We model the heating and cooling processes in the hydrogen- and helium-rich
zones of the envelope of SN 1987A from t = 200 to 1200 days after outburst
and use these results to calculate the light curves of the most prominent
emission lines. For the first 600 days, heating and cooling processes are i
n equilibrium. The main heating mechanism is direct heat deposition by nont
hermal electrons, and the main cooling mechanism is collisional excitation
of trace elements such as Ca II, Fe II, and C I, followed by the emission o
f a line photon. After 600 days adiabatic cooling becomes important, and th
e cooling and heating rates are no longer in equilibrium Dust, formed in th
e Fe/Co/Ni zone after t similar to 400 days, plays an important role in the
formation of the emission lines. It both modifies the internal UV radiatio
n held that excites the ions and reduces the escaping line fluxes by extinc
tion. The pseudocontinuum opacity in the envelope due to the many absorptio
n lines of metals, which we model crudely by a simple power law, is also im
portant for the emerging spectrum.
Our results for the temperature evolution do not depend strongly on our ass
umptions. We find that the temperatures of the hydrogen and helium zones ev
olve from T approximate to 6000 K at t = 200 days to T approximate to 1000
K at t = 1200 days. The ionized fraction of hydrogen evolves from x(H) appr
oximate to 6 x 10(-3) at t = 200 days to x(H) approximate to 3 x 10(-4) at
t = 1200 days. With abundances determined from observations of the circumst
ellar ring, the model can account for the light curves of most strong emiss
ion lines of H I, He I, Ca II, and Fe II, but some discrepancies remain. Es
pecially interesting is the H beta light curve, which exhibits a clear plat
eau when H beta is still optically thick, but Pa alpha is already optically
thin. In all our models this phase appears to occur later than in the obse
rvations.
For t greater than or similar to 800 days, the infrared emission lines of F
e II are produced mainly by primordial iron in the H/He envelope, not by ne
wly synthesized iron. The fluxes of C I and O I lines that our model predic
ts are much higher than observed, and they may require a significant adjust
ment in abundance or mass of the different composition zones to make them a
gree with observations. Our models also indicate that the total helium mass
in the core of the remnant (upsilon < 2500 km s(-1)) must lie in the range
2-5 M.. The hydrogen mass in the core is less well constrained, because th
e hydrogen line strength does not vary much as long as most of the nontherm
al energy is deposited in hydrogen. The ratio of the fluxes of the Br gamma
and the He I 2.058 mu m lines is slightly more sensitive, and it indicates
a helium mass-to-hydrogen mass ratio similar to 1:2.