Hi,
The reason you get observation differences is due to a couple of different reasons
1. Space is curved, and is like an onion with it's dimensional layering each of which run at different epochs.
2. Although the warping of space (3 degrees approx) the effect on one's observation of another point can be dramatically altered with only a small displacement from one's last observation position, it is kind of like looking thru a warped piece of glass, the light is refracted at very different angles with only a small amount of displacement.
The distort effect is even greater when the different epoch layers are take into account.

In your model you should be able to say, observe a pulsar and approximate its energy level at at time, and also observing a point at which it was in the past you should be able to observe its energy levels at that time.

>>Why not use the neural network software and skew the nodes to simulate multiple epoch layers.
>
>
>Use it how? Add it to the rules of the model?
>
>Also, please explain what you mean by "multiple epoch layers"?
>
>
>For what it is worth, I think a neural network is needed. But its implementation has to be carefully considered.