Extended Data Fig. 6: Flux variations expected from microlensing simulations.

Microlensing is only expected to vary the total magnification by a factor of 2–3 over time, consistent with the observed steady flux over 3.5 years. a, The simulated microcaustic network arising from stars and stellar remnants within the lensing cluster. The cluster caustic is the extreme magnification horizontal region near the middle of the image, with individual cusps from microlenses still visible beyond the cluster caustic. We estimate Earendel will move relative to the microlens network at ~1,000 km s⁻¹ in some unknown direction. b, Predicted magnification fluctuations over time arising from this motion in the 1M_⊙ pc⁻² case (blue) and the 10M_⊙ pc⁻² case (purple), assuming that the relative motion is at an angle of 45°. Grey bands highlight a factor of 2 (dark) and a factor of 4 (light) change in magnification. c, The likelihood of magnification variations between two observations separated by different times, again for both the 1M_⊙ and 10M_⊙ pc⁻² cases. Note the ‘more is less’ microlensing effect that reduces variability in the observed images when the density of microlenses increases.