I've been trying an exercise from a book I'm self-studying, and I can't seem to get the blowup right. I'm not sure where I'm going wrong, so a nudge in the right direction would be helpful.

$\textbf{Summary of problem}$:

Given the following family of parametrized nodal cubics

$$ \mu: \mathbb{P}^1 \to \mathbb{P}^2$$ $$[s:t] \to [3b^2st^2: t^3 - 3s^2t: s^3]$$

there is a corresponding family of dual curves of quartics $$ [s: t] \to [s^2(s^2- t^2) : 2b^2s^3t: b^2t^2( t^2 + 3s^2)] $$ From here, it is clear that the member corresponding to $b = 0$ has three points of indeterminacy, all when $t=1$ and when $s=0$ or $s= \pm 1$. I need to show that the limit of these quartics should be a three-component curve, two simple lines and double line.


The idea is that when I blow up at these points of indeterminacy, I should have that the exceptional divisors get mapped to three lines. The double line comes out cleanly by blowing up at $s=0$, but I'm having issues with the others. I'll do the blowup at $[1:1]$ to illustrate my problem.

We can work with the chart $t= 1$. Here, we have an affine chart $\mathbb{A}^2$ with coordinates $(b,s)$, and we want to blow up at $(0,1)$. This results in the variety $\{ ( (b,s), [b_1: s_1]) | b_1(s-1) = s_1b\} \subset \mathbb{A}^2 \times \mathbb{P}^1$. Again, I restrict to an affine chart, say when $s_1 \neq 0$. In this chart, we have $b = b_1(s-1)$, so we can rewrite our dual map as $$(s, b_1) \to [s^2(s^2- 1) : 2b_1^2(s-1)^2s^3: b_1^2(s-1)^2(1 + 3s^2)]$$ This is the total transform of our map, so to get the proper transform, we divide out be $s-1$ $$(s, b_1) \to [s^2(s + 1) : 2b_1^2(s-1)s^3: b_1^2(s-1)(1 + 3s^2)]$$ Our exceptional divisor is given by $s = 1$, and thus plugging in $s= 1$ in the above map, we get that the image of our exceptional divisor is $[2: 0: 0]$.

From the above, I'm getting that the exceptional divisor gets mapped to a point, not a line. The $b^2$ terms are problematic here, as I'm getting an extra copy of $(s-1)$ in the last two coordinates. Doing the same calculations on the other open affine gives me that it maps to a point as well.

Any help would be appreciated, particularly to whether I'm resolving the map correctly. Thanks!


The quartic has equation $3x^2y^2b^4+4y^4-4x^3zb^6-6xy^2zb^2-y^2z^2$ so when $b=0$, you're left with $4y^4-y^2z^2$ or $y^2(4y^2-z^2)$.

To find the implicit equation I used Macaulay2 on



Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.