The Tate–Šafarevič group of a polarised K3 surface
Daniel Huybrechts and Dominique Mattei · 2025
Read The Tate–Šafarevič group of a polarised K3 surface by Daniel Huybrechts and Dominique Mattei (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2507-22703.txt.
Introduction
Daniel Huybrechts and Dominique Mattei extend a classic algebraic geometry result to K3 surfaces in a 2025 note. For an elliptic K3 surface, the Tate-Shafarevich group collects the twists that can still be compactified into a K3 surface. The authors had earlier defined an analogue of this group for a polarised K3 surface, which is a K3 surface paired with a chosen divisor class.
Here they prove that, when the polarising class is primitive, this group matches exactly the torsors for the Jacobian of the generic curve in the linear system. Each such torsor that admits a good hyperkähler compactification appears once. The result mirrors the classical elliptic picture in the broader polarised setting.
Catalog record
| CREATOR | Daniel Huybrechts and Dominique Mattei |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 161 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2507-22703
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
The Tate–Šafarevič group of a polarised K3 surface Daniel Huybrechts and Dominique Mattei Abstract . In [ 17 ] we generalised the notion of the Tate–Šafarevič group X ( S / ℙ 1 ) \Sha(S/\mathbb{P}^{1}) of an elliptic K3 surface S ℙ 1 S\lx@xy@svg{\hbox{\raise 2.5pt\hbox{\kern 1.0pt\hbox{\ignorespaces\ignorespaces\ignorespaces\hbox{\vtop{\halign{\entry@#!@&&\entry@@#!@\cr&\crcr}}}\ignorespaces{\hbox{\kern-1.0pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\hbox{\kern 1.0pt\raise-2.5pt\hbox{$\textstyle{\ignorespaces\ignorespaces\ignorespaces\ignorespaces}$}}}}}}}\ignorespaces\ignorespaces\ignorespaces\ignorespaces{}{\hbox{\lx@xy@droprule}}\ignorespaces{\hbox{\kern 14.0pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\lx@xy@tip{1}\lx@xy@tip{-1}}}}}}{\hbox{\lx@xy@droprule}}{\hbox{\lx@xy@droprule}}{\hbox{\kern 14.0pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\hbox{\kern 1.0pt\raise-2.5pt\hbox{$\textstyle{}$}}}}}}}\ignorespaces}}}}\ignorespaces\mathbb{P}^{1} to the Tate–Šafarevič group X ( S , h ) \Sha(S,h) of a polarised K3 surface ( S , h ) (S,h) . In the present note, we complement the result by proving that for a polarised K3 surface ( S , h ) (S,h) with h h primitive the group X ( S , h ) \Sha(S,h) parametrises bijectively all Pic 0 ( 𝒞 η ) {\rm Pic}^{0}({\cal C}_{\eta}) -torsors that admit a ‘good’ hyperkähler compactification. Here, 𝒞 η {\cal C}_{\eta} is the scheme-theoretic generic curve of the linear system | h | |h| . The result is seen as the analogue of the classical fact that the Tate–Šafarevič group X ( S / ℙ 1 ) \Sha(S/\mathbb{P}^{1}) of an elliptic K3 surface is the subgroup of the Weil–Châtelet group of all twists that can be compactified to a K3 surface. 0 0 footnotetext: The first author is supported by the ERC Synergy Grant HyperK (ID 854361). 1. Introduction Let (S,h) be a polarised K3 surface over \mathbb{C} . In particular, we assume that the polarisation h is primitive. We write {\cal C}\lx@xy@svg{\hbox{\raise 2.5pt\hbox{\kern 1.0pt\hbox{\ignorespaces\ignorespaces\ignorespaces\hbox{\vtop{\halign{\entry@#!@&&\entry@@#!@\cr&\crcr}}}\ignorespaces{\hbox{\kern-1.0pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\hbox{\kern 1.0pt\raise-2.5pt\hbox{$\textstyle{\ignorespaces\ignorespaces\ignorespaces\ignorespaces}$}}}}}}}\ignorespaces\ignorespaces\ignorespaces\ignorespaces{}{\hbox{\lx@xy@droprule}}\ignorespaces{\hbox{\kern 14.0pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\lx@xy@tip{1}\lx@xy@tip{-1}}}}}}{\hbox{\lx@xy@droprule}}{\hbox{\lx@xy@droprule}}{\hbox{\kern 14.0pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\hbox{\kern 1.0pt\raise-2.5pt\hbox{$\textstyle{}$}}}}}}}\ignorespaces}}}}\ignorespaces|h| for the universal curve over the associated linear system and denote its scheme geometric fibre by {\cal C}_{\eta} , which is, therefore, a smooth integral curve over the function field k(\eta) of |h| . In [ 17 ] we introduced the Tate–Šafarevič group \Sha(S,h) of (S,h) and associated with any class \alpha\in\Sha(S,h) the {\rm Pic}^{0}({\cal C}_{\eta}) -torsor {\rm Pic}_{\alpha}^{0}({\cal C}_{\eta}) of \alpha -twisted invertible sheaves. The group \Sha(S,h) was defined cohomologically by means of the special Brauer group {\rm SBr}(S) , see [ 17 , § 4] and § 3 for more information. The classical analogue of the above situation is the case of an elliptic K3 surface S\lx@xy@svg{\hbox{\raise 2.5pt\hbox{\kern 1.0pt\hbox{\ignorespaces\ignorespaces\ignorespaces\hbox{\vtop{\halign{\entry@#!@&&\entry@@#!@\cr&\crcr}}}\ignorespaces{\hbox{\kern-1.0pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\hbox{\kern 1.0pt\raise-2.5pt\hbox{$\textstyle{\ignorespaces\ignorespaces\ignorespaces\ignorespaces}$}}}}}}}\ignorespaces\ignorespaces\ignorespaces\ignorespaces{}{\hbox{\lx@xy@droprule}}\ignorespaces{\hbox{\kern 14.0pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\lx@xy@tip{1}\lx@xy@tip{-1}}}}}}{\hbox{\lx@xy@droprule}}{\hbox{\lx@xy@droprule}}{\hbox{\kern 14.0pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\hbox{\kern 1.0pt\raise-2.5pt\hbox{$\textstyle{}$}}}}}}}\ignorespaces}}}}\ignorespaces\mathbb{P}^{1} . Here, instead of a primitive ample h one considers the pull-back of {\cal O}(1) , which is primitive but of course not ample. The Tate–Šafarevič group \Sha(S/\mathbb{P}^{1}) parametrises by definition all K3 twists of S\lx@xy@svg{\hbox{\raise 2.5pt\hbox{\kern 1.0pt\hbox{\ignorespaces\ignorespaces\ignorespaces\hbox{\vtop{\halign{\entry@#!@&&\entry@@#!@\cr&\crcr}}}\ignorespaces{\hbox{\kern-1.0pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\hbox{\kern 1.0pt\raise-2.5pt\hbox{$\textstyle{\ignorespaces\ignorespaces\ignorespaces\ignorespaces}$}}}}}}}\ignorespaces\ignorespaces\ignorespaces\ignorespaces{}{\hbox{\lx@xy@droprule}}\ignorespaces{\hbox{\kern 14.0pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\lx@xy@tip{1}\lx@xy@tip{-1}}}}}}{\hbox{\lx@xy@droprule}}{\hbox{\lx@xy@droprule}}{\hbox{\kern 14.0pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\hbox{\kern 1.0pt\raise-2.5pt\hbox{$\textstyle{}$}}}}}}}\ignorespaces}}}}\ignorespaces\mathbb{P}^{1} and, according to Artin–Tate [ 25 ] and Grothendieck [ 11 ] , there exists an isomorphism \Sha(S/\mathbb{P}^{1})\simeq{\rm Br}(S) . In this note, we complement the cohomological definition of \Sha(S,h) by the following geometric characterisation. Theorem 1.1 . Assume (S,h) is a complex K3 surface together with an ample and primitive class h . Then, there exists a natural bijection between the Tate–Šafarevič group \Sha(S,h) and the set of isomorphism classes of {\rm Pic}^{0}({\cal C}_{\eta}) -torsors admitting a ‘good’ hyperkähler model \textstyle{\Sha(S,h)\ignorespaces\ignorespaces\ignorespaces\ignorespaces} \textstyle{\{\,A={\rm Pic}^{0}({\cal C}_{\eta})\text{\rm-torsor}\mid\exists~A\subset X\text{ \rm good HK model}\,\}.} By definition, a good hyperkähler model compactifies A\lx@xy@svg{\hbox{\raise 2.5pt\hbox{\kern 1.0pt\hbox{\ignorespaces\ignorespaces\ignorespaces\hbox{\vtop{\halign{\entry@#!@&&\entry@@#!@\cr&\crcr}}}\ignorespaces{\hbox{\kern-1.0pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\hbox{\kern 1.0pt\raise-2.5pt\hbox{$\textstyle{\ignorespaces\ignorespaces\ignorespaces\ignorespaces}$}}}}}}}\ignorespaces\ignorespaces\ignorespaces\ignorespaces{}{\hbox{\lx@xy@droprule}}\ignorespaces{\hbox{\kern 14.0pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\lx@xy@tip{1}\lx@xy@tip{-1}}}}}}{\hbox{\lx@xy@droprule}}{\hbox{\lx@xy@droprule}}{\hbox{\kern 14.0pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\hbox{\kern 1.0pt\raise-2.5pt\hbox{$\textstyle{}$}}}}}}}\ignorespaces}}}}\ignorespaces\eta to a Lagrangian fibration X\lx@xy@svg{\hbox{\raise 2.5pt\hbox{\kern 1.0pt\hbox{\ignorespaces\ignorespaces\ignorespaces\hbox{\vtop{\halign{\entry@#!@&&\entry@@#!@\cr&\crcr}}}\ignorespaces{\hbox{\kern-1.0pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\hbox{\kern 1.0pt\raise-2.5pt\hbox{$\textstyle{\ignorespaces\ignorespaces\ignorespaces\ignorespaces}$}}}}}}}\ignorespaces\ignorespaces\ignorespaces\ignorespaces{}{\hbox{\lx@xy@droprule}}\ignorespaces{\hbox{\kern 14.0pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\lx@xy@tip{1}\lx@xy@tip{-1}}}}}}{\hbox{\lx@xy@droprule}}{\hbox{\lx@xy@droprule}}{\hbox{\kern 14.0pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\hbox{\kern 1.0pt\raise-2.5pt\hbox{$\textstyle{}$}}}}}}}\ignorespaces}}}}\ignorespaces|h| with all fibres required to have at least one generically reduced component, see § 5.1 for the precise definition and further details. Quite possibly, under our assumptions, every torsor admitting a compactification by some smooth projective hyperkähler variety at all also admits a hyperkähler compactification that is good in our sense. The result has two parts (see also § 6.1 ): (i) The {\rm Pic}^{0}({\cal C}_{\eta}) -torsors parametrised by \Sha(S,h) are pairwise non-isomorphic. This will be addressed in § 4 . (ii) Among all {\rm Pic}^{0}({\cal C}_{\eta}) -torsors, one obtains exactly those that admit a geometrically satisfying compactification, see ( 2.1 ) below for the well-known analogue of elliptic K3 surfaces. This part will be dealt with in § 5 . The injectivity statement (i) relies on the techniques developed in [ 17 ] (and could have been proved there already), but for the the surjectivity in (ii) we make extensive use of recent work of Kim [ 19 ] . Acknowledgment. The first author wishes to thank the organisers and the participants of the program ‘Arithmetic geometry of K3 surfaces’ at the Bernoulli Center for the invitation and for questions and comments. We wish to thank the anonymous referee of [ 17 ] whose numerous comments and constructive suggestions have been invaluable also for writing this paper. We are most grateful to Yoonjoo Kim for helpful comments on a first version of the present note. 2. Review: Elliptic K3 surfaces In order to motivate our results and to increase readability, we briefly recall the notion of the Tate–Šafarevič group of an elliptic K3 surface, emphasising the modular perspective. 2.1. Classical theory We start with an elliptic K3 surface S_{0}\lx@xy@svg{\hbox{\raise 2.5pt\hbox{\kern 1.0pt\hbox{\ignorespaces\ignorespaces\ignorespaces\hbox{\vtop{\halign{\entry@#!@&&\entry@@#!@\cr&\crcr}}}\ignorespaces{\hbox{\kern-1.0pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\hbox{\kern 1.0pt\raise-2.5pt\hbox{$\textstyle{\ignorespaces\ignorespaces\ignorespaces\ignorespaces}$}}}}}}}\ignorespaces\ignorespaces\ignorespaces\ignorespaces{}{\hbox{\lx@xy@droprule}}\ignorespaces{\hbox{\kern 14.0pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\lx@xy@tip{1}\lx@xy@tip{-1}}}}}}{\hbox{\lx@xy@droprule}}{\hbox{\lx@xy@droprule}}{\hbox{\kern 14.0pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\hbox{\kern 1.0pt\raise-2.5pt\hbox{$\textstyle{}$}}}}}}}\ignorespaces}}}}\ignorespaces\mathbb{P}^{1} over an algebraically closed field k , which one might as well assume to be \mathbb{C} , as we will apply Hodge theory eventually. Note that by definition, an elliptic K3 surface comes with a section. If we only assume that the smooth fibres are isomorphic to elliptic curves without fixing the origin by means of a section then we will speak of genus one fibred K3 surfaces. A twist of S_{0}\lx@xy@svg{\hbox{\raise 2.5pt\hbox{\kern 1.0pt\hbox{\ignorespaces\ignorespaces\ignorespaces\hbox{\vtop{\halign{\entry@#!@&&\entry@@#!@\cr&\crcr}}}\ignorespaces{\hbox{\kern-1.0pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\hbox{\kern 1.0pt\raise-2.5pt\hbox{$\textstyle{\ignorespaces\ignorespaces\ignorespaces\ignorespaces}$}}}}}}}\ignorespaces\ignorespaces\ignorespaces\ignorespaces{}{\hbox{\lx@xy@droprule}}\ignorespaces{\hbox{\kern 14.0pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\lx@xy@tip{1}\lx@xy@tip{-1}}}}}}{\hbox{\lx@xy@droprule}}{\hbox{\lx@xy@droprule}}{\hbox{\kern 14.0pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\hbox{\kern 1.0pt\raise-2.5pt\hbox{$\textstyle{}$}}}}}}}\ignorespaces}}}}\ignorespaces\mathbb{P}^{1} consists of a K3 surface S together with a genus one fibration S\lx@xy@svg{\hbox{\raise 2.5pt\hbox{\kern 1.0pt\hbox{\ignorespaces\ignorespaces\ignorespaces\hbox{\vtop{\halign{\entry@#!@&&\entry@@#!@\cr&\crcr}}}\ignorespaces{\hbox{\kern-1.0pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\hbox{\kern 1.0pt\raise-2.5pt\hbox{$\textstyle{\ignorespaces\ignorespaces\ignorespaces\ignorespaces}$}}}}}}}\ignorespaces\ignorespaces\ignorespaces\ignorespaces{}{\hbox{\lx@xy@droprule}}\ignorespaces{\hbox{\kern 14.0pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\lx@xy@tip{1}\lx@xy@tip{-1}}}}}}{\hbox{\lx@xy@droprule}}{\hbox{\lx@xy@droprule}}{\hbox{\kern 14.0pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\hbox{\kern 1.0pt\raise-2.5pt\hbox{$\textstyle{}$}}}}}}}\ignorespaces}}}}\ignorespaces\mathbb{P}^{1} and an open immersion {\rm Pic}^{0}(S/\mathbb{P}^{1}_{\rm sm})\,\lx@xy@svg{\hbox{\raise 2.5pt\hbox{\kern 1.0pt\hbox{\ignorespaces\ignorespaces\ignorespaces\hbox{\vtop{\halign{\entry@#!@&&\entry@@#!@\cr&\crcr}}}\ignorespaces{\hbox{\kern-1.0pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\hbox{\kern 1.0pt\raise-2.5pt\hbox{$\textstyle{\ignorespaces\ignorespaces\ignorespaces\ignorespaces}$}}}}}}}\ignorespaces\ignorespaces\ignorespaces\ignorespaces{\hbox{\kern 1.0pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\lx@xy@hook{1}}}}}}{\hbox{\lx@xy@droprule}}\ignorespaces{\hbox{\kern 14.0pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\lx@xy@tip{1}\lx@xy@tip{-1}}}}}}{\hbox{\lx@xy@droprule}}{\hbox{\lx@xy@droprule}}{\hbox{\kern 14.0pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\hbox{\kern 1.0pt\raise-2.5pt\hbox{$\textstyle{}$}}}}}}}\ignorespaces}}}}\ignorespaces S_{0} of the relative Jacobian of its smooth part S\lx@xy@svg{\hbox{\raise 2.5pt\hbox{\kern 1.0pt\hbox{\ignorespaces\ignorespaces\ignorespaces\hbox{\vtop{\halign{\entry@#!@&&\entry@@#!@\cr&\crcr}}}\ignorespaces{\hbox{\kern-1.0pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\hbox{\kern 1.0pt\raise-2.5pt\hbox{$\textstyle{\ignorespaces\ignorespaces\ignorespaces\ignorespaces}$}}}}}}}\ignorespaces\ignorespaces\ignorespaces\ignorespaces{}{\hbox{\lx@xy@droprule}}\ignorespaces{\hbox{\kern 14.0pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\lx@xy@tip{1}\lx@xy@tip{-1}}}}}}{\hbox{\lx@xy@droprule}}{\hbox{\lx@xy@droprule}}{\hbox{\kern 14.0pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\hbox{\kern 1.0pt\raise-2.5pt\hbox{$\textstyle{}$}}}}}}}\ignorespaces}}}}\ignorespaces\mathbb{P}^{1}_{\rm sm} . In other words, we fix an isomorphism of the minimal smooth compactification of the relative Jacobian {\rm Pic}^{0}(S/\mathbb{P}^{1}_{\rm sm}) with S_{0} relative over \mathbb{P}^{1} . By definition, the Tate–Šafarevič group is the group of all twists: \Sha(S_{0}/\mathbb{P}^{1})\coloneqq\{\,\text{twists}\,\}/_{\simeq}. At this point, it is simply a set of isomorphism classes. Its group structure becomes apparent by observing that any twist S\lx@xy@svg{\hbox{\raise 2.5pt\hbox{\kern 1.0pt\hbox{\ignorespaces\ignorespaces\ignorespaces\hbox{\vtop{\halign{\entry@#!@&&\entry@@#!@\cr&\crcr}}}\ignorespaces{\hbox{\kern-1.0pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\hbox{\kern 1.0pt\raise-2.5pt\hbox{$\textstyle{\ignorespaces\ignorespaces\ignorespaces\ignorespaces}$}}}}}}}\ignorespaces\ignorespaces\ignorespaces\ignorespaces{}{\hbox{\lx@xy@droprule}}\ignorespaces{\hbox{\kern 14.0pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\lx@xy@tip{1}\lx@xy@tip{-1}}}}}}{\hbox{\lx@xy@droprule}}{\hbox{\lx@xy@droprule}}{\hbox{\kern 14.0pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\hbox{\kern 1.0pt\raise-2.5pt\hbox{$\textstyle{}$}}}}}}}\ignorespaces}}}}\ignorespaces\mathbb{P}^{1} of S_{0}\lx@xy@svg{\hbox{\raise 2.5pt\hbox{\kern 1.0pt\hbox{\ignorespaces\ignorespaces\ignorespaces\hbox{\vtop{\halign{\entry@#!@&&\entry@@#!@\cr&\crcr}}}\ignorespaces{\hbox{\kern-1.0pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\hbox{\kern 1.0pt\raise-2.5pt\hbox{$\textstyle{\ignorespaces\ignorespaces\ignorespaces\ignorespaces}$}}}}}}}\ignorespaces\ignorespaces\ignorespaces\ignorespaces{}{\hbox{\lx@xy@droprule}}\ignorespaces{\hbox{\kern 14.0pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\lx@xy@tip{1}\lx@xy@tip{-1}}}}}}{\hbox{\lx@xy@droprule}}{\hbox{\lx@xy@droprule}}{\hbox{\kern 14.0pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\hbox{\kern 1.0pt\raise-2.5pt\hbox{$\textstyle{}$}}}}}}}\ignorespaces}}}}\ignorespaces\mathbb{P}^{1} defines a torsor of the generic fibre E_{0}=S_{0\eta} . More precisely, \Sha(S_{0}/\mathbb{P}^{1}) can be realised as a subgroup of the Weil–Châtelet group of E_{0} : \Sha(S_{0}/\mathbb{P}^{1})\,\lx@xy@svg{\hbox{\raise 2.5pt\hbox{\kern 1.0pt\hbox{\ignorespaces\ignorespaces\ignorespaces\hbox{\vtop{\halign{\entry@#!@&&\entry@@#!@\cr&\crcr}}}\ignorespaces{\hbox{\kern-1.0pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\hbox{\kern 1.0pt\raise-2.5pt\hbox{$\textstyle{\ignorespaces\ignorespaces\ignorespaces\ignorespaces}$}}}}}}}\ignorespaces\ignorespaces\ignorespaces\ignorespaces{\hbox{\kern 1.0pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\lx@xy@hook{1}}}}}}{\hbox{\lx@xy@droprule}}\ignorespaces{\hbox{\kern 14.0pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\lx@xy@tip{1}\lx@xy@tip{-1}}}}}}{\hbox{\lx@xy@droprule}}{\hbox{\lx@xy@droprule}}{\hbox{\kern 14.0pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\hbox{\kern 1.0pt\raise-2.5pt\hbox{$\textstyle{}$}}}}}}}\ignorespaces}}}}\ignorespaces\text{WC}(E_{0})\simeq H^{1}(\eta,E_{0}). It is known that the fibres of S_{0}\lx@xy@svg{\hbox{\raise 2.5pt\hbox{\kern 1.0pt\hbox{\ignorespaces\ignorespaces\ignorespaces\hbox{\vtop{\halign{\entry@#!@&&\entry@@#!@\cr&\crcr}}}\ignorespaces{\hbox{\kern-1.0pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\hbox{\kern 1.0pt\raise-2.5pt\hbox{$\textstyle{\ignorespaces\ignorespaces\ignorespaces\ignorespaces}$}}}}}}}\ignorespaces\ignorespaces\ignorespaces\ignorespaces{}{\hbox{\lx@xy@droprule}}\ignorespaces{\hbox{\kern 14.0pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\lx@xy@tip{1}\lx@xy@tip{-1}}}}}}{\hbox{\lx@xy@droprule}}{\hbox{\lx@xy@droprule}}{\hbox{\kern 14.0pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\hbox{\kern 1.0pt\raise-2.5pt\hbox{$\textstyle{}$}}}}}}}\ignorespaces}}}}\ignorespaces\mathbb{P}^{1} and of any twist