Cluster Analysis for ENSO Events using HYSPLIT


The previous cursory examination of the decadal cluster mean trajectories suggested that there might be an ENSO effect in the tropical Pacific region. To determine what such trajectory clusters might look like, the existing database of trajectories was partitioned into two periods, months that showed either an El Niño or La Niña influence. This was determined by using the existing partition developed by the NOAA's Physical Sciences Laboratory for the top 24 strongest events. The cluster analysis was limited to the years 1950 through 2014. The calculations shown here are for only one location: 0N 150W.

The two tables and figures shown below are for the El Niño periods on the left and the La Niña periods on the right. In both cases the flow was dominated by the eastern trades with 92% of the trajectories during El Niño periods and 98% of the trajectories during La Niña periods. The remaining flow periods showed much more directional variability during El Niño and with much lower wind speeds, averaging 8.7 m/s for cluster #1, compared with 12.0 m/s for cluster #2 during the La Niña months.

The trajectories for this analysis had already been precomputed for the decadal clustering study. The data extraction and clustering for other locations using the ENSO filter in not currently available through the web interface.



El Niño (167 months) La Niña (165 months)
Cluster
Numb
Percent
 
Duration
(h)
Temp
(oC)
Press
(hPa)
H2O
(g/kg)
Wspd
(m/s)
191.8634.016.4849.310.28.7
22.324.117.1849.89.73.8
32.263.217.4849.910.44.4
41.021.617.3849.89.93.5
52.722.717.0849.99.94.0
Mean100.0261.916.5849.310.28.3
Cluster
Numb
Percent
 
Duration
(h)
Temp
(oC)
Press
(hPa)
H2O
(g/kg)
Wspd
(m/s)
10.320.515.4850.09.85.0
298.01040.314.8848.89.412.9
31.415.815.2850.28.710.1
40.213.715.3849.79.14.8
50.130.014.6848.69.43.7
Mean100.0562.114.8848.99.412.8