Up: Solar system objects in survey
Subsections
5 Scientific results
The procedure from Sect. 3 resulted in a list of
potential SSO candidates in the ISOSS. Mainly for the following
reasons, not all of them were visible in the slew data:
- 1.)
- The structured and bright cirrus background caused
source confusion and limited the point source extraction
.
It also affected strongly the analysis of extended structures,
like from a cometary coma.
- 2.)
- The sensitivity limit of approximately 1 Jy at 170
m
allowed only the detection of bright asteroids and comets,
which are already well known through other observing programmes
and techniques (IRAS, occultation measurements, radar, ...)
and through dedicated ISO measurements.
- 3.)
- The source extraction from a 4-pixel camera is difficult due
to the high slewing speed and the variety of impact parameters.
The resulting fluxes or flux limits had usually larger error bars
than comparable pointed observations, where sources were usually
either centred in the C200 array or on a single pixel.
- 4.)
- For some faint sources the allowed maximal offset of 5
was too large to produce a noticable signal increase.
In some cases the object was visible, but a reliable flux determination
from the ISOSS was not possible. In these cases upper or lower limits
are given.
Tables 4, 5 and 6
include all SSO predictions which are within 5
of
the slew centre, fulfill the flux requirements and have not been used
in Sect. 4.
For the scientific comparison between ISOSS fluxes and model predictions
we did the following calibration steps: We determined the ISOSS calibrated
inband fluxes through the
different methods and corrected them by an
estimated factor based on the slopes visible in Figs. 4
and 5. As a last step we applied the colour correction to
obtain monochromatic flux densities at 170
m
(ISOSS values in Tables 4, 5).
5.1 Planets
The inner planets were not visible for ISO due to pointing
constraints. Mars, Jupiter and Saturn exceeded the saturation
limits, Pluto was below the 1.0 Jy limit. Therefore, only
Uranus and Neptune were seen, but were already used to extend the ISOSS
calibration (Sects. 4.1, 4.2
and 4.3). However, the bright planets had to be included
in the search programme with the objective to identify close-by slews.
For the extremely IR bright planets, diffraction effects of the optical
system in combination with certain satellite-planet constellations produced
bright spots, spikes and ring like structures around the planets,
which are visible in the slew data. Mars, Jupiter and Saturn, with
170
m brightnesses between 10 000 and 400 000 Jy, influenced
slews up to 1
distance, Uranus and Neptune (>200 Jy)
up to 10
distance. These slews have been identified
(with the above mentioned SSO extraction method in combination with a large
search radius) and the planet influence can now be taken into
account for further scientific catalogues based on ISOSS.
Some planetary satellites are bright enough to be visible in principle.
However, close to Jupiter (the maximal distance for the Galilean
satellites is about 11
)
and Saturn (the maximal angular
distance for the 8 largest satellites is about 10
)
no
170
m fluxes can be derived, due to the strong planet influences.
The measured Uranus and Neptune flux values from
Tables 1, 2 and 3
can also be used as input to future models of planetary atmospheres.
Current models are based on Voyager IRIS data from 25 to 50
m
and sub-millimetre data beyond 350
m (Griffin & Orton
1993) with an interpolation in between. The ISOSS
observations make this wavelength range directly accessible for
model tests in the far-IR.
5.2 Asteroids
After establishing new methods for the flux calibration, based now
additionally on measurements of Uranus, Neptune,
Ceres, Pallas, Juno and Vesta,
the monochromatic flux densities at 170
m of the remaining
asteroids were derived. 23 out of the 56 asteroid "hits''
were already used in this calibration context in the previous
Sect. 4. The remaining 33 hits can be split in 3 groups:
18 asteroid predictions have reliable IRAS detections,
but only low quality ISOSS detections. The reasons for the poor ISOSS
fluxes are manifold: slew offsets, bright backgrounds,
technical problems, low fluxes, etc.
For these 18 asteroids Tedesco et al. (1992)
calculated already diameters and albedos, based on IRAS
observations. Upper flux limits from ISOSS would therefore
not give any new information.
Table 4:
ISOSS observational results for asteroids with IRAS
detections. Note, that the ISOSS results were first flux
corrected according to Sect. 4 and then colour
corrected.The uncertainties in the table, given in brackets for
method 3a, are statistical errors of weighted results from
all 4 pixels.
TDT |
Date/Time |
SSO |
 |
 |
Method |
Slew |
Remarks |
No. |
|
|
(Jy) |
(Jy) |
|
speed |
|
(1) |
(2) |
(3) |
(4) |
(5) |
(6) |
(7) |
(8) |
11080300 |
06-Mar.-96 11:42:00 |
(5) Astraea |
<4 |
1.3 |
2 |
fast |
upper limit, cirrus bgd. |
12780300 |
23-Mar.-96 02:50:17 |
(5) Astraea |
<4 |
1.6 |
2 |
fast |
upper limit, cirrus bgd. |
85480200 |
18-Mar.-98 13:48:12 |
(7) Iris |
<3 |
2.1 |
2 |
fast |
in cirrus knot |
25381400 |
27-Jul.-96 03:22:14 |
(15) Eunomia |
2.9 |
2.5 |
1 |
fast |
ok |
85480300 |
18-Mar.-98 17:22:43 |
(89) Julia |
<3 |
1.7 |
2 |
fast |
upper limit |
18780100 |
22-May-96 05:40:00 |
(344) Desiderata |
4.5 |
5.8 |
1 |
fast |
ok |
18280800 |
17-May-96 03:50:00 |
(532) Herculina |
5.3(1.0) |
5.5 |
3a |
stop |
ok |
21780900 |
21-Jun.-96 06:01:48 |
(532) Herculina |
3.4(0.3) |
3.8 |
3a |
stop |
ok |
83380500 |
25-Feb.-98 11:36:26 |
(1036) Ganymed |
<4 |
0.1 |
2 |
fast |
upper limit |
7 asteroids (9 hits) have reliable IRAS
detections and also good quality ISOSS detections (see
Table 4).
For these asteroids we could successfully derive
fluxes and upper limits with our newly established
calibration, based on clear detections. The IRAS diameter
and albedo values in the following are all taken from the
Minor Planet Survey (MPS, Tedesco et al. 1992).
The main-belt asteroid Astraea has been well observed by
different techniques, including IRAS, radar and lightcurve
observation. The combination of all measurements led to
the description of the object as a rotating ellipsoid with a well
determined spin vector (Erikson 2000) and axis dimensions of
km
(Magri et al. 1999).
Using the TPM with default thermal parameters for main-belt
asteroids (Müller et al. 1999) together with the
shape, size and spin vector information gave fluxes of
1.3
0.3 Jy and 1.6
0.4 Jy at the 2 ISOSS epochs
(see Table 4). The measured ISOSS upper limits
are in agreement with the TPM predictions.
As for Astraea, a shape model has been established
for Iris based on a combination of radiometric, lightcurve
and occultation data (Magri et al. 1999).
The corresponding TPM prediction gives 2.1
0.2 Jy at
the ISOSS epoch, with an additional lightcurve variation
of about 25% (min to max). The measured lower flux limit is
in agreement with the calculations.
Eunomia was one of the best observed asteroids by IRAS:
7 epochs distributed over almost one month, each time observed with
high S/N in all 4 bands. The MPS diameter
is 255.3
15 km and the albedo 0.21
0.03. Two single
chord occultation measurements led to diameters of >309
5 km
(Overbeek 1982) and >232 km (Stamm 1991).
The TPM prediction (based on MPS
diameter and albedo and on shape and spin vector by Erikson
2000) gives 2.5
0.5 Jy at the ISOSS epoch,
with the main error contribution coming from the large lightcurve
variation. The measured ISOSS flux of 2.9 Jy agrees within the
errorbars.
IRAS observed this asteroid 4 times within 2 weeks, each time
with high S/N in all 4 bands. The MPS diameter
is 151
3.1 km and the albedo 0.18
0.01. No shape
or spin vector is available, but the possible lightcurve
amplitudes range between 0.10 and 0.25 mag (Lagerkvist et al.
1989). The TPM prediction (based on MPS
diameter and albedo together with a spherical shape) gives 1.7
0.1 Jy
at the ISOSS epoch, with an additional maximal lightcurve variation
of approximately 25% (min to max). The measured upper limit
agrees with this prediction.
This asteroid was observed by IRAS extensively at 9 epochs during
a period of 2 months with high S/N in all bands. The MPS diameter
is given with 132.3
5.5 km and the albedo 0.06
0.01.
No shape and spin vector exists currently for Desiderata,
but a 0.17 mag lightcurve amplitude has been stated by Lagerkvist
et al. (1989). The TPM prediction (based on MPS
diameter and albedo together with a spherical shape) gives 5.8
0.5 Jy
at the ISOSS epoch, with an additional maximal lightcurve variation
of approximately 17% (min to max). Assuming an ISOSS observation
at lightcurve minimum and a diameter at the lower end of MPS diameter
range would result in a TPM flux which is only a few percent above the
measured ISOSS value, but well within the ISOSS measurement error bars.
7 IRAS observations with high S/N in either 3 or 4 bands were
obtained between March and October 1983. The The MPS diameter
is given with 222.2
7.6 km and the albedo 0.17
0.01.
The occultation diameter of 217
15 km is based on several
chords in combination with information on the pole orientation and a
lightcurve fit (Bowell et al. 1978).
MPS and occultation diameters agree nicely. The
complete shape and spin vector solutions derived from lightcurve
observations are given in Erikson (2000). Using the
full information for Herculina (see also Müller &
Lagerros 1998 for details) led to
170
m fluxes of 5.5
0.4 Jy and 3.8
0.3 Jy.
These values have
been calculated using the exact lightcurve phase and amplitude
at the time of the ISOSS observations. The almost perfect
agreement between predicted and measured fluxes (see Table 4) confirms
in an independent way the reliable calibration of this
new flux extraction method for the ISOSS.
IRAS saw Ganymed only twice and in both cases only
the 25
m flux was useable for the radiometric
calculations, resulting in a diameter of 31.7
2.8 km
and an albedo of 0.29
0.06. A single chord occultation
measurement gave a lower diameter limit of 16 km (Langans
1985
).
The large lightcurve amplitude
of 0.45 mag (Lagerkvist et al. 1989) adds
more uncertainties to the model calculations. Purely based
on the given diameter and albedo, the TPM predicts
approximately 0.1 Jy for the time of the ISOSS observation,
which is well below the detection limit of this observing mode.
5 asteroids (6 hits) have no IRAS detection, but fulfilled
the conservative flux requirements for the ISOSS asteroid search
(see Table 5). Unfortunately 4
sources (Euterpe, Ingeborg, Cruithne
and 1991 CS) have only marginal ISOSS detections
and establishing upper flux limits was difficult.
The results of Table 5 per Object:
Kristensen (1984) has determined a size of
190
19 km for this asteroid from an occultation event.
A second occultation a few years later gave a high quality
173.5 km diameter (Stamm 1989; Blow 1997).
Recent HST images (Storrs et al. 1999) revealed an elongated disk
with a long axis of 235 km and a short axis of 165 km, which
corresponds to an effective diameter of 197 km. Given the uncertainties
involved we adopt the occultation result which is perfectly consistent
with both techniques (see also Lagerros et al. 1999).
The full light curve and shape information
has been taken from Erikson (2000). The TPM predictions
gave 4.1
0.8 Jy and 3.4
0.7 Jy, respectively
(see Table 5). Adopting the HST results instead
led to about 5% and 10% higher fluxes.
Based on the ISOSS flux of 3.8 Jy, the TPM allowed the calculation
of an effective projected diameter of 178 km and an albedo
of pV=0.15 at the epoch of the ISOSS observation.
A possible 20% ISOSS flux uncertainty would correspond to
about 10% diameter uncertainty, resulting in a
size of the rotating ellipsoid of
km with 10% minimum uncertainties.
Within the different uncertainties and based on the shape and spin
vector solutions, the results agree nicely. The 3 methods - occultation,
HST direct imaging and ISOSS radiometric method - led to
comparable diameters and albedos.
No IRAS observations are available. We used instead the largest
extension from an occultation measurement (Dunham 1998)
together with a shape and spin-vector model (Erikson 2000),
H, G values (Piironen et al. 1997) and an albedo of
0.13 related to the occultation cross section. The TPM prediction
was 1.5 Jy with a large uncertainty due to the limited size knowledge.
This is well within the detection limits, but the source was too far
from the slew center to determine an upper flux limit.
There exists hardly any information about this asteroid. Based on its
H-value of 10.1 mag, together with a typical S-type (Tholen 1989)
albedo of 0.155 one can calculate an approximate diameter of 32 km.
The corresponding flux calculation for the ISOSS epoch gave 0.2 Jy,
which is clearly below the detection limit. Even under the assumption
of an extreme albedo of 0.03 the asteroid flux at 170
m
would only be 1.3 Jy and therefore hardly detectable. Like in the
case of Euterpe, Ingeborg had a slew center offset
which was close to the maximal allowed 5
.
Cruithne is currently the only known object on a horseshoe
orbit around the Earth (Christou 2000). It was also part of
a special near-Earth object observations
programme (Erikson et al. 2000a). Based on an unweighted
mean of typical C and S-type asteroids (pV=0.12) and an H-value
of
,
they calculated a diameter of 3.7 km and
a slow rotation period of 27.4 hours. Although the observing
geometry with only 0.37 AU from Earth was almost ideal, the 170
m
flux prediction was only 0.1 Jy. Even an extremely low albedo (leading to
a diameter of about 8 km) would only give 0.3 Jy well below the detection
limit. Therefore an upper limit from a background analysis would not give
any new information.
The case of 1991 CS is similar to Cruithne: a near-Earth
asteroid, at only 0.14 AU from Earth at the time of the ISOSS slew and
with an H-value of 17.4 mag. A radar campagne resulted in an estimated
diameter of 1.1 km, an albedo of 0.14 and a rotation period of 2.39 hours
(Pravec et al. 1998). The TPM predicts a 170
m
flux below 0.1 Jy and even for extreme albedo values the flux would be below
0.3 Jy and therefore not detectable for ISOSS.
The Juno observations in Tables 1 and 3 have flux ratios systematically higher than
ratios from comparable sources.
Calibrating the ISOSS values with the corresponding newly established
methods 1 and 3a resulted in an average observation over model ratio
of 1.14, indicating that the model diameter of Juno is about
7% too low. Müller & Lagerros (2002a) analysed 11
independent ISO observations, taken with the long wavelengths ISOPHOT
detectors. They find a similar mean ratio of 1.13
0.10, which
confirms the tendency to higher diameter values. Both investigations
indicate that the effective diameter should be close to 260 km, compared
to the published values of 241.4 km (Müller & Lagerros 1998)
and 233.9
11.2 km (Tedesco et al. 1992).
For Vesta the situation is not that clear. The values in
Table 1 are not conclusive since the Vesta
fluxes cover the difficult transition region between little flux loss and
the more than 40% flux loss for sources brighter than 25 Jy
(see Fig. 4).
It seems that two of the measured fluxes (TDT 07881200, 79781500)
are higher than one would expect from other sources with
similar brightness. This contradicts the findings by Redman et al.
(1998; 1992). They state for Vesta
an extremely low emissivity of 0.6 in the submillimetre. Assuming
that the emissivity is already lower in the far-IR
would mean that the Vesta points in
Fig. 4 should lie below the general trend and not
above. The measurement from methods 3a (TDT 57581500) and 3b (TDT 61580800)
agree within the errorbars with the model predictions.
As in Müller & Lagerros (2002a), we see no clear
indications of far-IR emissivities lower than the default values
given in Müller & Lagerros (1998)
5.3 Comets
The results of the positional search through the ISOSS pointing data,
combined with the flux estimates are given in
Table 6.
The table columns are: (1-3) same as in Table 1,
(4-5) ISO-centric coordinates (2000.0), (6-7) Sun and Earth distance
at the time of the observation, (8-9) ISOSS and model flux.
2P (Encke):
The comet has been detected at a slew end position on an
extremely high background close to the galactic plane.
The flux increase towards the comet nucleus corresponds
to about 5-10 Jy. The coma extension and its brightness
profile could not be determined due to the high background
brightness. The model flux at this close encounter with
Earth (0.26 AU) may have been strongly overestimated due to
the large apparent size of the central coma which was
assumed to be of constant brightness.
22P (Kopff):
The ISOSS slew ends again on the comet, but this time
the source is located on a clean low background. The signal
pattern is similar to that of a point-source with
0.5-1 Jy,
which is close to the detection limit. An upper flux
limit of 2 Jy can be given, which is in good agreement with
the simple model calculations (Table 6).
96P (Machholz 1):
The comet has not been detected. The position calculation
showed that the source was just outside the slew path, but
within the specified 5
search limit. The low model
flux indicated already the difficulty to detect the coma
or the comet nucleus.
103P (Hartley 2):
Only a poor detection of an extended source was found, although
the comet was on a low background. ISOPHOT observations close to the
ISOSS observing epoch show that Hartley 2 had a colour
temperature of 285 K (Colangeli et al. 1999).
This is 30 K colder than the calculated model temperature.
A second reason for the discrepancy between a low ISOSS flux
and the predicted 25 Jy is probably the too large apparent size of
the central coma which was assumed to be of constant brightness.
At an Earth distance of 0.82 AU the model comet core covers a
significant part of the aperture. Both effects together might
explain the model value.
104P (Kowal 2):
A source of approximately 1 Jy was detected by one pixel
at the predicted position of the comet, but confusion with
a close IRAS source could not be excluded.
In the first case the slew length was only 1
which was not
sufficient for the data analysis. The second case was a clear
detection by one pixel (Method 2). The derived flux of 1 Jy
is in agreement with the calculated upper limit of 2 Jy.
13481800: the slew passed over the coma with the nucleus only
30
outside the closest pixel. A weak signal of
2
has been detected in this pixel.
16280600: the slew ended on Hale-Bopp and the integrated
4-pixel flux was determined to 9.3
1.8 Jy (Method 3a).
31580500: method 3a was applicable and a 170
m flux of
30.9
7.3 Jy has been derived.
32081300: slew over the comet nucleus, with one pixel
crossing centrally, two pixels in 1
distance and one pixel in
2
distance. The slew crossed the nucleus under an angle of
45
relative to the orientation of the dust tail
(PsAng
,
ISOSS PosAng
).
The measured brightness profile clearly deviates from that of a
point-source. The asymmetric profile is stronger towards the east,
i.e. on the tail-side of the nucleus.
Due to a nearby cirrus ridge, the dust tail extension is confirmed out
to 2
only (but would be probably larger on a flat background).
32280200: the slew crossed the dust tail of Hale-Bopp
under an angle of approx. 30
in 8
distance from
the nucleus (
,
).
A signal increase at the position of the dust tail can be seen, but
the signal pattern is difficult to discriminate from the cirrus
structures in the background, hence quantitatively not helpful.
At the closest comet approach of 4
a second signal increase
can be seen which coincides with the position angle of the negative
of the target's heliocentric velocity vector
(PsAMV
).
The signal increase is either related to the large cometary coma at
a distance of only r=2.86 AU from the sun or a kind of trail
formation in the direction of PsAMV similar to what Reach et al.
(2000) found for comet Encke.
32580600: the detectors moved centrally along the dust tail
and cross over the comet nucleus (
,
). The measured brightness
profile clearly deviates from a point-source profile (see
Fig. 6). A dust tail extension of more than
4
can be seen where the satellite approaches the nucleus.
The signal in anti-tail direction decreases more rapidly
(see also Sect. 5.3.2).
77081500: method 3a was applicable again and a 170
m flux
of 43.8
4.0 Jy was derived.
86880300: the slew passes in 4
distance ahead of the comet
tail under an angle of approx. 70
with the sun direction
(
,
).
A signal change of 2
extended over 15
can clearly
be seen. Due to the viewing geometry (the phase angle is only
11
)
coma and tail are difficult to separate and the signal
increase is most likely connected to the dust emission of the extended
coma and tail structures of Hale-Bopp. Here, as in 32280200,
the signal maximum coincides with the PsAMV angle of
.
A connection might be possible between the 170
m signal pattern
and large particles forming an elongated structure behind the comet
nucleus while it is moving away from perihelion.
87380400: method 3a was applicable again and a 170
m flux
of 15.0
2.9 Jy was derived.
![\begin{figure}
\par\includegraphics[width=8.8cm,clip]{tmueller_fig6.ps}
\end{figure}](/articles/aa/full/2002/26/aa2270/Timg56.gif) |
Figure 6:
An ISOSS profile of Hale-Bopp in comparison
with the signal pattern of a point source. The slewing
speeds for these two hits were identical and already quite low
(2 /s), resulting in a high sampling rate. The offsets are
defined with respect to the positions from the ephemeris
calculations and have not been shifted for the two sources relative
to each other. Hale-Bopp shows an asymmetrically extended
profile. Note that the profiles are not deconvolved. |
Figure 6 illustrates a measured signal profile
from a central slew over Hale-Bopp (TDT 32580600) in
comparison with a slew over the point source Neptune
(TDT 72081600). For these two detections, the geometrical
configurations and the slewing speeds (2
/s) have been
identical. The detectors moved first centrally over the dust
tail (left side of the peak) and then over the nucleus of
Hale-Bopp (peak). An asymmetric signal profile between
and
can be seen. In the case of Neptune
the signal increase starts approximately 2
ahead of the
true position, which is related to a combination of the
Airy disk with the slew speed and read out frequency
(see also Hotzel et al. 2001). The slight shift
between the two peaks is most probably related to the possible
positional uncertainties of ISOSS data (see
Sect. 4.4). This Hale-Bopp asymmetry
has not been seen in dedicated 170
m maps (Peschke et al. 1999) which were taken at r=3.904 AU (as compared
to r=2.816 AU in Fig. 6).
The fluxes derived from Method 3a can be compared to results from
Grün et al. 2001 through the following corrections:
1) Flux corrections according to Fig. 5;
2) Normalization to a standard aperture diameter of 23
assuming that the coma brightness scales linearly with
aperture diameter (
ca23=0.120);
3) Point-spread-function correction which takes into account
the differences of a point source and a
-coma
(
);
4) Colour correction
which changes for different distances
from the sun (
).
The first two measurements (16280600, 31580500) were obtained
on the same days as the ones in Grün et al. (2001).
The calibrated and reduced 23
fluxes agree within
the errorbars. The third observation (77081500), taken 5 days
earlier than the dedicated Hale-Bopp observation,
lead to a flux of 5.64
0.56 Jy, as compared to 2.66 Jy.
This large difference can not be explained by epoch or geometry
differences. However, the dedicated measurement was mis-pointed
by 24
which caused large uncertainties in the applied
corrections. The ISOSS flux provides therefore valuable information
for the colour temperature determination and, through grain size
models, might give clues whether icy grains were present in the coma
in December 1997 at almost 4 AU post-perihelion.
The last measurement of Method 3a (87380400) was obtained when
Hale-Bopp was already 4.9 AU from the sun. The
calibrated and reduced 23
flux was 1.97
0.39 Jy.
This is the most distant thermal far-IR observation of Hale-Bopp
post-perihelion. A comparison of the flux with a dedicated
observation (Grün et al. 2001;
Jy) at
a similar distance from the sun, but pre-perihelion, shows that the
dust emission post-perihelion was higher as the comet receeded from
the sun. In fact, the higher far-IR fluxes post-perihelion are
related to contributions from large particles which have been
accumulated during the passage around the sun and which stay on
similar orbits as the nucleus.
Two measurements (32280200, 86880300) show signal peaks a few arcminutes
away from the nucleus in anti-orbital velocity (trail) direction.
It seems that the emitting dust particles are not homogeneously
distributed and are concentrated in a narrow region of the outer
parts of the dust coma towards the trail direction.
These features are not seen in slews over other parts of the outer coma.
Reach et al. (2000) observed for the first time the dust
trail formation in comet Encke in the mid-IR. The ISOSS data provide now evidence for this process in the far-IR where
the emission is strongly connected to the largest particles.
Up: Solar system objects in survey
Copyright ESO 2002