1
$\begingroup$

I'm looking for good (as simple as it is possible) reference for the local discriminant variety.

I need it in the following situation: I have an unfolding $F: (\mathbb{K}^n \times \mathbb{K}^p, 0) \to (\mathbb{K}^n \times \mathbb{K}^p, 0) $ of a germ $f=f_0: (\mathbb{K}^n,0) \to (\mathbb{K}^n, 0) $ in the form $F(x, t)=(f_t(x), t)$. Here $\mathbb{K}= \mathbb{R}$ or $\mathbb{C}$, I don't need other fields, the germs $f$ and $F$ are analytic (or smooth). Let $\mathcal{D} \subset \mathbb{K}^p$ be the set of those $t $ parameter values for which $f_t$ has singular zeros, i.e. points $x \in \mathbb{K}^n $ with $f_t(x)=0$ and $ \det (\mbox{Jac}_x (f_t)) =0$. I'm looking for references for the fact that $\mathcal{D}$ is a local analytic (or smooth) variety.

I also ask for an advice for the terminology: Is $\mathcal{D}$ called local discriminant variety? For a fix $t$ can I call $f_t$ the perturbation of $f$ corresponding to the parameter value $t$, or does it have another name?

Additional information

  • $\mathcal{D}$ is the generalization of the discriminant of 1-variable polynomials in the following way: $n=1$, $f=x^{p-1}$, and $t$ is the vector of the coefficients of a degree-$p-1$ polynomial $f_t$. $\mathcal{D}$ defined above is the zero set of the discriminant of $f_t$, i.e. the resultant of $f_t$ and $f'_t$.
  • A multi-version for polynomials called discriminant variety can be found in the Maple help with references.
  • On the other hand in a new book Singularities of mappings the notion of the discriminant is used in another context, namely for the set of the singular values of a map.

Actually I need only for the fact that $\mathcal{D}$ has at least 1 complex codimension in the complex case. I guess I can prove it. The set $\mathcal{S}$ of singular points of $F$ in the source is the zero set of $\det (\mbox{Jac} (F))$, hence $\mathcal{S} $ has 1 complex codimension in $\mathbb{C}^n \times \mathbb{C}^p$. Its image $F(\mathcal{S})$ is the set of singular values of $F$, i.e. the discriminant in the sense of the Mond-Nuno-Ballesteros book. It has at least 1 complex codimension. Then $\mathcal{D}=F(\mathcal{S}) \cap \{x=0\} \subset \mathbb{C}^p$. I guess that the intersection is transverse, therefore the codimension is at least 1. Anyway, this argument does not answer my original question, just its original motivation.

$\endgroup$
2
  • 4
    $\begingroup$ Please use the Contact Us form to have your accounts merged, to regain full control over your posts. $\endgroup$
    – Glorfindel
    Commented Jan 26, 2022 at 20:50
  • $\begingroup$ I just realized: $t \in \mathcal{D}$ if and only if $(0, t)$ is a singular value of $F$, that is, $(0, t)$ is an element of the discriminant of $F$ in sense of the Mond--Nuno-Ballesteros book. The ideal of $\mathcal{D}$ can be somehow inherited from this observation (?) $\endgroup$ Commented Jan 26, 2022 at 23:42

1 Answer 1

1
$\begingroup$

In the complex case $\mathcal D$ is analytic when the zero locus of $f_0$ has isolated critical point. The set germ $S$ of pairs $(x,t)$ in $\mathbb C^n\times\mathbb C^p$ such that $f_t(x)=0$ and $\det(d_xf_t)=0$ is analytic. The discriminant $\mathcal D$ is the image of $S$ by the projection $\pi(x,t)=t$. Since $S\cap\{t=0\}=\{0\}$ (as set germs), the projection $\pi:S\to \mathbb C^p$ is finite and hence its image is analytic by the Remmert's Finite Mapping Theorem.

$\endgroup$
2
  • $\begingroup$ Thank you very much Juan! I guess the explicit condition ($f$ has isolated critical point) and referring to Remmert's Finite Mapping Theorem were missing for me. Do you know an explicit reference for the statement? I checked the Mond--Nuno-Ballesteros book, it uses the discriminant in a different sense. De Jong Local analytic geometry uses the discriminant only for polynomials, therefore I haven't found reference there. I'm also not sure whether it is convenient to call $\mathcal{D}$ ''discriminant''. $\endgroup$ Commented Feb 6, 2022 at 9:55
  • $\begingroup$ The Remmert's Finite Mapping Theorem (or Proper Mapping Theorem) can be found for instance in the Lojasiewicz book "Introduction to complex analytic geometry". Birkhäuser Verlag, Basel, 1991, in \S 5, Chapter V. The precise statement is that the image of any finite analytic map germ $f:(X,x)\to (Y,y)$ is analytic of the same dimension as $(X,x)$. $\endgroup$ Commented Feb 6, 2022 at 21:27

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .