The strongly intensive observable , constructed from charged-pion multiplicities in separated forward (F) and backward (B) pseudorapidity intervals , is investigated within the UrQMD transport model for Pb–Pb collisions at SPS energies. The dependence of on is studied as a function of collision energy, centrality, and acceptance to explore the longitudinal structure of fluctuations and correlations in particle production. For GeV, exhibits a non-monotonic behavior, increasing from values close to unity at small , reaching a maximum at intermediate separation, and subsequently decreasing towards unity at large . A decomposition of into scaled variance and covariance terms shows that this behavior originates from their different dependences. Qualitatively, exhibits similar behaviour when rapidity is used in place of , demonstrating the robustness of the observed structure against the choice of longitudinal variable. With decays of and mesons disabled, is moderately increased in the small range revealing that the selected resonance decays partly contribute to the short-range component of the correlations, while the persistence of at larger indicates fluctuation and correlation effects extending over broader pseudorapidity intervals. The magnitude of , together with its fluctuation and covariance components, increases from central to peripheral collisions, with the fluctuation term dominating over the covariance contribution at large . With increasing collision energy, the pseudorapidity interval over which remains above unity becomes significantly larger, whereas at lower energies, where resonance dynamics dominate particle production, exhibits an approximately flat dependence on . An analytical acceptance-scaling relation for is derived under the assumption of independent binomial sampling and validated using a random-thinning procedure within a fixed pseudorapidity window, showing very good agreement between the analytical expectation and the direct calculation from randomly thinned events. These results demonstrate the sensitivity of to the interplay between multiplicity fluctuations and forward–backward correlations in relativistic heavy-ion collisions.