Parameterization

Exercise objective:

·   Parameterization of ORYZA2000 under conditions of potential production.

·   Using the parameterization programs DRATES and PARAM.

Suggested reading:

Chapters 7.3 and 8.2 of the book, ORYZA2000: modelling lowland rice.

This set of exercises illustrates how to parameterize (or calibrate) ORYZA2000 for a particular rice variety grown in a particular environment. We limit ourselves here to calibrating the most important crop characteristics that determine the response of the crop to environmental characteristics under conditions of potential production.

Exercise:

Most parameters in the crop data file are valid for all rice varieties, and only a few need to be empirically determined. This should be done using data from well-designed field experiments without any water or nutrient limitations, and without any pest, disease or weed infestation. Of course, ORYZA2000 can be calibrated using data from non-potential production conditions as well, but then it should be realized that the parameters derived implicitly include the effect of growth limitations. Thus, such parameter values cannot be used to estimate potential production, and one should be careful in extrapolating results beyond the experimental conditions. The following measurements should be made at frequent intervals (ideally 2 weeks or less) during the growing season: green leaf area index, dry biomass of stems (including leaf sheaths), green leaves, yellow plus brown leaves, and panicles. The panicle should be cut from its supporting stem at the first node (that supports or supported a stemlet with spikes) from the bottom, the remainder belongs to the stem. The separation of green and yellow leaves is a judgmental call. When leaves are partially yellow, the yellow parts should be cut with scissors and added to the yellow leaves, the remaining green parts are to be added to the whole green leaves of the plants. Dead leaves that have fallen off should be collected and added to the other dead and yellow leaves. If dead leaves have not systematically been collected and measured, then their weight can be estimated from the weight of the green leaves as follows: assuming that green leaves reach their maximum weight at flowering, the difference in weight at each sampling after flowering indicates the loss as dead leaves, and can hence be interpreted as the weight of dead leaves. It is important to take at least crop samples at transplanting, maximum tillering, panicle initiation (PI), flowering, half-way flowering and maturity, and physiological maturity. All measurements should be made on the same day (or, if there are 1-2 day differences, then the values should be entered in the experimental data file as having been taken on the same day).

In these exercises, we parameterize ORYZA2000 for a Chinese rice variety called Jin Dao 305 (JD305), grown near Beijing in the summer of 1987. The data that we use were not obtained from a real experiment, but were combined especially for these exercises. Hence, they are fictive data (and although they were derived from a real experiment in 2001, they have no real value). Box VI.1 gives the details of the fictive experiment. Twelve crop cuts were mimicked during the growing season, of which the first was at emergence in the seedbed and the second at the day of transplanting. Box VI.2 gives the measured leaf area indices and dry weights of crop organs.

Box VI.1. Experimental details of a fictive potential production field experiment for model parameterization.

Box VI.2. Fictive measured values of leaf area index (LAI; m2 m-2), and dry weights (kg ha-1) of total above-ground biomass (WAGT), stems (WST), green leaves (WLVG), yellow and dead leaves (WLVD) and panicle/storage organs (WSO).