PacificNorth American PNA Pattern Monthly means PNA Winter
















- Slides: 16

Pacific-North American (PNA) Pattern (Monthly means) � PNA Winter correlated with globally gridded temperatures (Jan-Feb-Mar (JFM) averages 1950 -2011)

The test for significant correlation In this example the T value is conform with the null hypothesis x that there is no correlation. We accept H 0 at the chosen two sided significance test-level of p (for example 0. 05). Calculated T-value

The test for significant correlation In this example the T value is unusually far in the tails of x theoretical T-distribution for a zero correlation. We reject H 0 at the chosen two sided significance test-level of p (for example 0. 05) Calculated T-value and say: there is a significant correlation at the (two-sided) 5 -% significance level

Pacific-North American (PNA) Pattern (Monthly means) � PNA Winter correlated with globally gridded temperatures (Jan-Feb-Mar (JFM) averages 1950 -2011) Masked out non-significant correlations (two-sided t-test at 5% level)

THE ROLE OF SAMPLE SIZE FOR SIGNIFICANCE TESTS (CORRELATION) �

Time series analysis ENSO Index: Tropical Pacific Sea Surface Temperatures Seasonal forcast outlook: El Nino is coming!

EXAMPLES OF TIME SERIES 7 days Source: http: //keelingcurve. ucsd. edu/ (retrieved 2014 -04 -30)

EXAMPLES OF TIME SERIES 1 week 12 months Source: http: //keelingcurve. ucsd. edu/ (retrieved 2014 -04 -30)

EXAMPLES OF TIME SERIES 1 year 60 years Source: http: //keelingcurve. ucsd. edu/ (retrieved 2014 -04 -30)

EXAMPLES OF TIME SERIES 60 years 300 years Source: http: //keelingcurve. ucsd. edu/ (retrieved 2014 -04 -30)

EXAMPLES OF TIME SERIES 800 ? Source: http: //keelingcurve. ucsd. edu/ (retrieved 2014 -04 -30)

EXAMPLES OF TIME SERIES Reconstructions from air bubbles trapped in Antarctic Ice Cores 300 years 800, 000 years Source: http: //keelingcurve. ucsd. edu/ (retrieved 2014 -04 -30)

SAMPLING RATES AND TIME SERIES LENGTH � Atmospheric Environmental sciences study processes on a wide range of time scales! from tenth of seconds in turbulence (or even less, e. g. lightning processes) Weather (hours to weeks) � Seasonal variability (weeks to months) � Climate variability (years-decades) � Ice age cycles (10 -100, 000 years) � Geological processes (plate tectonics) > 10, 000 yrs �

SAMPLING RATES AND TIME SERIES LENGTH � Atmospheric Environmental sciences study processes on a wide range of time scales! � The sampling rate at which we observe these processes must match the typical time-scales of the variations

Aliasing � http: //youtu. be/Ukot. Zy 3 l. Qqo

R-SCRIPTS � scripts: � � � � & DATA rmean. R timeseries_noise 1. R timeseries_record 1. R timeseries_pna_daily. R timeseries_co 2_800000. R timeseries_insolation_800000. R data: � � � � norm. daily. pna. index. b 500101. current. ascii 20140226. csv timeseries_co 2_800000. asc timeseries_insolation_JJA_800000. asc sound_staple 1. csv sound_plasticbag 1. csv sound_noisy_tone 1. csv sound_water_fountain_sample. csv