Data associated with the manuscript entitled "Asymmetric Cooling of the Atlantic and Pacific Arctic during the Past Two Millennia: A Dual Observation-Modeling Study". Transient simulations for the past 2,000 years (past2k) have been proposed for the Paleoclimate Model Intercomparison Phase 4 (PMIP4) contribution to the Coupled Model Intercomparison Project Phase 6 (CMIP6) (Jungclaus et al., 2017). Forcing data for the past2k experiments was compiled by PMIP4. However, there was no land-cover change available for prior to 850 CE when this past2k CESM simulation was started. We therefore constructed a land cover forcing (Dataset1.tar) for this past2k simulation from 1 CE to 849 CE as a superposition of HYDE3.1 cropland changes (Goldewijk et al. 2010, 2011) onto the land cover forcing for the CESM Last Millennium Ensemble runs (LME, Otto-Blisner et al. 2016) for 850 CE. Land cover forcing for this past2k CESM simulation from 850 CE to 2005 CE is taken from LME, which implements grassland changes in addition to cropland changes as compiled by PMIP3. PMIP4 Easy Volcanic Aerosol data (EVA, Toohey et al. 2016) was used to generate volcanic aerosol forcing file for the past2k simulation with CESM. EVA provides space-time distribution of volcanic aerosol mass, effective radius, and optical depth. CESM reads in space-time distribution of volcanic aerosol mass, but assumes constant effective radius while computing optical depth. In order for CESM to match EVA optical depth, EVA aerosol mass needed to be scaled up by a factor of 1.67 based on a set of sensitivity experiments for the Tambora eruption in 1815. The forcing file (Dataset2.nc) is included for others to use. Data citation: Zhong, Y., Jahn, A., Miller, G. H., Geirsdottir, A. (2018). Asymmetric Cooling of the Atlantic and Pacific Arctic during the Past Two Millennia: A Dual Observation-Modeling Study. Submitted to GRL.