Quadrature for quadrics (Q6146098)

\textit{H. Shahgholian} [J. Math. Anal. Appl. 171, No. 1, 61--78 (1992; Zbl 0768.31003)] stated the question on why exteriors of ellipses and parabolas in the plane are null-quadrature domains, but the exterior of hyperbolas is not so. This question has been studied in the setting of potential theory and partial differential equations since the 90's. The point of view in the present paper is different, because the author works in terms of real algebraic geometry, more precisely in the so-called multi-sheeted algebraic domains, or quadrature Riemann surfaces. Both fields, partial differential equations and real algebraic geometry, are linked by means of potential theory. So, the aim of the paper becomes the integration of holomorphic functions over domains bounded by algebraic curves of degree 2-quadrics -- in particular ellipses, hyperbolas and parabolas. In this way conic sections and potential theory appear, and the rôle of conic sections in gravitational theory is known since I. Newton. In this paper the author makes a detailed research on quadrature properties of planar domains bounded by curves of degree 2. Since the exterior has infinite area, problems are avoided by using spherical instead of Euclidean area measure. Besides, domains are allowed to be multi-sheeted, that is to say, covered several times by a branched covering map from a real compact Riemann surface having an anticonformal involution whose fixed points disconnect the surface. Boundaries of these quadrature Riemann surfaces are algebraic curves. The study of algebraic curves in projective spaces is the subject of Sections 2 and 3 of the article, and definitions and properties of quadrature domains are considered in Section 4. The main result in this Section is Theorem 4.6, and it is worth noting that the computations leading to it extend and complete those made by the author and \textit{V. G. Tkachev} in [Comput. Methods Funct. Theory 11, No. 2, 591--615 (2011; Zbl 1246.30072)]. Once made all this general work, Sections 5 to 8 are devoted to study systematically the special cases which come from each of the involved curves: the ellipse, the hyperbola, the parabola, and inversions of them. A special relevance has Section 6 on the hyperbola, since this case behaves in a way very different to the others. In particular, hyperbolas are multi-sheeted quadrature domains, and null-quadrature domains when working with Euclidean rather than with spherical area measure. It is also to be noted that the inversions of the three curves (Section 8) are respectively the hippopede, the lemniscate and the cardioid. The paper as a whole gives a deep and clarifying insight on this important topic.