Recognizing the uncertainties associated with fire emission, a crucial factor influencing the fire aerosol predictions, we have initiated studies to improve fire emission for subseasonal to seasonal (S2S) forecasts. Two global aerosol/chemistry forecast models are currently under development and fully coupled with the Unified Forecast System (UFS), encompassing ocean, sea ice, wave and land surface components for S2S forecasts at NOAA. 1) UFS-Aerosols: the second-generation of UFS coupled aerosol system has been collaboratively developed by NOAA and NASA since 2021, which embeds NASA’s 2nd-generation GOCART model in a National Unified Operational Prediction Capability (NUOPC) infrastructure. It is planned to be implemented into the Global Ensemble Forecast System (GEFS) v13.0 for ensemble prototype 5 (EP5) experiments early this year. 2) UFS-Chem: an innovative community model of chemistry online coupled with UFS, which is a wide collaboration between NOAA Oceanic and Atmospheric Research (OAR) laboratories and NCAR. The aerosol component based on the current operational GEFS-Aerosols v12.3, has been implemented into UFS-Chem utilizing the Common Community Physics Package (CCPP) infrastructure with updates to wet deposition, dust and fire emission etc. Both these two global aerosols forecast models include the direct and semi-direct radiative feedback from online aerosols prediction. Various global fire emission data and their ensemble product are employed to quantify the uncertainties associated with fire aerosol predictions. The capabilities of UFS-Aerosols and UFS-Chem in medium-range and S2S predictions of fire aerosol are assessed and compared using observations from reanalysis data, ground-based measurements, and satellite data. Additionally, blending and machine learning methods have been developed to predict fire emission and improve the S2S prediction.