Data Envelopment Analysis in Uncertainty Conditions to Improve the Energy Consumption in Mazandaran Kiwi Orchards

One of the principal requirements for sustainable agriculture is efficient energy use. Energy use in agriculture has been increasing in response to the growing global population, limited arable land and desire for higher living standards. It should be noted that agriculture contributes significantly to atmospheric GHG emissions, with 10-12% of the net global CO2 (carbon dioxide) emissions. The scientific community believes global warming will pose one of the major environmental challenges in the future, with the bulk of GHG originating from fossil fuel consumption. Kiwifruit is an economically important fruit crop in northern Iran, because the northern region of Iran is a suitable, natural habitat for kiwifruit cultivation. The high kiwifruit production in Iran has reached a point that Iran is now well-known on global markets and in recent years this fruit has contributed a large share to agricultural exports. More recently, Mazandaran horticulturists have been encouraged to produce more kiwifruit. Increased production leads to greater energy consumption by Iranian kiwifruit orchards due to the added application of inputs, such as fertilizers and fuel. Besides, where there is no clear energy consumption pattern in agricultural production, especially fruit orchards, a lot of energy dissipates in the fruit production cycle. Therefore, it seems necessary to provide a model for the energy consumption of kiwifruit orchards in Mazandaran Province to prevent excessive energy utilization. Energy analysis is one of the methods has been used to evaluate the status of agricultural production. In this regard, many researchers have used Data Envelopment Analysis (DEA) for optimization the energy consumption in agricultural productions. DEA is recognized as a methodology widely used to evaluate the relative efficiency of a set of decision-making units (DMUs) involved in a production process. Although DEA is a powerful tool to measure efficiency but the uncertainty in the applied data in this model is inevitable and there is need to use different models that be able to control this uncertainty.

In this study, in order to determine the efficiency of kiwifruit orchards in Mazandaran province and in terms of uncertainty of input data, the Robust Data Envelopment Analysis model (RDEA) and Fuzzy Interval Data Envelopment Analysis model (FIDEA) were used. The method incorporates the degree of conservatism in the maximum probability bound for constraint violation. The required data were collected by distributing and completing a questionnaire and face-to-face interview using random sampling method in 1397-98.

The results showed that the average technical efficiency of all kiwi fields in RDEA model at three levels of probability include: 10, 50 and 100% is equal to 0.93, 0.96 and 0.98%, respectively. The results of FIDEA model showed that if the level of parameter α (optimal use of production factors) increases, the average efficiency of kiwi fields will increase. The highest energy savings are related to chemical pesticides and the lowest amount of savings is related to the chemical fertilizers and electricity inputs, respectively. So, holding the training courses on the correct and optimal use of production inputs from an economic and managerial point of view and improving the level of knowledge of farmers and factors involved in kiwi production in Mazandaran province can improve the efficiency and save energy consumptions.

Evaluating the performance of many activities by a traditional DEA approach requires precise input and output data. However, input and output data in real-world problems are often imprecise or vague. To deal with imprecise data, this study uses RDEA and FIDEA approaches as a way to quantify vague data in DEA models. It is shown that the approaches can be a useful tool in DEA models without introducing additional complexity into the problem. A case study of kiwifruit orchard units is presented to illustrate the reliability and flexibility of the models. As a result, efficiency decreases as the constraint violation probability increased. Additionally, the RDEA approach provides both a deterministic guarantee about the efficiency level of the model, as well as a probabilistic guarantee that is valid for all symmetric distributions.