This manuscript presents new trajectory datasets for the surface, planetary boundary layer, and clouds, complementing existing ARM measurements. While the integration of HYSPLIT with ERA5 data is not novel, as it has been employed in previous studies, this paper uniquely synthesizes ARM radiosonde, radar and lidar data to enhance our understanding and potentially benefit future aerosol and cloud research. The study currently focuses on the EPCAPE site during 2023-2024. It is valuable to expand the datasets to more ARM sites and extend the temporal coverage. I believe these datasets will be of great interest to the aerosol, cloud or aerosol-cloud interaction community, as well as useful for field campaigns at ARM sites. The paper is overall well-structured and clearly written.

The primary concern with the manuscript is the inadequate assessment of uncertainties and the lack of a comprehensive quality check of the datasets. Although ensemble runs are run to indicate uncertainty, this primarily addresses uncertainties associated with the coarse resolution of meteorological fields. I recommend that the authors elaborate on uncertainties in the ERA5 data, uncertainties in the ARM measurements used to initiate the HYSPLIT, and propagation of uncertainties through the modeling processes. Additionally, it is not clear for example, whether the trajectory products on land cover and thermodynamic variables exhibit similar uncertainty levels, and how these uncertainties vary with the length of the backward or forward trajectory period (e.g., 2 days vs. 10 days).  Addressing these aspects is crucial for users applying these datasets and interpreting results.

Some specific comments:

• Introduction: The manuscript adequately explains the utility of ARM observations and air mass trajectory analysis. However, given the common usage of the HYSPLIT model in prior studies, it would be beneficial for the authors to clarify what unique insights these new datasets provide, and what specific scientific questions they enable users to address beyond the capabilities of traditional HYSPLIT applications.
• 2.1 Surface Trajectory Dataset: There is a potential issue with trajectories initiating at low altitudes as they may hit the ground and lose accuracy. Have the authors observed this issue in their datasets? A discussion on the impact of terrain on data quality would enhance this section.
• Table 1: Identification of potential applications for backward and forward ARMTRAJ-CLD and ARMTRAJ-ARSCL trajectories would provide clarity and usefulness to the readers.
• #198: The mention of “the resulting 8 sets of cloud deck base, top, and free-tropospheric heights” is unclear. For better context, consider placing the sentence from paragraph #190 before this statement.
• Figure 4, please include standard deviation of these numbers.