!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! !!!!!!!!!!!!! MJO Simulation Diagnostics - LEVEL 1 !!!!!!!!!!!!!! !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! ! History ! [Created] Daehyun Kim - 20May2007 ! [Revised] Daehyun Kim - 29Apr2008 ! Can be applied to the model which has 360 day calendar ! leap.eq.2 => leap.eq.0 (in the c-shell script) ! Description ! This document contains descriptions for ! the calculating & plotting system of level_1 MSD ! (MJO Simulation Diagnostics). ! Please visit ! http://climate.snu.ac.kr/mjo_diagnostics/index.htm ! to see the figures and detailed descriptions of level_1 MSD. ! The upper limit directory of this system is 'msd'. ! This document will give descriptions starting from this ! directory. (e.g. msd/level_1/sample/ano/ano.sh) ! $var : {variable}_{source of data} ! (e.g. olr_av, olr_cam, u850_snu) !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! # Before the calculation & plotting The daily data should be located in the directory msd/level_1/$var/data *Fortran binary data, direct access (e.g. msd/level_1/olr_av/data/daily.19790101_20051231.gdat) Example data provided are olr_av, u850_n1 and u200_n1. olr_av : OLR (AVHRR) u850_n1 : 850hPa zonal wind (NCEP1) u200_n1 : 200hPa zonal wind (NCEP1) !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! NOW, YOU HAVE DATA IN A PROPER DIRECTORY. LET'S START! 1. Variance Maps * Order of calculations & plotting a. ano : calculate daliy climatology and anomaly Calculation :: msd/level_1/csh/ano/ano.sh Results :: msd/level_1/$var/data/daily.anom.${period}.gdat ===> anomaly data obtained by subtracting climatological daily annual cycle b. filter : filtering daily anomaly using Lanczos filter Calculation :: msd/level_1/csh/filter/filter.sh Results :: msd/level_1/$var/data/daily.filt.20-100.lanz.100.period.gdat ===> 20-100 day filtered data using 201-points Lanczos filter c. var : calculate variance using anomaly and filtered data Calculation :: msd/level_1/csh/var/var.sh Results :: msd/level_1/$var/var/raw.[all/win/sum].gdat :: msd/level_1/$var/var/fil.[all/win/sum].gdat Plotting :: msd/level_1/csh/var/fig.sh Figures :: msd/level_1/fig/var/$var/all/*gif - all season msd/level_1/fig/var/$var/sum/*gif - summer msd/level_1/fig/var/$var/win/*gif - winter 2. Time Series Power Spectra * Order of calculations & plotting a. aave : area averging before calculate power spectra Calculation :: msd/level_1/csh/aave/aave.sh Results :: msd/level_1/$var/data/${region}.[win/sum].series ===> area averaged time series b. tsps : calculate power spectra using area averaged data Calculation :: msd/level_1/csh/tsps/tsps.sh Results :: msd/level_1/$var/tsps/${region}.[win/sum] ===> power spectra Plotting :: msd/level_1/csh/tsps/fig.sh Figures :: msd/level_1/fig/tsps/$var/${region}.[win/sum].gif 3. EOF's * Order of calculations & plotting a. anom : go to 1.a b. filter : go to 1.b c. 5x5 : interpolate filtered (1.b) data to have 5'x5' resolution Calculation :: msd/level_1/csh/5x5/5x5.sh Results :: msd/level_1/$var/data/daily.5x5.anom.${period}.gdat ===> interpolated anomaly data (for projected PC) :: msd/level_1/$var/data/daily.5x5.filt.20-100.lanz.100.${period}.gdat ===> interpolated filtered data (for eof) d. eof : calculate EOFs (empirical orthogonal function) Calculation :: msd/level_1/csh/eof/eof.sh Results :: msd/level_1/$var/eof/[win/sum]/eof.pct ===> percentage variance :: msd/level_1/$var/eof/[win/sum]/eof.pct.gdat ===> used in percentage variance plotting :: msd/level_1/$var/eof/[win/sum]/eof.ev ===> eigen vectors :: msd/level_1/$var/eof/[win/sum]/eof.ts ===> PC time series :: msd/level_1/$var/eof/[win/sum]/eof.ts.pr ===> projcected PC time series e. pcps : calculate power spectra using projected PCs (3.b) Calculation :: msd/level_1/csh/eof/pcps.sh Results :: msd/level_1/$var/eof/[win/sum]/pcps.ts0[1/2/3/4/5].[win/sum] ===> power spectra from projcected PC time series Plotting :: msd/level_1/csh/eof/fig.sh Figures :: msd/level_1/fig/eof/$var/eof.[win/sum].gif ===> eigen vectors :: msd/level_1/fig/eof/$var/pct.[win/sum].gif ===> percentage variance :: msd/level_1/fig/eof/$var/pcps.ts0[1/2/3/4/5].[win/sum].gif ===> power spectra using projected PCs 4. Lag Correlations * Order of calculations & plotting a. ano : go to 1.a b. filter : go to 1.b c. 5x5 : go to 3.d d eof : go to 3.d e. pcl : calculate lead-lag correlation coefficients between PC1 and PC2 from 3.d Calculation :: msd/level_1/csh/pcl/pcl.sh Results :: msd/level_1/$var/eof/[win/sum]/pcl.llreg_2d.gdat ===> lag correlation coefficients Plotting :: msd/level_1/csh/pcl/fig.sh Figures :: msd/level_1/fig/pcl/$var/pcl.[win/sum].gif ===> lag correlation between PC1 and PC2 from EOF f. zm : meridionally/zonally averaging filtered data (1.b) Calculation :: msd/level_1/csh/zm/zm.sh Results :: msd/level_1/$var/data/daily.filt.20-100.lanz.100.10S10N.${period}.gdat ===> meridionally averaged (10S-10N) :: msd/level_1/$var/data/daily.filt.20-100.lanz.100.80E100E.${period}.gdat ===> zonally average (80E-100E) :: msd/level_1/$var/data/daily.filt.20-100.lanz.100.115E135E.${period}.gdat ===> zonally average (115E-135E) g. lgc : calculate lead-lag correlation coefficients between area averaged time series (2.a) and meridionally/zonally averaged filtered data (4.f) Calculation :: msd/level_1/csh/lgc/lgc.sh Results :: msd/level_1/$var/lgc/east.[win/sum].llreg_2d.gdat ===> eastward propagation :: msd/level_1/$var/lgc/north[1/2].llreg_2d.gdat ===> northward propagation (1 = indian ocean, 2 = western Pacific) Plotting :: msd/level_1/csh/lgc/fig.sh Figures :: msd/level_1/fig/lgc/$var/east.[win/sum].gif :: msd/level_1/fig/lgc/$var/north.[io/wp].gif ! * Contact : Daehyun Kim (kim@climate.snu.ac.kr) !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! !!! Last updated in 26May2007 by Daehyun Kim !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!