A remark on generalized abundance for surfaces (Q6145007)

\textit{V. Lazić} and \textit{T. Peternell} [Publ. Res. Inst. Math. Sci. 56, No. 2, 353--389 (2020; Zbl 1466.14019)] formulated the Generalized Abundance Conjecture stating that, given a klt pair \((X,\Delta)\) such that \(K_X+\Delta\) is pseudoeffective and a nef line bundle \(L\) such that \(K_X+\Delta+L\) is nef, there is a semi-ample \(\mathbb Q\)-divisor \(M\) numerically equivalent to \(K_X+\Delta+L\). [loc. cit.] also contains a proof of the conjecture in the case of surfaces. The paper under review contains a variation of the above result in dimension 2, proving that, if one drops the assumption that \(K_X+\Delta\) is pseudoeffective and assumes instead that that the numerical equivalence classes of \(K_X+\Delta\) and \(L\) are independent, then \(K_X+\Delta+L\) is nef. An example (due to Lazić) shows that an analogous result does not hold in dimension >2.