Troubleshooting
Exercise objective:
· Correcting errors in measured crop variables in the experimental data file.
· Editing data files outside the FSEWin Shell.
Suggested reading:
See also Exercise: Error handling in The basics section.
Exercise:
This set of exercises deals with some commonly-made errors in the experimental data file in relation to the use of PARAM. For ORYZA2000, it does not matter how many measured crop variables are listed in the experimental data file, nor on what dates they have been measured. ORYZA2000 does not use the measured data but just lists them in the output files for easy plotting and comparison with simulated values. However, for PARAM to successfully compute the crop parameters discussed in the previous exercises, all measurements of crop variables in the experimental data file should be recorded on the same day. If, in reality different variables were measured on 2-3 consecutive days, then the measured values should still be entered as having been recorded on a single day. Also, the type of measured variables required is strictly specified.
In the following exercises, we will not use the FSEWin Shell to edit input and output files, but will use the text editor ‘Notepad’ that is generally available under Windows. If Notepad is not available on your computer, use any other text editor of your preference. Remember that all input data files should be in ASCII format without any formatting and without the use of <tabs>. For example, if you use Word as text editor, make sure that you save your edited file as unformatted text only.
Ex-VI.17. Close the FSEWinRunOnly Shell (don’t click away this time, but really close the Shell) and return to your file manager under Windows. Open the file MANAGE.TST with Notepad and locate the measured crop variables. Put an asterisk ‘*’ in front of all the LAI measurements:
*Leaf Area Index (m2 leaf / m2 ground):
*LAI_OBS =
*1987.0, 119.0, 0.11,
*1987.0, 145.0, 0.75,
*1987.0, 165.0, 0.25,
*1987.0, 185.0, 1.75,
*1987.0, 200.0, 4.46,
*1987.0, 214.0, 7.51,
*1987.0, 230.0, 8.72,
*1987.0, 251.0, 8.57,
*1987.0, 262.0, 6.29,
*1987.0, 280.0, 3.01,
*1987.0, 292.0, 1.33,
*1987.0, 293.0, 1.07
Close the file MANAGE.TST and run the program PARAM.EXE. Open the file PARAM.OUT and study its contents. Note the following messages:
** WARNING: LAI and WLVG observations do not match in time
** => SLA can not be calculated
and
** WARNING: no LAI data: can not calculate RGRL
To calibrate parameters that govern leaf area growth (the relative leaf growth rate RGRL and the specific leaf area SLA; see ‘Exercise: the PARAM program; leaf area’), we need measured values for leaf area index (LAI). If no LAI values are supplied, the program PARAM gives a warning that parameters related to leaf growth cannot be computed and continues with the computation of other crop parameters. For these calculations, measured data on dry weights of green leaves, yellow and dead leaves combined, stems, and storage organs need to be supplied. If any of these is absent, PARAM is terminated and gives an error message.
Ex-VI.18. Open the file MANAGE.TST with Notepad and locate the measured crop variables. Put an asterisk ‘*’ in front of all the WLVG measurements, as done for LAI in exercise Ex-VI.17 above. Close the file MANAGE.TST and run the program PARAM.EXE. Note the error message produced: “Error in PARAM: Observations on WLVG and WLVD do not match in time”.
If the dry weights of green leaves, dead leaves, stems or storage organs has not been measured and entered in the experimental data file, the program PARAM can not perform the computations of the partitioning tables, loss rate of green leaves and stem reserves. It will send an error message to the screen and terminate execution. The error message is generic and just informs us that WLVG and WLVD measurements are not the same in time. In this case, there are no WLVG measurements at all!
Ex-VI.19. Open the file MANAGE.TST with Notepad and locate the measured crop variables. Remove the asterisk ‘*’ in front of all the LAI and WLVG measurements. Close the file, run the program PARAM.EXE and note that all crop parameters for calibration are computed. Next, open MANAGE.TST again, and change one of the dates in the measured LAI list. For example, locate the third line in the measured LAI list:
1987.0, 165.0, 0.25,
and change the date of measurement from 165.0 in 166.0. Close MANAGE.TST and run the program PARAM.EXE. Open the file PARAM.OUT and study its contents. Note the following warning message:
** WARNING: LAI and WLVG observations do not match in time
** => SLA can not be calculated
The specific leaf area (SLA) is calculated using the measured leaf area index (LAI) and the measured weight of green leaves (WLVG). This is done at each date of measurement. Therefore, if the dates of measured LAI and WLVG are not the same, the SLA calculation would be wrong. PARAM checks whether the dates of LAI and WLVG recordings are the same at each measurement, and does not calculate SLA when this is not the case. PARAM can still calculate other crop parameters for calibration, and therefore PARAM continues its execution and there is no error message sent to the screen.
Ex-VI.20. Open the file MANAGE.TST and restore the date in the third line in the measured LAI list from 166.0 in 165.0. Now locate the fourth line in the measured WLVG list:
1987.0, 185.0, 661.2,
and change the date of measurement from 185.0 into 184.0. Close MANAGE.TST and run the program PARAM.EXE. Note the error message produced: “Error in PARAM: Observations on WLVG and WLVD do not match in time”. Also note that this error is produced just after the line "Default set, Year: 1987, Day: 184.00".
If the dates of measurement of the dry weights of green leaves, dead leaves, stems or storage organs do not match, the program PARAM can not perform the computations of the partitioning tables, loss rate of green leaves and stem reserves. It will send an error message to the screen and terminate execution. In this case, the error message points to a mismatch in dates of measured WLVG or WLVD. Since the error message follows the day of simulation 184, we know we have to look for errors around day 184 in the measured values of WLVG or WLVD.