BackTrajectory-ClusterAnalysis

---

This repository houses code to generate cluster-associated representative air parcel trajectories associated with cloud water samples collected at the summit of Whiteface Mountain (NY).

Methods

This repository houses code that generates HYSPLT back trajectories for each hour that the cloud water collector was deployed during a 12-hour cloud event, then finds an average trajectory for each 12-hour cloud water sample. Cluster analysis is then done on trajectories, using partitioning around medoids, to find a representative trajectory that samples could be associated with.

---

(While the data were processed using the R code in this directory, this map and trajectory heights were plotted with Python)

Code Requirements

This code uses several packages from the R library:

• scales
• tidyverse
• lubridate
• tibbletime
• data.table
• splitr
• openair

Running the Code

R scripts need to be run in this order:

1. HysplitMetCleanupScript.R
2. HypslitTrajectoryScript.r
3. CloudClusterProcessingScriptCorrected.R

Reach Out For Met Data Files

---

This code was originally written and compiled by Dr. Christopher Lawrence