Data analysis Page 1

60 p
1. Photometry and spectroscopy of Nova Del 2013

a) From the light curve plot the Modified Julian Dates can be read with an error of about 0.5. The
peak of the maximum brightness is very narrow and obviously placed at MJD0 = 56 520.5 with a
value of m0 = 4.5m , therefore:
MJD0 = 56 520.5 ± 0.5 (2 p)

m0 = 4.5m ± 0.05m (1 p)
b) The brightness values of 2m and 3m decline are 6.5m ± 0.05m and 7.5m ± 0.05m , respectively. (2 p)
The corresponding Modified Julian Dates are MJD2 and MJD3 . Because of the poorly defined slopes
on the light curve around these dates, their acceptable error is larger than in other parts of the light
curve, let say it is about 1d , so:
MJD2 = 56 531.5 ± 1 , MJD3 = 56 543.5 ± 1 (2 p)

t2 = 11d ± 1d , t3 = 23d ± 1d (2 p)
c) The text of this part does not ask for calculating the individual errors of the formulae, but it is worth
estimating them here, just for the sake of completeness. (Students won’t do it.)
(a) The form of the function is
c − log t2
(1.1) M = a + b arctan , a = −7.92, b = −0.81, c = 1.32, d = 0.23,
d
so its derivative:
b 1 b ∆t2
(1.2) M 0 = − ñ ã2 ô → ∆M = − ñ ã2 ô
c − log t2 t2 c − log t2 t2
Å Å
d log(10) 1 + d log(10) 1 +
d d
The value and its error calculated from the formulae above (error is not necessary):
(a)
M0 = −8.63m ± 0.06m (1 p)
(b) The form of the function is
(1.3) M = a + b log t2 , a = −11.32, b = 2.55
so its derivative:
b b ∆t2
(1.4) M 0 = → ∆M =
t2 log(10) log(10) t2
The value and its error calculated from the formulae above (error is not necessary):
(b)
M0 = −8.66m ± 0.10m (1 p)
(c) The form of the function is
(1.5) M = a + b log t3 , a = −11.99, b = 2.54
so its derivative:
b b ∆t3
(1.6) M 0 = → ∆M =
t3 log(10) log(10) t3
The value and its error calculated from the formulae above (error is not necessary):
(c)
M0 = −8.53m ± 0.05m (1 p)
 Data analysis Page 2

The standard deviation of a dataset can be calculated as

Õ
Pn
(xi − x̄)2
i=1
(1.7) σ = ,
n−1

where n is the number of data points, xi is the ith individual data value and x̄ is their mean,
x̄ = (x1 + x2 + . . . + xn )/n.
The mean and the standard deviation of the three absolute maximum brightness values:

M0 = −8.61m ± 0.07m (2 p)

d) The color excess E(B − V ) is the difference between the observed color index of the star and the
intrinsic color index predicted from its spectral type:

(1.8) E(B − V ) = (B − V ) − (B − V )0 = AB − AV

The total extinction is quantified by AV (at 5550 Å). The ratio of total-to-selective extinction:

AV
(1.9) R = → AV = RE(B − V ), where R = 3.1 (2 p)
E(B − V )

With the given value and error of E(B − V ):

AV = 3.1 × E(B − V ) = 3.1 × (0.184m ± 0.035m ) = 0.57m ± 0.11m (2 p)

e) According to the formula for the distance modulus:

(1.10) mV − MV = −5 + 5 log d + AV → (1 p)
mV − MV + 5 − AV
(1.11) log d = →
5
(1.12) d = 10(mV −MV +5−AV )/5 , (2 p)

where the distance d is in parsecs.

Since ∆ax /∆x = ax ln a, therefore the error of the distance d:

∆d = 10(mV −MV +5−AV )/5 × ln 10 × ∆((mV − MV + 5 − AV )/5) (2 p)

The error of (mV − MV + 5 − AV )/5 can be estimated with the sum of the errors of mV , MV and
AV , so:

∆((mV − MV + 5 − AV )/5) ≈ 0.05 (2 p)

With the data:

d = 3220 pc and ∆d = 338 pc, (2 p)

so the distance to the nova:

d ≈ 3.2 ± 0.3 kpc (2 p)

 Data analysis Page 3

f) The well known Doppler formula between the wavelength displacement and radial velocity:
λ − λ0 ∆λ vr ∆λ
(1.13) = = → vr = c, (1 p)
λ0 λ0 c λ0
where λ is the measured wavelength of the line feature, λ0 is the rest wavelength of the line, c is the
speed of light, and vr is the radial velocity to be calculated.
The wavelength of the P Cygni absorption peak should be extracted from the figure with an error
of about 1 Å. The main part of this error is coming from the "definition" of the peak position of the
Gaussian-like profile. This will result in inaccuracy of about ∆vr = ±50 km s−1 in radial velocities.
(1 p)
The wavelengths and radial velocities should be something like these:

MJD WL RV
56518.986 6527 -1636
56519.813 6531 -1454
56520.843 6534 -1317
56521.835 6537 -1179
56522.829 6542 -951
56523.827 6544 -860

6 points for the wavelength values and 6 points for the radial velocity values. (12 p)
Radial velocities within the range of ±50 km s−1 of the RV values listed in the table should be given
full marks, but velocities in the range of ±100 km s−1 are still acceptable with half marks.

g) See the attached figure as an example for the acceptable solution. The plotted data are taken from
the table above. For the sake of simplicity the absolute values of the radial velocities have been used
for making the graph. (6 p)

h) It is obvious from the plot, that the radial velocities lie along a straight line.
To estimate the size of the expanding envelope we need to calculate the area below the t − vr graph
between the first and last date.
Hence the graph is a straight line, this is very simple: we have to determine the area of the hatched
region which is a trapezoid.
If the two bases and the height of the trapezoid are a, b, and m, respectively, then the area of the
trapezoid is:
a+b
(1.14) T = m (3 p)
2
In our case a = vr1 , b = vr6 , and m = t6 − t1 . (1 p)
We could use the fitted line (dashed) as the upper side of the trapezoid, but this would be a bit
complicated – because of the difficulties of the fitting process –, and not necessary at all. Instead of
this we use the line connecting the first and last radial velocity points as this runs very close to the
fitted line.
The result: R ≈ 3.5 AU (3 p)

i) The apparent angular diameter of the spherical envelope seen from the Earth:
R R
Å ã
(1.15) ϑ = 2 × arctan ≈2 (2 p)
d d
 Data analysis Page 4

Using the values of d ≈ 3.2 kpc and R ≈ 3.5 AU, 5 days after the discovery the angular diameter of
the envelope is:

ϑ = 0.002200 = 2.2 mas (2 p)

A less formal solution:

• By definition a parsec (1 pc) is the distance from the Sun to an astronomical object that has
a parallax angle of one arcsecond, i.e. it represents the distance at which the radius of Earth’s
orbit (1 AU) subtends an angle of one arcsecond.
• Because of the very small angles the distance is a linear function of the parallax. This means
that the radius R ≈ 3.5 AU of the envelope subtends one arcsecond viewing from a distance of
d ≈ 3.5 pc, and one milliarcsecond from a distance of 1000d ≈ 3500 pc = 3.5 kpc.
• Since this value is close to the distance of Nova Del 2013 determined earlier, we can conclude
that the apparent angular diameter of the spherical shape envelope 5 days after the discovery
was about 2 milliarcseconds.
 Nova Del 2013 – Johnson V light curve
4.0

m0,V = 4.5

5.0

6.0

m2,V = 6.5

7.0

m3,V = 7.5

8.0
MJD0,V = 56 520.5

mV (mag)
t2,V = 11
Data analysis

9.0

MJD2,V = 56 531.5

10.0
t3,V = 23

11.0
MJD3,V = 56 543.5

12.0
56510 56520 56530 56540 56550 56560 56570 56580 56590 56600 56610
MJD (days)
Page 5
 Nova Del 2013 – Spectra around Hα line
1.0

0.9

0.8

0.7

0.6

t0 − 2d , MJD = 56518.986
0.5

t0 − 1d , MJD = 56519.813
0.4

relative flux + constant

Data analysis

t0 + 0d , MJD = 56520.843
0.3

t0 + 1d , MJD = 56521.835
0.2

t0 + 2d , MJD = 56522.829
0.1

t0 + 3d , MJD = 56523.827
0.0
6450 6470 6490 6510 6530 6550 6570 6590 6610 6630 6650 6670 6690 6710 6730
λ (Å)
Page 6
 Nova Del 2013 – Radial velocities of Hα absorption components

(vr1 + vr6 )
R=
2
|vr | dt ≈ (t6 − t1 ) ≈ 3.5 AU
Zt6

t1
1500

1000

|vr | [km/s]
Data analysis

1248 km/s

500

0
4.841 days = 418 262 s

56519 56520 56521 56522 56523 56524

MJD (days)
Page 7
 Data analysis Page 8

90 p
2. Triply eclipsing hierarchical triple stellar system

i) a) See the following table. (10 p)

event no. contacts components BJD ϕ1 ϕ2

1 I A, B 2455476.1096 0.1226 0.9747
II A, C 2455476.4245 0.4703 0.9816
III A, B 2455477.9677 0.1743 0.0155
IV A, B 2455478.4722 0.7313 0.0266
2 I A, B 2455521.5217 0.2643 0.9734
3 III A, C 2455568.9434 0.6248 0.0163
4 I A, C 2455612.4733 0.6882 0.9736
III A, C 2455614.3571 0.7682 0.0150
5 III A, B 2455659.9241 0.0808 0.0171
IV A, C 2455660.2422 0.4320 0.0241

b) See the following table. (5 p)

event no. closer component

1 A
2 A
3 A
4 A
5 A

All these events occur close to ϕ2 = 0 (or ϕ2 = 1) which means by definition that star A eclipses
stars B and C. Therefore, star A is closer to the observer.
For full mark there is no need for explanation. 1 point for each event with correct answer.
c) In the moments of the 1st and last (4th) contacts:

(2.1) RA + RB,C = dA–B,C , (1 p)

while in the case of the 2nd and 3rd (inner) contacts:

(2.2) RA − RB,C = dA–B,C , (1 p)

where dA–B,C stands for the sky-projected distance of the disks of the occulting star A and
occulted component B or C.
Let ~r1 the radius vector directed from star B toward star C, and ~r2 another radius vector
directed from the centre of mass of stars B and C toward star A.
By the use of these two (Jacobian) vectors, the position vectors connecting stars B and C with
star A, can be written as:
mC
(2.3) ~rBA = ~rA − ~rB = ~r2 + ~r1 , (1 p)
mBC

mB
(2.4) ~rCA = ~rA − ~rC = ~r2 − ~r1 . (1 p)
mBC

Using the facts that (irel = 0°) and, furthermore, i1 = i2 = 90°, it is worthy to introduce the
orbital plane as reference frame. Let the origin of our frame of reference be the centre of mass
 Data analysis Page 9

of the close binary (denoted as O2 in figure above). Furthermore, let axis X directed toward
the line of sight. Therefore, axis Y is located in the tangential plane of the sky.
In this frame of reference, the components of vectors ~r1 and ~r2 can simply be written as

(2.5) ~r1 = a1 [cos(φ1 ); sin(φ1 )] , (1 p)

(2.6) ~r2 = a2 [cos(φ2 ); sin(φ2 )] , (1 p)

as φ1 takes the values of 2kπ in the moments, when star C eclipses star B, and similarly,
φ2 = 2kπ when star A "eclipses" the centre of mass of the inner binary.
One can also notice that there is a simple relation between these position angles and the
photometric phases defined above as

(2.7) φ1,2 = 2πϕ1,2 (1 p)

We are interested in the sky-projected distances of the stellar disks, which, in this frame of
reference, are equal to the y components of the vector equations (2.3 and 2.4). Accordingly,
 ï ò
 m C a1 

(2.8) dA–B = a2 sin(2πϕ2 ) + sin(2πϕ1 )  , (1 p)
mBC a2
 ï ò
 mB a1 

(2.9) dA–C = a2 sin(2πϕ2 ) − sin(2πϕ1 )  . (1 p)
mBC a2
From this point we can use different combinations of eclipsing events given in the table. There-
fore, here we use one possible order only for illustration.
 Data analysis Page 10

We notice that both outer contacts (i.e. contacts I and IV) of the first event stars A and B are
involved. Therefore, we can write the same equation for both contacts as follows:
 
RA + RB  mC a1 
(2.10) = sin(2πϕ2 ) + sin(2πϕ1 ) . (1 p)
a2 mBC a2
Therefore, we have two independent equations and two unknown variables, namely
RA + RB m C a1 aB
(2.11) , and = . (2 p)
a2 mBC a2 a2

The computation is as follows:

For the 1st contact sin 2πϕ2 = −0.158 296, sin 2πϕ1 = 0.696 364, while for the 4th one sin 2πϕ2 =
0.166 356, sin 2πϕ1 = −0.993 105. Consequently
aB aB
(2.12) 0.158 296 − 0.696 364 = 0.166 356 − 0.993 105 , (1 p)
a2 a2

aB
(2.13) = 0.027 161 (1 p)
a2
and, therefore,
RA + RB
(2.14) = 0.1394. (1 p)
a2
Now, considering the also available 3rd contact moment of the same event (in which similarly,
stars A and B play the roles), one can get

RA − RB
(2.15) = |sin(2π × 0.0155) + 0.027 161 × sin(2π × 0.1743)| (1 p)
a2

RA − RB
(2.16) = 0.1214 (1 p)
a2
Combining these results, we obtain that
RA 1 RA + RB RA − RB
Å ã
(2.17) = + = 0.1304 (2 p)
a2 2 a2 a2

RB 1 RA + RB RA − RB
Å ã
(2.18) = − = 0.0090 (2 p)
a2 2 a2 a2
We repeat a very similar calculation for the 3rd contacts of events 3 and 4, in which cases star
B is substituted with star C.
For the 3rd contact of the third event sin 2πϕ2 = 0.102 237, sin 2πϕ1 = −0.706 218, while the
3rd contact of the fourth event sin 2πϕ2 = 0.094 108, sin 2πϕ1 = −0.993 469.
Therefore,
aC aC
(2.19) 0.102 237 + 0.706 218 = 0.094 108 + 0.993 469 (1 p)
a2 a2

aC
(2.20) = 0.028 298 (1 p)
a2
 Data analysis Page 11

and, accordingly,
RA − RC
(2.21) = 0.1222 (1 p)
a2
Taking into account that RA /a2 is already known from the previous stage, the dimensionless
relative radius of star C can be obtained simply as
RC RA RA − RC
(2.22) = − = 0.1304 − 0.1222 = 0.0082 (2 p)
a2 a2 a2
The other possibility is, however, that e.g. from the 1st contact of event 4 we calculate the sum
(RA + RC )/a2 , as

RA + RC
(2.23) = |sin(2π × 0.9736) − 0.028 298 × sin(2π × 0.6882)|
a2

RA + RC
(2.24) = 0.1389,
a2
and then we obtain that
RA 1 RA + RC RA − RC
Å ã
(2.25) = + = 0.1306,
a2 2 a2 a2

RC 1 RA + RC RA − RC
Å ã
(2.26) = − = 0.0084.
a2 2 a2 a2

********** Several additional, equivalently acceptable scenarios **********

We can write three equations for (RA + RB )/a2 :

RA + RB aB
= 0.158 296 − 0.696 364 (1st event Ist contact)
a2 a2
RA + RB aB
= 0.166 356 − 0.993 105 (1IV)
a2 a2
RA + RB aB
= 0.166 433 − 0.995 966 (2I)
a2 a2
Combining them one can get:

contacts aB /a2 (RA + RB )/a2

1I–1IV 0.027 161 0.1394
1I–2I 0.027 159 0.1394
1IV–2I 0.026 988 0.1396

Similarly, for (RA − RB )/a2 :

RA − RB aB
= 0.097 235 + 0.889 001 (1III)
a2 a2
RA − RB aB
= 0.107 236 + 0.486 152 (5III)
a2 a2
Combining them one can get:
 Data analysis Page 12

contacts aB /a2 (RA − RB )/a2

1III–5III 0.024 826 0.1193

Now for (RA + RC )/a2 :

RA + RC aC
= 0.165 116 − 0.925 553 (4I)
a2 a2
RA + RC aC
= 0.150 847 − 0.414 376 (5IV)
a2 a2
The only combination gives:

contacts aC /a2 (RA + RC )/a2

4I–5IV 0.027 914 0.1393

Finally, for (RA − RC )/a2 :

RA − RC aC
= 0.115 353 + 0.185 529 (1II)
a2 a2
RA − RC aC
= 0.102 237 + 0.706 218 (3III)
a2 a2
RA − RC aC
= 0.094 108 + 0.993 469 (4III)
a2 a2
The combinations are:
contacts aC /a2 (RA − RC )/a2
1II–3III 0.025 190 0.1200
1II–4III 0.026 295 0.1202
3III–4III 0.028 299 0.1222
********** End additional, equivalently acceptable scenarios **********
Now, we are in the position to calculate the inner mass ratio q1 , as follows:

aB /a2 0.027 161

(2.27) q1 = = = 0.9598 (2 p)
aC /a2 0.028 298

Furthermore, we can also get the ratio of the semi-major axes as

a1 aB aC
(2.28) = + = 0.027 161 + 0.028 298 = 0.055 459 (2 p)
a2 a2 a2
The requested results with the range of the full marks (summary):

RA
0.1150 ≤ ≤ 0.1450
a2
RB
0.0075 ≤ ≤ 0.0105
a2
RC
0.0070 ≤ ≤ 0.0100
a2
a1
0.045 ≤ ≤ 0.065
a2
0.80 ≤ q1 ≤ 1.25
 Data analysis Page 13

d) The quickest method to obtain q2 comes with the double use of Kepler’s third law. Writing this
law for both the inner and outer orbits, and dividing them one can get, that
Å ã3 Å ã2
a1 P2 mBC q2
(2.29) = = . (4 p)
a2 P1 mABC 1 + q2
Therefore
a1 3 P 2 2
Å
ã Å ã

a2 P1
(2.30) q2 = Å ã3 Å ã2 (2 p)
a1 P2
1−
a2 P1

0.055 4593 × 50.206 7632

(2.31) q2 ≈ ≈ 0.7543 (2 p)
1 − 0.055 4593 × 50.206 7632

ii) The circular velocity of star A is

2π
(2.32) vA = rA . (1 p)
P2
The amplitude of the RV curve is equal to maximum value of the line-of-sight component of that
velocity, i.e.

(2.33) KA = vA sin i2 . (1 p)

The sinusoidal component in the occurrence of the eclipsing minima times comes from the light-
travel time effect caused by the revolution of stars B and C around the centre of mass of the whole
triple system. In this regard the motion of components B and C can simply be substituted by the
movement of their centre of mass along the outer orbit. Therefore, the total variation of the inner
binary’s distance to the Earth is the line-of-sight component of the diameter of the orbit of the centre
of mass of the inner binary around the centre of mass of the complete triple system, i.e.

(2.34) ∆zBC = 2rBC sin i2 . (2 p)

Therefore, the amplitude of the light-travel time sine wave is

∆z rBC sin i2 cAETV
(2.35) AETV = = → rBC = . (2 p)
2c c sin i2
Such a way, using that
mBC rA
(2.36) q2 = = , (1 p)
mA rBC
one can obtain, that
P2 K A
(2.37) q2 = . (1 p)
2πcAETV
which gives a second, independent determination of the outer mass ratio q2 .
The uncertainty can be estimated either as
 Å
∆P2 2 ∆KA 2 ∆AETV 2
ã Å ã Å ã
(2.38) ∆q2 = q2 + +
P2 KA AETV
 Data analysis Page 14

or
Å     ã
 ∆P2   ∆KA   ∆AETV 
(2.39) ∆q2 = q2  + +  (2 p)
P2   KA   AETV 
With numerical values:

q2 = 0.621 ± 0.047 or q2 = 0.621 ± 0.048 (1 p)

Then, the semi-major axis of the outer orbit can be calculated in the following alternative ways:
1 + q2 P2 K A 1 + q 2
(2.40) a2 = rA =
q2 2π sin i2 q2

AETV
(2.41) a2 = rBC (1 + q2 ) = c (1 + q2 )
sin i2

P 2 KA AETV
(2.42) a2 = rA + rBC = +c ,
2π sin i2 sin i2
One of the three equations above: (2 p)
We assumed that i2 = 90°, therefore sin i2 = 1. Then

a2 = (60.712 ± 2.851) × 106 km = (87.293 ± 4.099) R = (0.406 ± 0.019) AU (2 p)

The uncertainties from the three different formulae above:

 Å
∆P2 2 ∆KA 2
ã2
∆q2
ã Å ã Å
(2.43) ∆a2 = a2 + +
P2 KA q2 (1 + q2 )

∆a2 ≈ 2.851 × 106 km ≈ 4.10 R ≈ 0.019 AU

 Å ã2 ã2
∆AETV ∆q2
Å
(2.44) ∆a2 = a2 +
AETV 1 + q2

∆a2 ≈ 4.946 × 106 km ≈ 7.11 R ≈ 0.033 AU

 Å ã2 ã2
∆P2 P2
Å
(2.45) ∆a2 = KA + ∆KA + (c∆AETV )2
2π 2π

∆a2 ≈ 2.849 × 106 km ≈ 4.10 R ≈ 0.019 AU

One of the three formulae above with its uncertainty: (3 p)
In what follows, we use the smallest uncertainties (first row above) but the others, and also the
nonquadratic ones are acceptable, too.
The total mass of the triple now can be calculated directly from Kepler’s third law or, as an alternative
way, one can obtain it from the equations on the centripetal accelerations.

4π 2 GmBC
(2.46) rA 2
=
P2 a22
 Data analysis Page 15

4π 2 GmA
(2.47) rBC 2
= ,
P2 a22

where the only unknown quantities are the masses, so they can be calculated separately from the
two equations or, summing the masses we can get back (formally) Kepler’s third law:

4π 2 a32
(2.48) mABC = = (4.312 ± 0.607) M , (2 p)
G P22

from which
mABC
(2.49) mA = = (2.660 ± 0.383) M (1 p)
1 + q2
q2
(2.50) mBC = mABC = (1.652 ± 0.245) M (1 p)
1 + q2

iii) We obtained the total mass (mBC ) of the inner binary in the second part of the problem. Combining
this with the inner mass ratio (q1 ) obtained in the first part, one can get that

mBC 1.652 M
(2.51) mB = = = 0.843 M (3 p)
1 + q1 1 + 0.960
and

(2.52) mC = q1 × mB = 0.960 × 0.843 M = 0.809 M (3 p)

The dimensionless radius of each star relative to the semi-major axis was calculated in the first
part, while the semi-major axis of the outer orbit (a2 ) was obtained in the second part. Their
multiplication gives the physical radii of the stars:

(2.53) RA = 0.1304 × 87.29 R = 11.381 R (3 p)

(2.54) RB = 0.0090 × 87.29 R = 0.786 R (3 p)

(2.55) RC = 0.0082 × 87.29 R = 0.712 R (3 p)