Incorrect values for the probed grid cells after using "maximum_8hour_mean" stats in a tile plot

[yadongxuEPA]

Describe the bug This issue was first reported on CMAS forum by a user under this link https://forum.cmascenter.org/t/bizarre-calculation-of-md8a-in-verdi/4186

To Reproduce Steps to reproduce the behavior:

1. Launch VERDI GUI
2. Load testing CAMx dataset: "CAMx.SIP_WF4_1_33kn2023OSATnoctrlv1.20170713.avrg.grd02.nc" from /work/MOD3APP/verdi/VERDI_2.1.4_testing_files
3. Select "O3(ppmv)" from variables list of the above model data and click "Tile Plot"
4. Click on Controls >> Set Row and Column Ranges to set the columns as 40-60 and rows as 60-80 and turn on the grid lines
5. Select "maximum_8hour_mean" from "Stats" drop-down menu
6. Click on Controls >> Probe to read the MD8A values from the color map

Expected behavior VERDI can display the calculated maximum daily 8 hour averages (MD8A) correctly for CAMx output file.

Screenshots If applicable, add screenshots to help explain your problem. The screenshot below showed that there is a mismatch between the color in the grid cell (50,64) and the probed value at the lower left corner. the color map showed that the grid cell's value is between 0.048-0.057, but the probed value is 0.006

[yadongxuEPA]

I also tested this issue with the example CAMx file "camx.example.grd01" came with the VERDI builds, found the same problem.
Furthermore, I found that this issue has been long-existing (including VERDI_2.0_20210311 builds).
Please note that the color map did display the right color after selecting "maximum_8hour_mean" Stats. However, the "probe" function could not display the correct value when clicking on a specific grid cell.

A temporal solution for this user is to get the grid cell values from the exported shapefiles (instead of using 'probe' function on the color map) after applying "maximum_8hour_mean" Stats. But we still need to resolve the issue of incorrect "probed" values.

[lizadams]

Testing using VERDI_2.1.5_mac_20240124.tar.gz

./verdi.command -f $cwd/../CAMx.SIP_WF4_1_33kn2023OSATnoctrlv1.20170713.avrg.grd02.nc -s "O3[2]" -subDomain 40 60 60 80 -g tile

When I use the GUI and select

Stats> maximum_8hr_mean

Select Controls > Probe

I get the following error, and no values appear in the lower left corner of the GUI window.

2024.01.25 17:09:55.620 [AWT-EventQueue-0] ERROR anl.verdi.plot.gui.FastTilePlot - Invalid Range Exception in FastTilePlot.Probe ucar.ma2.InvalidRangeException: first (-55) must be >= 0 at ucar.ma2.Range.(Range.java:148) ~[netcdfAll-4.5.5-SNAPSHOT.jar:?] at ucar.ma2.Range.(Range.java:134) ~[netcdfAll-4.5.5-SNAPSHOT.jar:?] at ucar.ma2.Array.sectionNoReduce(Array.java:679) ~[netcdfAll-4.5.5-SNAPSHOT.jar:?] at anl.verdi.data.AbstractDataFrame.sliceCopy(AbstractDataFrame.java:193) ~[?:?] at anl.verdi.plot.gui.FastTilePlot.probe(FastTilePlot.java:3251) ~[?:?] at anl.verdi.plot.gui.FastTilePlot.access$36(FastTilePlot.java:3220) ~[?:?] at anl.verdi.plot.gui.FastTilePlot$AreaFinder.mouseReleased(FastTilePlot.java:3506) ~[?:?] at java.awt.AWTEventMulticaster.mouseReleased(AWTEventMulticaster.java:298) ~[?:?] at java.awt.Component.processMouseEvent(Component.java:6621) ~[?:?] at javax.swing.JComponent.processMouseEvent(JComponent.java:3398) ~[?:?] at anl.verdi.plot.gui.FastTilePlot.processMouseEvent(FastTilePlot.java:1231) ~[?:?] at java.awt.Component.processEvent(Component.java:6386) ~[?:?] at java.awt.Container.processEvent(Container.java:2266) ~[?:?] at java.awt.Component.dispatchEventImpl(Component.java:4996) ~[?:?] at java.awt.Container.dispatchEventImpl(Container.java:2324) ~[?:?] at java.awt.Component.dispatchEvent(Component.java:4828) ~[?:?] at java.awt.LightweightDispatcher.retargetMouseEvent(Container.java:4948) ~[?:?] at java.awt.LightweightDispatcher.processMouseEvent(Container.java:4575) ~[?:?] at java.awt.LightweightDispatcher.dispatchEvent(Container.java:4516) ~[?:?] at java.awt.Container.dispatchEventImpl(Container.java:2310) ~[?:?] at java.awt.Window.dispatchEventImpl(Window.java:2780) ~[?:?] at java.awt.Component.dispatchEvent(Component.java:4828) ~[?:?] at java.awt.EventQueue.dispatchEventImpl(EventQueue.java:775) ~[?:?] at java.awt.EventQueue$4.run(EventQueue.java:720) ~[?:?] at java.awt.EventQueue$4.run(EventQueue.java:714) ~[?:?] at java.security.AccessController.doPrivileged(AccessController.java:400) [?:?] at java.security.ProtectionDomain$JavaSecurityAccessImpl.doIntersectionPrivilege(ProtectionDomain.java:87) ~[?:?] at java.security.ProtectionDomain$JavaSecurityAccessImpl.doIntersectionPrivilege(ProtectionDomain.java:98) ~[?:?] at java.awt.EventQueue$5.run(EventQueue.java:747) ~[?:?] at java.awt.EventQueue$5.run(EventQueue.java:745) ~[?:?] at java.security.AccessController.doPrivileged(AccessController.java:400) [?:?] at java.security.ProtectionDomain$JavaSecurityAccessImpl.doIntersectionPrivilege(ProtectionDomain.java:87) [?:?] at java.awt.EventQueue.dispatchEvent(EventQueue.java:744) [?:?] at java.awt.EventDispatchThread.pumpOneEventForFilters(EventDispatchThread.java:203) [?:?] at java.awt.EventDispatchThread.pumpEventsForFilter(EventDispatchThread.java:124) [?:?] at java.awt.EventDispatchThread.pumpEventsForHierarchy(EventDispatchThread.java:113) [?:?] at java.awt.EventDispatchThread.pumpEvents(EventDispatchThread.java:109) [?:?] at java.awt.EventDispatchThread.pumpEvents(EventDispatchThread.java:101) [?:?] at java.awt.EventDispatchThread.run(EventDispatchThread.java:90) [?:?]

The probe now is looking for a value of (11,5) from the subdomain versus (1,1,51,64) from the larger domain, and VERDI throws the following error:

If I don't use the -subDomain option, but instead zoom into the same area from the GUI, then probing works.

 ./verdi.command -f $cwd/../CAMx.SIP_WF4_1_33kn2023OSATnoctrlv1.20170713.avrg.grd02.nc -s "O3[2]" -g tile

When I use the GUI and select

Stats> maximum_8hr_mean

Select Controls > Probe

Then it works, the correct value is available in the plot.

[yadongxuEPA]

Re-tested VERDI_2.1.5_linux64_20240124.tar.gz on Atmos, the probed grid cell values were still shifted toward +1 north-eastward as previously described. I don't think we planned to resolve this issue in VERDI 2.1.5. We can discuss the details later.

[cgnolte]

In my testing, using VERDI 2.1.5 on atmos interactively, in probe mode, clicking on a grid cell results in the coordinates and value of the grid cell shifted by 1 north and east (diagonally). The data value is consistent with the coordinate displayed -- it just isn't the grid cell the user clicked on.

[yadongxuEPA]

Re-tested VERDI_2.1.5_linux64_20240221.tar.gz on Atmos, found that the issue has resolved. The values were displayed correctly for the probed grid cells (with or without using the stats).

[yadongxuEPA]

Also tried from the command line then use probe the grid cells, it also worked.