There are a couple of issues that you need to add to your picture.
First, the temperature varies over time so it is not constant at a certain elevation throughout e.g. the summer.
Second, snow will melt from the energy provided to the snow pack. This energy is not exclusively expressed as temperature. The energy fluxes that reach the ground and can provide energy for melt include
- Long wave radiation (infra red)
- Short wave radiation (visible light)
- Sensible heat (heat content of the air convected towards the surface
due to turbulence)
- Latent heat (Air moisture advected to the surface through turbulence;
a phase change from gas to liguid during condensation expells heat)
- Energy contained in rain drops, i.e. the temperature of the rain water, falling on the surface
The radiation will in part be reflected so not all incoming radiation will be available for melting. A new fresh snow surface can reflect as much as 90% of the incoming radiation, we say the snow has an albedo of 0.9 where 1 is 100% reflection and 0 is no reflection at all. Snow surface commonly have lower albedo since there may be dust and other particles that can accumulate and lower the albedo making the snow pack absorb more radiation thereby allowing more energy to contribute to melting. Wet snow also lowers the albedo, particularly for long wave radiation.
A melting snow surface will always be at the melting point, essentially zero but particularly the radiation components will contribute energy to cause melting at subzero temperatures. It is common to experience wet snow with a thin frozen crust in spring when incoming radiation is high but air temperature may be below zero. The radiation causes subsurface melt while the cold air keeps the crust frozen.
When snow covers rocky ground some radiation will heat rocks that protrude the cover and help snow to melt in the proximity of such rocks. If the snow cover is thin the ground may be heated from absorbing the radiation and cause melting at the base of the snow pack. Radiation can penetrate about a meter of snow (depends on density) so when the snow pack is a few dm or so a fair portion of radiation will reach the ground beneath the snow pack.
Bare paches of ground will absorb more radiation that the snow cover surrounding the patches and become heat islands in the general snow cover. The air above these patches will warm up due to the warm ground and in the presence of wind provide warm air for melting in the near surroundings.
So, in conclusion, there are several processes we need to consider when trying to understand the snow line, one is the heat balance and the other is the micro-climate that result from surfaces having different albedo. Therefore the snow line elevation will move up in elevation during the melt season due to the cumulative effect of what has been discussed above.