Determining the best volume of potato irrigation water in tape drip irrigation system using by WOFOST model

Introduction

Determining the appropriate amount of water for potato cultivation has been investigated by various researchers around the world and based on various factors such as irrigation methods, available technology, local climate and other management factors, various amounts have been presented. Based on these studies, the optimal amount of irrigation water for each region should be determined by field experiments. But doing it requires spending a lot of time and money. On the other hand, Iran's facing drought conditions in the coming decades and the need to optimize water consumption in the agricultural sector increases the importance of determining the optimal amount of irrigation water for potato production. To solve this problem and increase the speed of decision-making, plant models have been developed. The WOFOST model is one of the plant models introduced by Wageningen University in the Netherlands to simulate the growth and yield of agricultural plants. Although the WOFOST model has been widely used by researchers to simulate agricultural plants; But its use for potatoes has been less interesting to researchers. On the other hand, the optimal amount of potato irrigation water has been suggested by researchers around the world in a wide range, and it is necessary to determine and optimize these amounts in each region. Based on this importance, the present research was conducted to determine the optimal limits of potato irrigation water with the aim of achieving optimal yield and high water productivity. WOFOST model was used to simulate different irrigation scenarios.



Materials and Methods

This research was conducted in two crop years in a research farm located in Kermanshah city in the form of complete random blocks. The investigated treatments included the supply of irrigation water at the levels of 100 (T1), 75 (T2) and 50 (T3) percent of the potato plant's water requirement in T-Tape drip irrigation system. Irrigation treatments were the same until the 5-leaf stage, and after that, irrigation treatments were applied. In order to determine the amount of irrigation water in each treatment, volumetric meters were used. The amount of readily available water and soil residue (W2) and the percentage of usable water discharge or Re in the plant root zone were calculated using equations (1) and (2).

(1)



(2)



The WOFOST model is a plant growth simulation model based on the carbon cycle and has a complex structure. This model simulates plant growth in three conditions of no limiting factor, water limitation and food limitation. In fact, in the WOFOST model, crop growth is simulated based on eco-physiological processes. Before recalibration and validation of the WOFOST model, sensitivity analysis was performed based on equation (3):

(3)



In this regard, Sc is the dimensionless sensitivity coefficient, Pm is the estimated value of the desired parameter based on adjusted input data, and Pb is the estimated value of the desired parameter based on the basic input data. After the sensitivity analysis, the WOFOST model was recalibrated using the data of the first year and validated using the data of the second year of cultivation.



Results and Discussion

The statistical results of observed and simulated yield and water productivity by WOFOST model are shown in Table (1). According to these results, in the calibration phase, the WOFOST model had an underestimation error. The accuracy of this model based on the NRMSE statistic was in the excellent category. The error rate of WOFOST model for determining yield and water productivity was equal to 1.6 tons per hectare and 0.17 kg/m3, respectively. Based on two statistics, EF and d, the efficiency of this model to determine crop yield was better than water productivity.



Table 1- Statistical results of observed and simulated yield and water productivity by the WOFOST model in the calibration stage

MBE RMSE NRMSE EF D Unit Parameter

0.5- 1.6 0.06 0.93 0.99 Tons/ha Yield

0.06- 0.17 0.04 0.04- 0.99 kg/m3 Water productivity



The statistical results of yield and water productivity in the validation stage are shown in table (2). Based on the MBE statistic, the WOFOST model had an overestimation error in the performance simulation and an underestimation error in the water productivity simulation. Based on the NRMSE statistic, the accuracy of this model was excellent for determining both parameters. As in the calibration phase, the efficiency of the WOFOST model was better in simulating performance than water productivity.



Table 2- Statistical results of observed and simulated yield and water productivity by the WOFOST model in the validation stage

MBE RMSE NRMSE EF D Unit Parameter

0.1 0.93 0.03 0.97 0.99 Tons/ha Yield

0.04- 0.14 0.03 0.40- 0.99 kg/m3 Water productivity



In general, to determine the optimal scenario, it should be kept in mind that the difference between potato yield in two irrigation depths of 634 and 487 mm was greater than other depths, while the difference between water productivity between these two depths was only 5%. Therefore, the irrigation water depth of 634 mm is chosen as the optimal depth for irrigation. These results are close to the values suggested by Doornbos and Kassam (1979) in FAO publication No.33.



Conclusion

The results showed that the WOFOST model had the necessary accuracy (NRMSE<0.1) and efficiency (d>0.99) to simulate the yield and water productivity of potato in Kermanshah region. Based on these results, different irrigation water management scenarios were investigated on the yield and water productivity of this crop. The difference between potato yield in two irrigation depths of 634 and 487 mm was greater than other irrigation intervals. For this reason, the yield and water productivity values were compared with each other in this interval. Based on all the results, the depth of 634 mm was determined as the optimal value for potato cultivation. Potato yield at this depth is about 25 tons per hectare and water productivity is about 2.4 kg/m3. These values decreased and increased by 7.3 tons per hectare and 0.8 kg/m3 compared to providing 100% of the plant's water needs (975 mm of irrigation water).