Extratropical extreme precipitation events are usually associated with large-scale flow disturbances, strong ascent and large latent heat release. The causal relationships between these factors are often not obvious, however, and the roles of different physical processes in producing the extreme precipitation event can be difficult to disentangle. Here, we examine the large-scale forcings and convective heating feedback in the precipitation events which caused the 2010 Pakistan flood within the Column Quasi-Geostrophic framework. A cloud-revolving model (CRM) is forced with the large-scale forcings (other than large-scale vertical motion) computed from the quasi-geostrophic omega equation with input data from a reanalysis data set, and the large-scale vertical motion is diagnosed interactively with the simulated convection.Numerical results show that the positive feedback of convective heating to large-scale dynamics is essential in amplifying the precipitation intensity to the observed values. Orographic lifting is the most important dynamic forcing in both events, while differential potential vorticity advection also contributes to the triggering of the first event. Horizontal moisture advection modulates the extreme events mainly by setting the environmental humidity, which modulates the amplitude of the convection's response to the dynamic forcings.When the CRM is replaced by either a single-column model (SCM) with parameterized convection or a dry model with a reduced effective static stability, the model results show substantial discrepancies compared with reanalysis data. The reasons for these discrepancies are examined, and the implications for global models and theoretical models are discussed.