Recent advancements in high-fidelity process simulations have significantly deepened our understanding of melt pool behavior in laser-based powder bed fusion (PBF-LB) of metals. However, ceramics present unique challenges due to their distinct material properties and complex thermophysical responses, which complicate the development of detailed models that can effectively supplement limited experimental data. In this study, we combine operando synchrotron X-ray tomographic microscopy with computational fluid dynamics–discrete element method (CFD-DEM) simulations to explore melt pool dynamics in alumina during PBF-LB[1]. In-situ observations reveal a melt pool that is shallow and wide—markedly different from metallic systems. This morphology is attributed to alumina’s low thermal conductivity, high laser absorption, and the presence of negative surface tension gradients, as confirmed by high-fidelity simulations. We identify key mechanisms driving this behavior, including limited thermal penetration, intensified surface-directed convection, and vortex formation at the melt pool surface due to heat exchange with the surrounding gas environment. By systematically analyzing the effects of laser power and scanning speed on melt pool geometry, we establish the first virtual process map and define parameter windows for stable PBF-LB processing of alumina. These findings offer critical insights into process optimization and represent a significant step toward advancing ceramic additive manufacturing.