I am trying to get seasonality parameters from MODIS EVI (MOD13Q1) for Indonesia. I have loaded data from 2013-2015, but the plotted time-series is shifted from what I expect. Indonesia has rice-growing seasons peaking in Jan (season 1, Oct-Mar) & Apr (season 2, Mar-Jun) but the plots I'm getting are peaking in Mar and Sep. I am definitely getting two distinct growing seasons, just offset from where I'd expect them.
Data: MODIS MOD13Q1 EVI for 2013-2015 (last 5 values for 2015 copied from 2014). 23 points/yr for 3 years Savitsky-Golay fitting Seasonal parameter 0 (2 seasons/yr)