As wildfires increase in frequency and intensity due to changing climatic conditions, accurate and timely forecasting of wildfire smoke becomes crucial for reducing public health risks, enhancing weather predictions, and supporting firefighting efforts. Traditional smoke forecasting models often rely on a persistence approach, which adjusts the latest satellite-based estimates of biomass burning emissions to account for a climatologically driven diurnal cycle (e.g., HRRR-Smoke). These models, however, typically do not account for the complex meteorological influences on wildfire behavior and smoke emissions, often simplifying such influences to effects such as precipitation. To bridge this gap, NOAA's Global Systems Laboratory is utilizing the experimental weather forecast model, the Rapid-Refresh Forecasting System (RRFS), which covers the domains of North and Central America at a 3 km grid resolution. This system, designated as RRFS-Smoke-Dust, incorporates 3D transport and mixing of smoke emitted from biomass burning, utilizing satellite fire radiative power (FRP) data. Within this framework, we leverage an innovative Hourly Wildfire Potential (HWP) index that integrates predicted variables such as 10-meter wind gusts, surface dewpoint depression, precipitation, and soil moisture availability. This integration assists in predicting the temporal evolution of FRP and refining persistence-based wildfire emissions forecasts. This new approach was tested in the RRFS-Smoke-Dust model for the 2019 fire season, coinciding with the FIREX-AQ campaign, and was assessed using ground-based PM2.5, AOD, and in-situ aircraft measurements. Preliminary results demonstrate that integrating the HWP index into the RRFS-Smoke-Dust model improves the accuracy of predicting smoke emissions compared to the traditional persistence method.