I’m trying to run a model in Aspect that has radioactive decay, and also uses two compositional fields with different densities, but I am noticing some strange behavior.
I start with a two layer model, with a basal layer that has slightly higher density. I also use the radioactive decay heating model, where the heating elements are enriched (by 20x) in the basal layer.
However, it appears that the initial density distribution is used to compute the heating throughout the simulation, instead of computing the heating with the updated density distribution as the two layers mix.
Initially the simulation starts with two layers, and after some time it looks like this:
The heating elements (top plots) generally partition to follow the compositional fields (bottom plots). However, there is a strange jump in the heat production that corresponds to the initial location of the basal layer. The change in heating is not that large, I had to adjust the scaling to see it. It is the transition from light red to dark red across the bottom of the plume.
It is visible in the background mantle as well, when the scale bar is properly adjusted.
My suspicion is that while it is using the compositional fields to properly determine the W/kg heating rate for each compositional field, it is using the initial density distribution to convert W/kg to the W/m^3 heating rates shown in the plots.This would explain why the differences in heating rate are small – the density difference between the top/bottom layer is only 59 kg/m^3.
Instead, aspect should use the density distribution at each timestep to convert the heating rate.
Does this look like the proper diagnosis? Any ideas on how to fix this?
(More images to follow - I could only put one image per post as a new user)