Leaf Area

This program runs in the same way as DRATES in the previous exercise. PARAM should always be run after the program DRATES and after the appropriate development rates have been entered in the crop data file that is specified under FILEI1 in the control file PARAM.IN (in our case, the file JD305.DAT). Remember that PARAM uses the same control file PARAM.IN as the DRATES program. As DRATES, PARAM uses the measured data provided in the experimental data file (in our example MANAGE.TST as specified under FILEIT in PARAM.IN) to compute crop parameter values.

Ex-VI.6. Click away the Shell (don’t close it, just minimize it) and return to your file manager under Windows. Find the directory C:\COURSE\PARAM\ and run the model PARAM.EXE. Ignore the error message ‘ERROR in OPSYS: no data for output file’ and just give a return. Return to the FSEWin Shell, and include the file PARAM.OUT following the same instructions as given in Exercise VI.2. Open PARAM.OUT and study its content. Note that first some feedback is given on the measured data used in the computations, then calibrated parameter values are given, and then a list is printed of daynumbers and calculated development stages on those dates.

A number of crop parameters that has to be calibrated are specified as functions of development stage. Therefore, PARAM first calculates the development stages for the dates of observations of LAI and crop weights as supplied in the experimental data file. The first tables ‘*Observed biomass values’ and ‘*Observed LAI values’ can be compared with the data supplied in the experimental data file to check whether any errors have been made.

Ex-VI.7. In PARAM.OUT, compare the data given under ‘*Observed biomass values’ and ‘*Observed LAI values’ with the data given in the experimental data file MANAGE.TST and in Box VI.1. Have any mistakes been made in encoding the information from Box VI.1 in the experimental data file?

Next, the development rate parameters are given that were used in the calculation of the development stages on the dates of crop measurements. The values of these development rate parameters should be the same as the ones calculated previously with the program DRATES. If this is not the case, then the values of the development rates as computed with DRATES should be entered in the crop data file that is used in the calibration (in our case, JD305.DAT).

Ex-VI.8. Check that the development rate parameters printed in PARAM.OUT are indeed the same as the ones computed with the program DRATES as printed in the file DRATE.OUT and as set in the crop data file JD305.DAT (see exercise Ex-VI.4).

The first parameter that is calibrated is the relative leaf growth rate, RGRL. This is the parameter RGRLMX (the maximum RGRL under unstressed conditions) in the crop data file. This parameter is only effective in the so-called exponential phase of growth, i.e. at LAI values below 1 (Chapter 3.2.9 of the book ‘ORYZA2000: modeling lowland rice’). The best way to compute RGRL is to plot natural logarithmic values of measured LAI (LNLAI) versus the temperature sum and to calculate RGRL as the slope of a fitted straight line through these data points. An alternative way is to calculate RGRL-values between two consecutive observations, and take the average. An accurate calibration of RGRL requires accurate measurements of LAI with a high frequency between emergence and LAI=1. We see, however, that in our example, there are not enough measurements to calculate accurately RGRL. There are only two computed values for RGRL: 0.006 and –0.005 (check this!). Moreover, the few data we have cover transplanting in the main field and cannot be used to calculate RGRL (these produce the –0.005 value). These data are typical for most field experiments! With the data we have at hand, we have no other option than to leave the value for RGRL uncalibrated. If, after further parameterization, we remain with inaccuracies in the simulation of LAI in the exponential growth stage, we can try to find a satisfactory value for RGRL by trial and error: adjusting the value of RGRLMX in the crop data file until a good fit between simulated and measured LAI is produced. Such a procedure is known as ‘curve fitting’, and should be performed with extreme caution. Values of RGRLMX can only be accepted that are within a biologically plausible range, so not too far off from the maximum (0.0085) and minimum (0.0040) values of RGRLMX and RGRLMN, respectively.

Following the exponential phase of growth, the crop enters a linear phase of growth in which there is a fixed ratio between weight and area of leaves, called specific leaf area (SLA; ha leaf kg-1 leaf). In most crops, SLA changes in the course of crop development, and in ORYZA2000, it is defined as a function of development stage.

Ex-VI.9. Look for the computed values of SLA in the file PARAM.OUT. Note that, except for the first two values, they are all 0.0026. The first two values are in the exponential phase of crop growth and we can discard them. Open the file JD305.DAT and scroll to the section that defines SLA. Since our computed SLA values are constant in time, we don’t need to fit a smooth function. Change the setting of SWISLA from ‘FUNCTION’ to ‘TABLE’, and enter the following simple table:

SLATB = 0.00,  0.0026,

2.50,  0.0026

Run ORYZA2000, and make graphs of simulated and measured LAI, WAGT, WSO, WLVG and any other variable of your choice versus time. Comment on the difference between simulated and measured values, and on the differences between Run 0 and Run 1.

Compared with the previous calibration step (Ex-VI.2-5, Figure VI.2), there is only a slight improvement in the simulation of LAI (Figure VI.3a) and WLVG (Figure VI.3b). The simulation of WAGT improved between days 180 and 220 (Figure VI.3c), whereas there is hardly any change in the simulation of WSO (Figure VI.3d). WSO is underestimated at the end of the season.

Figure VI.3a. Simulated and measured leaf area index (LAI; -); cv. JD305, Beijing, 1987; after calibration of development rates and specific leaf area.

Figure VI.3b. Simulated and measured dry weights (kg ha-1) of green leaves (WLVG); cv. JD305, Beijing, 1987; after calibration of development rates and specific leaf area.

Figure VI.3c. Simulated and measured dry weights (kg ha-1) of total above-ground biomass (WAGT); cv. JD305, Beijing, 1987; after calibration of development rates and specific leaf area.

Figure VI.3d. Simulated and measured dry weights (kg ha-1) of storage organs (WSO); cv. JD305, Beijing, 1987; after calibration of development rates and specific leaf area.