Dear ferretters, I am trying to plot an ocean tracer from an ORCA2.3 output file on a vertical section along a cruise track. I am using SAMPLEXY_CURV for this. Ferret accepts all my commands and plots something, but the data it plots is not what I expect for most latitudes (section_cruise_pix.pdf). I tested this also with another track (where latitudes were strictly monotonically increasing while longitudes zigzagged) in the Pacific. Here the output *is* what I expected. Of course I also plotted the tracer at a specific model grid longitude. This already resulted in slight inconsistencies. I will describe the process below. To give an impression of the tracer concentration I plotted a 1 km depth plot (plot_1km_pix.pdf) with the cruise track on top of it:shade/title="dAl (nM) at 1 km depth"/l=1/z=1000/i=90:150/j=36:124/levels=(0,24,1)(inf) dal[d=1]*1e9*maskdraw[d=3],lon,lat
plot/vs/line/over/d=2/title="Cruise track" lon_data,lat_data Then I did a vertical section plot through the i=125 grid section (plot_grid_pix.pdf):define axis/Y/from_data/name=latit/unit=degrees_north NAV_LAT[d=1,i=125,j=1:148] shade/j=38:123/title="On ORCA2 grid i=125"/levels=(0,24,1)(inf)/vlimits=5000:0:500 RESHAPE( dAl[d=1,i=125]*maskdraw[d=3,i=125], Y[gy=latit]+Z[gz=dAl[d=1]] ) * 1e9
The command RESHAPE converts the ORCA2 tracer dAl to degrees north. On first sight the result looks reasonable. If you look closer you can see a funny artefact and an inconsistency with the 1km plot, namely: - artefact: There is water under land mask around 35S, 0N and 8N. - inconsistency: In the 1km plot (plot_1km_pix.pdf) at 20S [dAl] > 8 nM for all longitudes in the Atlantic, while in the vertical section (plot_grid_pix.pdf) at 20S the value of [dAl] is less than 1 nM. Now for the third problem. I plotted the data over the cruise track with SAMPLEXY_CURV. This takes the same arguments as SAMPLEXY, plus the metric of the ORCA2.3 grid. DEFINE AXIS/Y/UNITS=degrees yaxAtl=LAT_DATA[d=2]LET X_ATL = SAMPLEXY_CURV( DAL[D=1]*MASKDRAW[D=3], LON,LAT, LON_DATA[D=2],LAT_DATA[D=2] )
LET DEG_ATL = RESHAPE( X_ATL, Y[GY=YAXATL] + Z[GZ=DAL[D=1]] )shade/title="Langs cruise track"/levels=(0,24,1)(inf)/vlimits=5000:0:500 deg_Atl*1e9
Data set 2 is attached as Atl-coords.csv. I do the RESHAPE here again just like in the previous case. If I plot X_ATL immediately, I get more or less the same result: the strange seemingly arbitrary interlacing that you see in section_cruise_pix.pdf. I am using Ferret 6.6.7 on Fedora 13. If you need more information, please let me know. If you have ever seen problems like this: any hint is welcome. Cheers, and thanks, Marco van Hulten -- Koninklijk Nederlands Meteorologisch Instituut Tel.: +31 30 2206 707 Please, do not top-post, quote what is useful and use free file formats. Se vi parolas Esperanton, bonvolu uzi ĝin en via respondo.
Attachment:
section_cruise_pix.pdf
Description: Adobe PDF document
Attachment:
section_grid_pix.pdf
Description: Adobe PDF document
Attachment:
plot_1km_pix.pdf
Description: Adobe PDF document
-48.59328,-48.7979 -47.166,-46.89542 -45.5211,-44.71436 -44.03842,-42.3572 -42.4017,-39.96188 -41.13252,-37.8426 -39.42862,-35.00938 -37.4544,-32.0936 -35.81406,-29.05176 -34.28968,-26.09046 -32.75606,-22.4732 -30.60778,-17.01852 -29.23588,-12.89818 -28.00404,-9.16548 -28.46154,-5.68232 -28.9156,-2.64296 -32.87892,-0.18682 -38.96672,0.7155 -39.6859,1.14654 -41.69956,2.54432 -43.75268,3.97464 -46.41688,5.97866 -48.88358,7.76656 -50.46946,9.54612 -52.0453,11.37302 -53.42096,13.16174 -54.80328,14.87934 -56.2684,16.83184 -57.61216,18.57286 -59.5308,20.45442 -61.84356,21.77658 -63.5832,22.3406 -65.55334,23.27478 -67.07272,24.7148 -67.80424,26.23872 -67.50084,28.09182 -66.52946,29.61548 -64.16674,31.66672 -64.49406,32.22892 -61.09858,32.55144 -58.04974,33.43344 -55.42994,34.32626 -53.29238,36.2078 -50.7119,37.55396 -48.08546,39.49661 -42.52596,44.8441 -39.65919,46.3122 -39.3993,47.80054 -42.4477,49.72166 -45.73246,51.82022 -45.83447,54.06338 -39.70618,58.60299 -37.90794,60.42817 -35.99736,62.3447 -34.25107,64.00026