Our mail server hiccuped, so I am not sure that this went out to everyone; therefore I am resending it:LSL_LOWPASS is a least-squares linear low pass filter. The first argument is the time series you wish to low-pass filter. The second argument is the filter's cutoff period. That is where the half-amplitude point of the filter is located. If your data are monthly and you want the half-amplitude point of the filter to be at 7 years, then argument 2 would be 7*12=84 months. The third argument is the filter span. The larger you make this, the sharper the filter will be in frequency space, but if your data have discontinuities the filter will ring (have more wiggles) across a larger time range and more points will be lost at the beginning and end of the filter output. If your filter cutoff period is 84 months, then you need a filter span of at least 84 months. See http://ferret.pmel.noaa.gov/Ferret/documentation/users-guide/appendix-a-external-functions/LSL-LOWPASS for more details.I suggest you try some runs like this:my_var = var_ano[l=1:`60*12`,x=40e:120e,y=20s:20n,k=1:20]let var_filter1 = lsl_lowpass(my_var, `7*12`, `7*12`)let var_filter2 = lsl_lowpass(my_var, `7*12`, `2*7*12`)let var_filter3 = lsl_lowpass(my_var, `7*12`, `4*7*12`)let var_filter4 = lsl_lowpass(my_var, `7*12`, `8*7*12`)and plot the filtered results to see how you like them.------------------------------------------------------------------------------------------Edward D. (Ned) Cokelet, Ph.D. OceanographerNOAA/PMEL off: (206) 526-68207600 Sand Point Way NE fax: (206) 526-6485Seattle, WA 98115-6349The contents of this message are mine personally and do not necessarily reflect any position of the Government or the National Oceanic and Atmospheric Administration.