The explicit factorization of the Cremona transformation which is an extension of the Nagata automorphism into elementary links (Q3022684)

This interesting paper deals with the famous example of an automorphism of \(\mathbb{C}^3\) constructed by M. Nagata, viz. \[ \sigma(x)= x- 2y(xz + y^2)- z(xz + y^2)^2,\;\sigma(y)= y+ z(xz + y^2),\;\sigma(z)= z. \] For a long time it was not clear if this automorphism is a product of affine and de Jonquiere automorphisms (i.e. whether a is tame). This question was recently answered by \textit{I. P. Shestakov} and \textit{U. U. Umirbaev} [J. Am. Math. Soc. 17, No. 1, 197-227 (2004; Zbl 1056.14085)] who showed that the Nagata automorphism is not tame. In the paper under review the author proves that \(\sigma\) is a product of 8 elementary links according to the Sarkisov program. The Sarkisov program developed by Sarkisov, Reid and Corti guarantees that any birational map between 3-dimensional Mori fiber spaces factorizes into elementary links.