Water management

Exercise objective:

      ·  Irrigation water management; the experimental data file.

Suggested reading:

Chapter 7.2 (the part on water management) of the book, ORYZA2000: modelling lowland rice.

The management of (irrigation) water is specified in the experimental data file.

Exercises:

Ex-IV.3. In the experimental data file, study the parameters in Section 6 Irrigation parameters. Note that the definition of all the parameters that specify the application of irrigation depend on the setting of parameter SWITIR. Run ORYZA2000, and fill out the explanations and units of the variables in Table IV.1 (use the variable listing on pages 215-230 of the book ORYZA2000: modelling lowland rice), and fill out column Sim2. In the chart option, note the presence of new variables. Make the following graphs: (a) amount and timing of irrigation (IR; mm) and rainfall (RAIN; mm) versus time; (b) depth of ponded water (WL0; mm) versus time; (c) water content in soil layer 1 (WCL1; m3 m-3), soil layer 4 (WCL4) and soil layer 7 (WCL7) versus time; and (d) actual transpiration rate (TRW; mm d-1) versus potential transpiration rate (TRC; mm d-1) (see Figs. IV.1a-d). Explain the results from the information in the experimental data file.

In all settings of SWITIR, except SWITIR = 1, ORYZA2000 determines the time of application of irrigation in dependence of user-defined irrigation criteria. In the example experimental data file, the parameter SWITIR is set to 2. This implies that irrigation is applied when the simulated depth of ponded water (in PADDY) reaches a minimum level. This minimum level is defined by the parameter WL0MIN, which is set to 10 mm. The amount of irrigation applied each time is defined by the parameter IRRI, which is set to 75 mm. In Figure IV.1a, we see that rainfall is very low and erratic, and that a large number of irrigations have to be given to keep ponded water on the soil surface. Figure IV.1b shows the dynamics of simulated depth of ponded water. The soil never falls dry and no drought occurs. In Figure IV.1c, we see that after transplanting (day 16), the various soil layers are continuously saturated with water. Only during the seedbed period, the water content drops after emergence, but this is reset to saturation when the crop is transplanted in the main field. It should be remembered that the water balance and the drought stress factors are only taken into account for the period that the crop grows in the main field. It is implicitly assumed that the seedbed is optimally supplied with water. The simulated irrigation application, IRC, only accounts for water applied to the main field.

In Table IV.1, the variable RAINC quantifies total rainfall received in the main field after transplanting (or after emergence in the case of direct-seeding!), whereas the variable RAINCU quantifies total rainfall received from the start of simulation (STTIME). In Table IV.1, the simulated yield with the water balance (Sim2) is similar to that simulated in the potential production mode. This is because enough irrigation is applied to keep the soil continuously submerged. The straight line in Figure IV.1d confirms that the actual transpiration is the same as potential transpiration: no drought stress occurs with the current irrigation parameter settings. Thus, for a given soil and a given season, we can use the water mode of ORYZA2000 to compute the amount and timing of irrigation required to realize potential (no water limitations) production. In this example, 750 mm of cumulative irrigation water is required to realize potential yield (column Sim2, Table IV.1).

Figure IV.1a. Amount and timing of irrigation (IR; mm) and rainfall (RAIN; mm) versus time; cv. IR72 grown with sufficient irrigation, IRRI, Los Baños, dry season 1992.

Figure IV.1b. Depth of ponded water (WL0; mm) versus time; cv. IR72 grown with sufficient irrigation, IRRI, Los Baños, dry season 1992.

Figure IV.1c. Water content in soil layer 1 (WCL1; m3 m-3), soil layer 4 (WCL4) and soil layer 7 (WCL7) versus time; cv. IR72 grown with sufficient irrigation, IRRI, Los Baños, dry season 1992.

Figure IV.1d. Actual transpiration rate (TRW; mm d-1) versus potential transpiration rate (TRC; mm d-1); cv. IR72 grown with sufficient irrigation, IRRI, Los Baños, dry season 1992.

Ex-IV.4. Make a rerun file varying the start time of simulation (remember that in the exploration mode, emergence day EMD is automatically set equal to STTIME!):

STTIME = 60.

STTIME = 120.

STTIME = 180.

Run ORYZA2000 and fill out columns Sim3, Sim4 and Sim5 of Table IV.1 for STTIME is 60, 120 and 180, respectively. Explain the results. A. With emergence later in the year, we enter the wet season and periods with higher temperatures. This means lower yields, because there is less radiation (more clouds) and shorter growth durations (check DAE!), more rainfall and less irrigation required.

Ex-IV.5. In the experimental data file, change the parameter SWITIR to 0 to simulate purely rainfed production (no irrigation will be given). Run ORYZA2000 with the rerun file as in Ex-IV.4. Fill out Table IV.2: Sim1 is the default run (SSTIME=4.), Sim2 is STTIME=60., etc. Explain the results.

Q. Why does yield now increase with later emergence, whereas in Table IV.1, it decreased with emergence date? Why is the yield in Sim2 lower than in Sim1, while total rainfall is a bit higher? Make the following graphs to help explain the results in Table IV.2: For all 4 runs: (a) above-ground weight (WAGT) versus time; and (b) cumulative rainfall (RAINCU) versus time. For runs 0 and 1: (c) above-ground weight (WAGT) and weight of panicles (WSO) versus development stage; (d) above-ground weight (WAGT) and weight of stems (WST) versus development stage; (e) number of spikes (grains) (NSP) versus development stage; (f) leaf rolling/spikelet sterility factor (LRSTRS) versus development stage; and (g) soil water tension in the second soil layer (MSPKA2) versus development stage.

Table IV.2.

View ANSWERS from the Tutorial_answer_sheet.pdf file.

In Table IV.2, yield under purely rainfed conditions generally increases with later emergence date because, with later emergence, we shift from the dry season into the rainy season. Yield increases as accumulated rainfall increases with later emergence (Figures IV.2a and IV.2b). Actual transpiration and evaporation also increase with later emergence date, because more (rain) water is available for these processes. To understand why yield in Sim2 (STTIME=60) is so much lower than in Sim1 (STTIME=4), although accumulated rainfall is twice as high, we need to study the figures carefully. By plotting crop and soil variables versus development stage instead of time, we can easier compare what happens to the crop in similar growth stages. In Figure IV.2c, we see that the weight of panicles in Sim2 stagnates shortly after flowering whereas that of Sim1 continues to increase. This is in contrast to the weight of the whole crop, which shows that Sim2 increases more than Sim1. The pattern in Figure IV.2c indicates that we deal with a case of sink limitation in Sim2. Though biomass continues to accumulate, dry weight does not accumulate in the panicles, but in the stems (Figure IV.2d). This is the consequence of drought around flowering that induced spikelet sterility. In Figure IV.2e, we see that the final number of spikes (grains) in Sim2 is smaller than in Sim1. In Figure IV.2f, we see indeed that the spikelet reduction factor (LRSTRS) around flowering is smaller for Sim2 than for Sim1. Finally, Figure IV.2g shows that the soil water tension around flowering is higher (indicates more severe drought) in Sim2 than in Sim1. The example of Sim1 and Sim2 illustrates the complex and dynamic interactions between soil water availability (soil water tension) and various crop growth processes. Often, we cannot explain differences in simulated yield from looking at the end results only (the data in OP.DAT), but we have to analyze dynamic patterns of different variables over time.

Figure IV.2a. Above-ground weight (WAGT; kg ha-1) versus time; cv. IR72 grown under rainfed conditions at four emergence dates, IRRI, Los Baños, dry season 1992.

Figure IV.2b. Cumulative rainfall (RAINCU; mm) versus time; cv. IR72 grown under rainfed conditions at four emergence dates, IRRI, Los Baños, dry season 1992.

Figure IV.2c. Above-ground weight (WAGT; kg ha-1) and weight of panicles (WSO; kg ha-1) versus development stage; cv. IR72 grown under rainfed conditions at emergence date 4 (Run 0) and at emergence date 60 (Run 1), IRRI, Los Baños, dry season 1992.

Figure IV.2d. Above-ground weight (WAGT; kg ha-1) and weight of stems (WST; kg ha-1) versus development stage; cv. IR72 grown under rainfed conditions at emergence date 4 (Run 0) and at emergence date 60 (Run 1), IRRI, Los Baños, dry season 1992.

Figure IV.2e. Number of spikes (grains) (NSP; no ha-1) versus development stage; cv. IR72 grown under rainfed conditions at emergence date 4 (Run 0) and at emergence date 60 (Run 1), IRRI, Los Baños, dry season 1992.

Figure IV.2f. Spikelet reduction factor/leaf rolling factor (LRSTRS; -) versus development stage; cv. IR72 grown under rainfed conditions at emergence date 4 (Run 0) and at emergence date 60 (Run 1), IRRI, Los Baños, dry season 1992.

Figure IV.2g. Soil water tension in soil layer 2 (i.e. at depth 15 cm) (MSKPA2; kPa) versus development stage; cv. IR72 grown under rainfed conditions at emergence date 4 (Run 0) and at emergence date 60 (Run 1), IRRI, Los Baños, dry season 1992.

Exercises IV.4 and IV.5 illustrated two extreme conditions of irrigation management: full irrigation so that potential production is realized, and zero irrigation so that purely rainfed conditions are simulated. The irrigation parameters in the experimental data file allow simulation of various intermediate irrigation scenarios, namely the selections SWITIR = 3 to 5. A specific type of water-saving irrigation is the so-called alternate wetting and drying (AWD). In AWD, irrigation is supplied after a number of days have passed since disappearance of ponded water. This can be simulated by setting SWITIR = 5 and specifying with the parameter WL0DAY how many days after disappearance of ponded water irrigation should be given.

Ex-IV.6. In the experimental data file, set SWITIR to 5 and WL0DAY to 0. Make a rerun file varying WL0DAY:

WL0DAY = 5

WL0DAY = 10

WL0DAY = 15

Run ORYZA2000 and fill out Table IV.3. Explain the differences. Make graphs of some variables versus time, e.g., depth of ponded water (WL0), weight of above-ground biomass (WAGT) or panicles (WSO), soil water content (WCL) or soil water tension (MSKPA) at certain soil depths, etc., etc.

The yield in Sim1 in Table IV.3 is close to potential (slightly lower than Sim1 in Table IV.1!), since we allow at maximum one day without standing water before re-irrigating (Figure IV.3a). The criterion WL0DAY = 0 means effectively that we can have 1 day without ponded water because of the 1-day time step of ORYZA2000. With AWD, we can reduce irrigation water input substantially, while maintaining high yields. The trade-off between water savings and yield reduction depends on soil type (see Chapter V: Water-limited production; soil properties). In this example, the soil has a high clay content. Water holding capacity is rather good and percolation rate low, so that a substantial proportion of the irrigation water is actually retained in the profile and is available for uptake by the roots. With increasing length of the periods with non-flooded soil (WL0DAY) between irrigations, the water productivity (calculated as weight of grains over total [irrigation plus rainfall from transplanting to harvest] water use) slightly increases: 1.30 (Sim1), 1.34 (Sim2), 1.40 (Sim3) and 1.47 (Sim4) kg grain kg-1 water. In Table IV.3, total accumulated rainfall is the same in all reruns, since they all start and mature at the same time. The amount of water transpired by the crop (TRWC) is significantly reduced with decreasing amount of irrigation. TRWC is 6, 15 and 22% smaller in Sim2, Sim3 and Sim4, respectively, than in Sim1. This is the reason why yields decline with increasing number of dry days between irrigations. The reductions in irrigation water use are mainly realized through reductions in percolation losses. A simple water balance states that the sum of water inflows equals the sum of water outflows (if we neglect changes in soil water storage). Inflows, in this situation, are rain and irrigation, and outflows are transpiration, evaporation and deep percolation (surface runoff is zero in all runs; check in OP.DAT!). Using the data in Table IV.3, we can calculate the sum of percolation losses for Sim1, Sim2, Sim3 and Sim4 to be 264, 159, 61 and 20mm, respectively. Thus, AWD with increasingly larger durations between irrigations increasingly reduces irrigation inputs through increasingly reduced percolation losses.

Table IV.3.

View ANSWERS from the Tutorial_answer_sheet.pdf file.

Figure IV.3a. Amount and timing of irrigation (IR; mm) versus time; cv. IR72 irrigated zero days after disappearance of ponded water, IRRI, Los Baños, dry season 1992. Compare to Figure IV.1a.

Figure IV.3b. Amount and timing of irrigation (IR; mm) versus time; cv. IR72 irrigated 15 days after disappearance of ponded water, IRRI, Los Baños, dry season 1992. Compare to Figure IV.1a.

Figure IV.3c. Above-ground weight (WAGT; kg ha-1) versus time; cv. IR72 grown under four alternate wetting and drying irrigation scenarios, IRRI, Los Baños, dry season 1992.

Figure IV.3d. Soil water tension in soil layer 2 (i.e. at depth 15 cm) (MSKPA2; kPa) versus time; cv. IR72 grown under four alternate wetting and drying irrigation scenarios, IRRI, Los Baños, dry season 1992.

The parameter settings SWITIR=3 and SWITIR=4 provide the same functionality as SWITIR=5 in Ex-IV.6. However, instead of specifying the number of non-flooded days between irrigations, now we have to specify the irrigation criterion as threshold soil water tension KPAMIN (kPa) when SWITIR=3 or threshold soil water content (m3 m-3) WCMIN when SWITIR=4. In addition, we have to specify at which depth, i.e. at which soil layer, this threshold values operates.

Ex-IV.7. Practice with the switch settings SWITIR=3 and SWITIR=4 in the same way as in Ex-IV.6. Change the setting of SWITIR in the experimental file, and make reruns varying the values of KPAMIN and WCMIN. Note the sensitivity of yield and (irrigation) water input to the values of KPAMIN and WCMIN, and make some graphs to examine the dynamics in crop growth and soil water balance.

The last option to control irrigation water management is by SWITIR=1. In that situation, we have to supply an array with actual irrigation applications as function of time: RIRRIT. This option works similar to the definition of fertilizer application FERTIL when we run ORYZA2000 with a nitrogen balance (see Chapter III: Nitrogen-limited production). The table RIRRIT gives the amount of irrigation as function of day of the year. RIRRIT is linearly interpolated in ORYZA2000 at each day of simulation and, therefore, each day of irrigation application should be preceded and succeeded by a day with zero irrigation application.

Ex-IV.8. In the experimental data file, select SWITIR = 1, and activate the table RIRRIT by removing the asterisk (*). Recognize that this default run is a purely rainfed situation. Create the following rerun file with supplementary irrigation every 20 days and every 10 days:

* rerun set 1

RIRRIT =   0., 0.,

25., 0.,

26., 75.,

27., 0.,

45., 0.,

46., 75.,

47., 0.,

65., 0.,

66., 75.,

67., 0.,

85., 0.,

86., 75.,

87., 0.,

366., 0.

* rerun set 2

RIRRIT =   0., 0.,

25., 0.,

26., 75.,

27., 0.,

35., 0.,

36., 75.,

37., 0.,

45., 0.,

46., 75.,

47., 0.,

55., 0.,

56., 75.,

57., 0.,

65., 0.,

66., 75.,

67., 0.,

75., 0.,

76., 75.,

77., 0.,

85., 0.,

86., 75.,

87., 0.

95., 0.,

96., 75.,

97., 0.,

366., 0.

Run ORYZA2000, fill out Table IV.4, and make some graphs to examine the dynamics of crop growth and soil water balance. Notice the large effect of a relatively small irrigation application on yield!

Table IV.4.

View ANSWERS from the Tutorial_answer_sheet.pdf file.

Figure IV.4a. Leaf area index (LAI; -) versus time; cv. IR72 grown under rainfed conditions and two supplementary irrigation scenarios, IRRI, Los Baños, dry season 1992.

Figure IV.4b. Cumulative irrigation water (IRC; mm) versus time; cv. IR72 grown under rainfed conditions and two supplementary irrigation scenarios, IRRI, Los Baños, dry season 1992.

Figure IV.4c. Actual transpiration rate (TRW; mm d-1) versus time; cv. IR72 grown under rainfed conditions and two supplementary irrigation scenarios, IRRI, Los Baños, dry season 1992.