S n o w

C o v e r

S t a b i l i t y,

A v a l a n c h e

I n i t ia t i o n

a n d

F o r e c a s t i n g

servedpattern ofdeformationand,as afirstapproxima-
tion,we rewrite the creep equation:
e_ij= Ad _ij+ Bs_ij+ Cs
ij^s
k j^+d ij^e
m_ij(1)

where e_ijisthe deformationrate tensor,e^misthe meta-
morphic (or capillary) strain rate,s_ijis the str^ij
ess tensor,d
is the Kronecker delta and A, B, and C are scalar functions^ij
of the stress invariants.

EVOLUTION OF MECHANICAL PROPERTIES

Mechanical properties such as strength and ductility un-
dergo significant improvement as a result of densification
and sintering. For example experimental data indicate the
shear strength of snow increases rapidly with density (eg.
Perla et al, 1982; Jamieson, 1995). Their results predict the
resistance to shear (other things being equal) will increase
by more than an order of magnitude as the snow density
increases from 60 kg m ^-3to 180 kg m^-3(typical forpartly
settled snow). It is also well known that the shear strength
of snow depends on temperature, the rate of shear as well
as thedensity.For examplemeasurementsof shear behavior
inthe laboratory(McClung,1977;deMontmollin,1982;
Fukuzawa and Narita,1992) indicate that snow doesnot
show a distinct failure at low rates of shear; the strength
increaseswith strain (strain-hardening). On the otherhand
at high rates,the stress increases toa peak and then de-
creasestoaresidualafterdisplacementsofjustafew
millimeters (strain-softening). Wethink thatsinteringproc-
essescontributetothisbehavior.Atwarmtemperatures
and low rates of deformation, grain bonds may form faster
than they break,but as the rate increases (orthe tempera-
ture decreases), bonds tend to break faster than they form.
Weexpectthe critical strainrateat whichbehavior changes
from strain-hardeningto strain-softening dependson both
temperature and snow texture.

SLAB FAILURE AT THE ONSET OF RAIN

Recent models of snow slope stability calculate the energy
neededtocauseazoneofbasalweakness(wherethe
stressesfromtheoverburden are notfullysupported) to
propagate. Figure 5a shows a physical picture of this con-
dition and McClung (1979) showed that a basal weakness
of length 2 L overlain by a slab of height H would propa-
gate if:

H[(t g-t r)L]2> ( tr-tr) v(2) 2E'H

where E'is the viscoelastic modulus of the slab,t_gis the
body weight shearstress,t_pis the peak shear stress near
the tip of the weak zone, t _ris the residual shear stress (after
softening) which is attained after a characteristic distance
v.
McClung consideredthe length L to be the only free pa-
rameterbut in some cases the effective depth H may also
be a freeparameter. For exampleavalanchecontrollers often
successfully "ski-cut" slopes by cutting across the bottom
and thenthe top of a slab. If such cuts were say 25 cm deep
and the basal weak zone was 50 cm below the surface, the
energy available to drive propagation (the LHS of eqn.2)
would double (even without accounting for the additional^{Fig. 5b Increase in available energy as a function of initial slab}
stress imposed by the skier).^{thickness for cases where theeffective slab depth decreases by 5, 10,}
and 20 cm.

79