From Table 3.9, it can be seen that reactions R13–R17 all have a radiated element
leading to energy loss from the system.
Altogether, 181 species were computed in the model of Rodgers and Charnley
(2002) with over 3500 individual reactions. One of the resulting figures is shown as
Fig. 3.49 and shows the abundances of different species as a function of
cometocentric distance. It should be noted that the values have been multiplied by
4πr
2 to give a flux through a spherical shell at each distance from the nucleus. This
linearises the function if the species is unaffected by reactions and expands as 1/r
2 .
Note also that the abscissa is logarithmic.
The top panel shows the parent molecules. Initially, the fluxes for these molecules
show the chosen starting conditions at the nucleus but at distances of >10
4 km from
the nucleus, the main reactions (e.g. photodissociation) begin to remove these
species. The central panel shows the daughter neutrals. In the case of the larger
daughter product, HC 3 N, its increase in local density with distance is slowed and
reversed as it undergoes reactions itself. Finally, the bottom panel shows ion species
and the electron density. This rises very rapidly with cometocentric distance.
There are several assumptions in the approach used to construct the coma
chemistry network described above. Firstly, the coma is assumed to be spherically
symmetric. The dayside-nightside asymmetry at the nucleus itself clear violates this
assumption. However, gas-gas collisions in the innermost coma where the gas is still
collisional will tend to move the spatial distribution to a more symmetric structure
and obviously the higher the local density near the nucleus, the more symmetric the
spatial distribution should become. Secondly, the coma is assumed to be compositionally uniform. Here again, the evidence suggests that compositional variations in
the innermost coma (e.g. the CO 2 /H 2 O ratio in Fig. 3.2) are actually significant and
this could easily challenge 1D models. This has not really been tested in a rigorous
manner to this point. Third, the calculations assume that there are no magnetic fields
acting on the ions. This is a significant issue that we will address when we discuss the
spatial distribution of ions below. However, this does indicate that a coupling
between ion-neutral reactions and effects resulting from ion pickup by the solar
wind are necessary to describe the spatial distributions more exactly. On the other
hand, this breaks the symmetry completely and increases the complexity enormously. Schmidt et al. (1988) were the first to tackle this problem and simplified
systems have been studied in detail by, for example, Rubin et al. (2014c). But
analysis of the complexities resulting from a detailed chemical network as used by
Rodgers and Charnley in a full 3D system has still not been attempted. Finally, as the
density falls off, the use of the fluid equations to simulate a system that is close to the
free molecular flow regime can introduce significant errors. For example, fast
products of dissociation reactions will not be able to transfer their energy to the
coma through collision if the mean free path becomes comparable to the size of the
coma itself. Hence, temperatures in the outer coma can be substantially
overestimated (Ip 1983).
One of main reasons for studying these chemical networks is to derive the nucleus
composition of parent species from coma in situ measurements (using mass spectrometers for example) by accounting for known reactions that can influence the
3.5 Reaction Chemistry and the Extended Coma
261
leading to energy loss from the system.
Altogether, 181 species were computed in the model of Rodgers and Charnley
(2002) with over 3500 individual reactions. One of the resulting figures is shown as
Fig. 3.49 and shows the abundances of different species as a function of
cometocentric distance. It should be noted that the values have been multiplied by
4πr
2 to give a flux through a spherical shell at each distance from the nucleus. This
linearises the function if the species is unaffected by reactions and expands as 1/r
2 .
Note also that the abscissa is logarithmic.
The top panel shows the parent molecules. Initially, the fluxes for these molecules
show the chosen starting conditions at the nucleus but at distances of >10
4 km from
the nucleus, the main reactions (e.g. photodissociation) begin to remove these
species. The central panel shows the daughter neutrals. In the case of the larger
daughter product, HC 3 N, its increase in local density with distance is slowed and
reversed as it undergoes reactions itself. Finally, the bottom panel shows ion species
and the electron density. This rises very rapidly with cometocentric distance.
There are several assumptions in the approach used to construct the coma
chemistry network described above. Firstly, the coma is assumed to be spherically
symmetric. The dayside-nightside asymmetry at the nucleus itself clear violates this
assumption. However, gas-gas collisions in the innermost coma where the gas is still
collisional will tend to move the spatial distribution to a more symmetric structure
and obviously the higher the local density near the nucleus, the more symmetric the
spatial distribution should become. Secondly, the coma is assumed to be compositionally uniform. Here again, the evidence suggests that compositional variations in
the innermost coma (e.g. the CO 2 /H 2 O ratio in Fig. 3.2) are actually significant and
this could easily challenge 1D models. This has not really been tested in a rigorous
manner to this point. Third, the calculations assume that there are no magnetic fields
acting on the ions. This is a significant issue that we will address when we discuss the
spatial distribution of ions below. However, this does indicate that a coupling
between ion-neutral reactions and effects resulting from ion pickup by the solar
wind are necessary to describe the spatial distributions more exactly. On the other
hand, this breaks the symmetry completely and increases the complexity enormously. Schmidt et al. (1988) were the first to tackle this problem and simplified
systems have been studied in detail by, for example, Rubin et al. (2014c). But
analysis of the complexities resulting from a detailed chemical network as used by
Rodgers and Charnley in a full 3D system has still not been attempted. Finally, as the
density falls off, the use of the fluid equations to simulate a system that is close to the
free molecular flow regime can introduce significant errors. For example, fast
products of dissociation reactions will not be able to transfer their energy to the
coma through collision if the mean free path becomes comparable to the size of the
coma itself. Hence, temperatures in the outer coma can be substantially
overestimated (Ip 1983).
One of main reasons for studying these chemical networks is to derive the nucleus
composition of parent species from coma in situ measurements (using mass spectrometers for example) by accounting for known reactions that can influence the
3.5 Reaction Chemistry and the Extended Coma
261