Estimate of the conformal scalar curvature equation via the method of moving planes. II (Q1281881)

This paper is motivated by the problem of finding a metric conformal to the standard metric of \(\mathbb R^n\) with bounded prescribed scalar curvature \(K(x).\) For a survey of the first achievements on this problem, see the book of \textit{Th. Aubin} [Some Nonlinear Problems in Riemannian Geometry, Springer (1998; Zbl 0896.53003)]. It is worth mentioning that the present paper is both hard and strong. It is strong since the results therein are nontrivial and of evident interest for the theory. However, it is hard because the assumptions and the main goals are mixed altogether, and it is difficult to separate the technicalities from the clarities arised by the results. In this paper the authors consider a sequence of positive \(C^2\) solutions \(u_i\) of the problem bellow \[ \triangle u_i + K_i(x)u_{i{}}^{p_i}\quad \text{ in} B_2 \tag{1} \] where \(K_i(x)\in C^1(\overline{B_2})\) is a sequence of \(C^1\) positive functions defined in \(\overline {B_2},\) the ball with center at \(0\) and radius \(2\) in \(\mathbb R^n\) with \(n\geqslant 3,\) and \(1< p_i\uparrow (n+2)/(n-2).\) Furthermore \(K_i\) is assumed to be bounded between two positive constants. It has been proved that for a sequence of solutions \(u_i\) of equation \((1)\), the blow-up (for a definition see p. 117 of the present paper) does not occur at a noncritical point of \(\{K_i\},\) see \textit{Y. Li} [J. Differ. Equations 120, No. 2, 319-410 (1995; Zbl 0827.53039)]. The authors throughout the paper assume that \(0\) is the only critical point of \(\{K_i\}.\) Unfortunately, it is not convenient to state the whole theorems here for the reasons given above. Nevertheless, to give a taste to the reader, the reviewer will remark that under some suitable analytic conditions the authors were able to prove that the sequence \(u_i\) is uniformly bounded in \(\overline{B_1}\). They also have shown that, under suitable conditions, if \(u_i\) is a sequence of solutions of \((1)\) with \(p_i\leqslant (n+2)/(n-2)\), \(\lim_{i\rightarrow \infty}p_i =(n+2)/(n-2)\) and \(\max_{\overline{B_1}}u_i\rightarrow \infty,\) then \(0\) is an isolated blow-up point. The authors main technique among others is the Alexandrov Reflection Principle, the improvement that has been done (in chronological order) by Serrin, Gidas-Ni-Nirenberg and the refinement of the method of moving planes due to \textit{L. A. Caffarelli, B. Gidas} and \textit{J. Spruck} [Commun. Pure Appl. Math. 42, No. 3, 271-297 (1989; Zbl 0702.35085)]. They also have applied Pokhozhaev's identity, the blowing-up analysis introduced by R. Schoen, some potential theory and hard analytic estimates. For part I see Commun. Pure Appl. Math. 50, 971--1017 (1997); Zbl 0958.35013).