A comparison of HYSPLIT backward trajectories generated from two GDAS datasets

Area of Science:

• Atmospheric Science
• Environmental Science
• Meteorology

Background:

• The Hybrid Single-Particle Lagrangian Integrated Trajectory (HYSPLIT) model is crucial for determining air mass origins.
• Discrepancies in HYSPLIT trajectories arise from variations in input meteorological datasets, particularly the Global Data Assimilation System (GDAS).

Purpose of the Study:

• To quantify differences in HYSPLIT backward trajectories generated using two GDAS datasets (GDAS1 and GDAS0P5) with varying resolutions.
• To identify the primary reasons for trajectory discrepancies, focusing on vertical velocity and data resolution impacts.

Main Methods:

• Generated HYSPLIT backward trajectories for Hong Kong throughout 2011 using GDAS1 and GDAS0P5 datasets.
• Analyzed trajectory differences, attributing them to vertical velocity calculation methods and meteorological data resolutions.
• Employed cluster analysis to assess trajectory sensitivity based on air mass origin and range.

Main Results:

• Significant trajectory differences were primarily caused by distinct vertical velocity calculation methods, especially the absence of vertical velocity in GDAS0P5.
• HYSPLIT trajectories showed sensitivity to horizontal and vertical resolutions, though to a lesser extent than vertical velocity.
• Trajectory sensitivity varied with air mass origin: north, northeast, and west trajectories were more affected by vertical velocity and resolution changes.

Conclusions:

• The GDAS1 dataset, incorporating vertical velocity, is more suitable for HYSPLIT backward trajectory analysis in the Pearl River Delta region.
• Understanding data resolution and vertical velocity impacts is critical for accurate air mass source apportionment studies.