Important warning (October 2025): we recently learned that the main definition of Gekeler ratio is incorrect. There is a conceptual error in the paper. There is, however, an easy fix. Erratum and full corrections will appear soon. In the meanwhile, if you are planning to use any part of this paper, please email me for corrections and clarification.
The correct version for stable orbital integrals appears in the appendix to On the pointwise convergence of the number of Abelian varieties over F_p with fixed trace .Important warning (October 2025): we recently learned that there is a small conceptual error in the paper, in addition to all the corrections discussed in Section 7 of the sequel . None of the main results are affected, and a corrected version will be posted to the ArXiv soon. In the meanwhile, if you are planning to use any part of this paper, please email me for corrections and clarification.
The preliminary corrected version (new text in blue) is here .