Extreme precipitation events are becoming more frequent and more intense due to climate change, causing devastating floods across the globe. Accurate and locally relevant prediction of extreme precipitation at both the weather and climate timescales is critically needed to address the increasing risk to infrastructure and human life. This presentation tackles the emergent relationship between extreme precipitation and temperature at the weather timescale and its implication for future predictions. As climate warms, extreme precipitation intensity is expected to increase with temperature exponentially; the rate of increase is subject to a wide range of uncertainties, despite the general expectation of a 7% per degree of warming based on the Clausius-Clapeyron(C-C) relationship. However, at the weather time scale, extreme precipitation intensity decreases when temperature exceeds a certain threshold. This “negative scaling” of extreme precipitation at high temperatures is likely a result of atmospheric moisture limitation under high temperatures. When controlled for surface air saturation deficit, extreme precipitation intensity at the weather time scale increases monotonically with local surface air temperature; the scaling in saturated atmosphere closely follows the C-C ratio, with a very low degree of regional variation. This emergent relationship is not well captured by earth system models; instead, they produce a wide range of scaling ratios in saturated atmosphere that are highly region-dependent. To identify potential sources of model biases, we examine how model resolution and storm types such as mesoscale convective systems may influence the emergent scaling relationship. To understand whether and how such model biases may propagate to influence its future projection, we assess the potential link between the rate of extreme precipitation increase with temperature at the climate timescale and the models’ scaling ratio at the weather timescale. It is suggested that the scaling ratio at saturation can be a useful model performance constraint for both weather forecast and climate projections.