS A OS: need only 380, 331, & 285 MeV. These have been shown
to be more reliable than lower energies and are sufficient to calculate
fOS.
MEGS B: using a higher beam energy (183 MeV or even 380 MeV). It
improves counts at the short wavelengths where we currently have
trouble. Plus, we can avoid using fuzz.
Performing y-scans at multiple energies may improve OS results for
MEGS A and explain the discrepancies for MEGS B.
8 December 2009
EVE Working Group Meeting
24