R. Heydari Kamalabadi; S.A. Hosseini Yekani; S.M. Mojaverian; A.R. Nikooie
Abstract
Introduction: Uncertainty existence in farmers crop production pulsed on important and necessity of science of risk management in the agricultural sector. The new risk management selects the best tools and techniques to minimize risks and consequences of decisions. Furthermore, determining the nature ...
Read More
Introduction: Uncertainty existence in farmers crop production pulsed on important and necessity of science of risk management in the agricultural sector. The new risk management selects the best tools and techniques to minimize risks and consequences of decisions. Furthermore, determining the nature of the risk of crops yield can provide useful information about how to manage the risk of the agricultural sector. One of the effects of climate change is caused damage in the agricultural sector. Dependence of crops to climate change is caused that climate factors have a determinative role in the occurrence of crops damaged. Performed studies on the economic effects of climate change have shown that climate change has a significant impact on agricultural yield and its production risk. Moreover, climate change influences crop yield and the risk of crop yield. Although several studies have been carried out about the impact of climate change on crops yield in Iran, the effect of climate change on crops yield risk is infrequently considered. Therefore, this article tries to offer a new way for calculating the risk of crops yield using of CVaR in the period 2017-2047 in the zayanderud agricultural system. The innovation of this study can be stated as follows:1) This study used of Value at Risk index, as one of the most important indicators of risk measuring, to measure the risk of crops yield, 2) For calculating of Value at Risk index, different studies are used from a famous probability distribution such as normal distribution, historical data or Monte-Carlo simulation, while in this study tried to calculate VaR index based on the forecasted scenarios of crops yield, and 3) In this study, in order to produce future scenarios of crops yield is used from ANN-PSO combined method for forecasting crops yield.
Materials and Methods: The method of this study includes the following steps:
1) The production of possible scenarios of temperature and precipitation using of AOGCM models: Today, one of the best tools for the production of climate scenarios is Atmosphere-Ocean General Circulation Models (AOGCM).But the main problem in the use of the output of the AOGCM models is the large spatial scale of their computational cells toward the area under study. LARS-WG model is also one of the most famous models to small scale for outputs of AOGCM models. In this study uncertainty related to AOGCM models, is used of for scenarios of all AOGCM models(including A1B, B1 and A2).
2)The production of scenarios of selected crop yield and available water in the period 2017-2047: The production of scenarios of selected crops yield and available water is performed using of combinedmethod of ANN-PSO.To combine neural network with particles warm optimization algorithm, from particles warm optimization algorithm is instead of training the neural network using gradient-based algorithms.
3) Measuring risk of crops yield using of VaR and CVaR indexes: VaR index is one of the most important criteria to measure downside risk that it determines the maximum amount of expected losses of a variable for a certain time period and specific confidence level. In this study (according to the non-normal distribution of crops yield scenarios) is used on the historical simulation approach.
Results and Discussion: In the first phase of research methodology, for producing of climate scenarios from daily available stats related to weather stations of Isfahan, Kabuotarabad, Kuohrang, and Daranwere used. Validation results of LARS-WG model showed that this model is well able to simulate changes of climate parameters. Eventually, 44 scenarios of the maximum temperature, minimum temperature and rainfall wereproduced in each studied stations and for each year. The results of the network design using trial and error methods revealed the best forecast combination model obtained with 3 and6 neurons in the input layer and hidden layer of neural network and assuming the initial population of 200, in PSO algorithm. Results of this step showed that ANN-PSO model is well able to forecast crops yield (wheat, barley, maize and alfalfa) and available water. Furthermore, calculating VaR and CvaR criterain confidence level %95 and for future period of 2017-2047, showed that the values of these two criterions for wheat, barley, maize and alfalfa were equal to (4240, 4205), (4062, 4057), (49061,48480) and (10875,10743) kg/ha. The comparison of the values of these two criterions with the values of last period also showed that for all selected crops, VaR and CvaR criterions is bigger in future period toward last period.
Conclusions: The new offered method can calculate the risk of crops yield due to climate change. The more accurate measuring of risk using of new methods such as CVaR can be suitable guidance for policy man to better management of production risk of crops.
