Experiment simulation

Exercise objective:

·   Simulating an irrigated/rainfed experiment with observations

The exercises so far in Chapter V were all carried out in the exploration mode, i.e. without any experimental observations. Here, we will practice with data from a real field experiment carried out in 1996 in Jakenan, East Indonesia. The treatments selected are a fully irrigated crop and a purely rainfed crop. The rice variety used was IR64, transplanted in the second part of the rainy season (Walik Jerami season). Both treatments received large quantities of macro nutrients (N, P and K), though probably not enough to realize potential production. The measured data set is insufficient for an accurate calibration: no leaf area was measured, and the measured soil water contents were considered unreliable. This situation, however, is typical for many experiments for which we want to use a crop growth model to help in data interpretation. The experimental data presented are part of a large data set, spanning 5 years, for which ORYZA2000 is still under calibration. For this exercise, we simplified some of the input data and use some parameter values for cv. IR64 that particularly suit the 1996 data.

Exercise: 

Ex-V.9. Start up the Shell FSEWinRunOnly. Open project C:\COURSE\JAKENAN96\ORYZAWin. Check the control file:

FILEON = 'RES.DAT'                              ! Output file

FILEOL = 'MODEL.LOG'                            ! Log file

FILEIR = 'c:\course\Jakenan96\RERUNS.DAT'       ! Rerun file

FILEIT = 'c:\course\Jakenan96\IRRIGATE.E96’     ! Experimental data

FILEI1 = 'c:\course\Jakenan96\IR64.J96'         ! Crop data

FILEI2 = 'c:\course\Jakenan96\SOILIR.S96’       ! Soil data

Check the rerun file:

FILEIT = 'c:\course\Jakenan96\RAINFED.E96'

FILEI2 = 'c:\course\Jakenan96\SOILRA.S96'

The default run simulates the irrigated treatment and the rerun the rainfed treatment. Open and examine the experimental data files and the soil data files. Note that in IRRIGATE.E96, no water balance is simulated. Also note that in the soil data file SOILRA.S96 measurements of soil water tension are included. No measurements of soil water content are given. In the experimental data file, note the absence of measurements of leaf area index. Run ORYZA2000 and examine the output. Make graphs of different crop and soil variables versus time, such as weight of above-ground dry matter, weight of panicles, soil water tension in layer 2 or 4, etc. Compare simulations with observations, and comment on the goodness-of-fit.

In this experiment, we see a good fit between simulated and observed biomass in the irrigated treatment (Figure V.3a). In the rainfed treatment, however, the simulated crop growth is better than the measured crop growth (Figure V.3b), although the simulated soil water tension values are reasonably good (Figures V.3c, V.3d). In this experiment, nitrogen fertilizer application was not very high. Under the alternate wet and dry conditions of the rainfed treatment, relatively large losses of nitrogen may have occurred because of alternating nitrification-denitrification processes, and nitrogen may have limited actual crop growth. Quite often, we can use simulation results of ORYZA2000 to set benchmark growth patterns and yield levels that might have been realized if no other yield-reducing factors had played any role. This helps in the interpretation of the experimental data. For this particular experiment, the analysis of nitrogen dynamics is still in progress.

Figure V.3a. Simulated and observed weight of above-ground biomass (WAGT; kg ha-1) and of panicles (WSO; kg ha-1) versus time; cv. IR64 grown under irrigated conditions; Jakenan, Indonesia, wet season 1996.

Figure V.3b. Simulated and observed weight of above-ground biomass (WAGT; kg ha-1) and of panicles (WSO; kg ha-1) versus time; cv. IR64 grown under rainfed conditions; Jakenan, Indonesia, wet season 1996.

Figure V.3c. Simulated and observed soil water tension at 10 cm depth (MSKAP2; kPa) versus time; cv. IR64 grown under rainfed conditions; Jakenan, Indonesia, wet season 1996.

Figure V.3d. Simulated and observed soil water tension at 20 cm depth (MSKAP4; kPa) versus time; cv. IR64 grown under rainfed conditions; Jakenan, Indonesia, wet season 1996.