بهره‌گیری از شاخص‌های اقلیمی و رویکرد تصمیم‌گیری چند معیاره در پیش‌بینی عملکرد محصول در راستای سیاست‌گذاری کشاورزی

در پیش‌بینی عملکرد محصول فناوری با هزینه کم و دقت بالا برای تسهیل در مدیریت کشاورزی مورد نیاز است. در این راستا عملکرد گندم در استان‌های آذربایجان شرقی و غربی با شاخص‌های اقلیمی (شاخص لانگ، دومارتن،کوپن 1،کوپن 2، کوپن 3، آنگستروم، سیلیانینوف، ایوانف، خشکی، شاخص باران موثر و شاخص پوشش گیاهی) مدلسازی شدند. در این نوع مدلسازی از ترکیب شاخص‌ها به صورت دو، سه و چهار شاخصه استفاده شد. در راستای تصمیم‌گیری در انتخاب شاخص اقلیمی مناسب در هر اقلیم در روند مدلسازی، براساس 5 آماره ارزیابی از تصمیم‌گیری چندمعیاره (topsis) استفاده شد که روش آنتروپی شانون در تعیین میزان وزن شاخص‌ها به‌کار گرفته شد. میزان متوسط افزایش آماره تشابه (sim) برای تمامی شاخص‌ها از استان آذربایجان شرقی به غربی برابر با 13/2 درصد است که نشان دهنده عملکرد بهتر شاخص‌های اقلیمی در استان آذربایجان غربی نسبت به شرقی است. استفاده از ترکیب شاخص‌ها به شرط استفاده از شاخصی با عملکرد بهتر در حالت تک شاخصه دقت بالایی دارد، به عنوان مثال میزان افزایش sim از حالت تک شاخصه به چهار شاخصه در استان آذربایجان شرقی 20/94 است. نتایج حاصل از تصمیم‌‌گیری چندمعیاره نشان داد که شاخص‌هایی ‌با شاخص ایوانف در استان آذر‌بایجان غر‌بی و شاخص خشکی در استان آذر‌بایجان شرقی تاثیر ‌بالایی ‌بر عملکرد گندم دارند. تعیین شاخص اقلیمی موثر در هر منطقه در پیش‌بینی عملکرد محصول به‌عنوان ابزار قوی برای تصمیم‌گیری در مدیریت و اصلاح محصول می‌باشد.

utilization of the climate indices and the multi-criteria decision-making approach in crop yield forecasting in line with policy making in agriculture

introductioncrop yield forecas ting requires low-cos t, high-precision technology to facilitate agricultural management. sus tainable agriculture is of great importance, as it s trives to optimize crop yields. meanwhile, increasing crop yield by reducing environmental impacts is an important and challenging task for sus tainable food supply in this century. wheat is one of the mos t important food products in the world and iran. due to its economic and nutritional value, decision-makers mus t monitor wheat crop growth and yield parameters during the season. therefore, this research aims to es timate crop yield using some climatic indicators. climate indices are different mathematical combinations of meteorological data. each climate index has a specific trend in the simulation of the surrounding environment, so it is necessary to identify the effective index in each region. in mos t of the previous research, the correlation between the indices and the yield of the product was used. however, this research proposed a robus t and comprehensive multi-criteria decision-making approach, which brings an innovative perspective to the crop yield issue. materials and methodsthe data and information of this research were related to two provinces in the northwes t of the country (eas t and wes t azerbaijan provinces). according to the de martonne index, the provinces are located in a semi-arid climate. climate indicators are quantities that can be used to determine droughts and weather variability. climate indicators play a key role in weather monitoring and forecas ting. in this regard, wheat yield was modeled with climate indices (de martonne index, koppen 1, koppen 2, koppen 3, angs tröm, selyaninov, ivanov, aridity, effective precipitation index, and vegetation index). in this type of modeling, the combination of indices was used in the form of two, three, and four indices. mcdm is a general term used to describe a set of methods for s tructuring and evaluating options based on multiple criteria and objectives. these methods provide targeted decisions because they can manage the inherent complexity and uncertainty of the issues, as well as the knowledge resulting from the participation of several factors. mcdm can make the quality of decisions clearer, more efficient, and more logical, which leads to jus tifiable and explainable choices. in addition, mcdm promotes the role of participants in the decision-making process, facilitates compromise and group decisions, and provides a suitable platform for s takeholders to share their personal preferences. to decide on the appropriate climate index selection in each climate of the modeling process, multi-criteria decision-making (topsis) was used based on five evaluation criteria, and shannon’s entropy was used to determine the weight of the criteria.results and discussionthe period used in this research was 22 years, and the period from 2018 to 2021 was considered the verification period. the hurs t coefficient was used to check the length of the s tatis tical period, and the average hurs t coefficient for all indicators is 0.65, which is greater than 0.5, so the length of the s tatis tical period of the series is acceptable. the values of the indices in wes t azarbaijan province have increased compared to eas t azarbaijan; for example, in the ivanov index, the rate of increase from eas t azarbaijan s tation to wes t azarbaijan is 1.38, in the vci index, it is equal to 51.01 and in the aridity index, it is 62.01. using a combination of indices is highly accurate under the condition of using an index with better performance in single-index mode; for example, the rate of sim increase from single-index to four-index mode in eas t azerbaijan province is 20.94. the trend of changes in s tatis tics is not the same in all cases, so to make a comprehensive conclusion about determining the optimal index based on the performance of all s tatis tics, the topsis method was used. in wes t azarbaijan province, in the case of a single index, firs t, the ivanov index and then the aridity index, and in eas t azarbaijan province aridity index, vci, and silyaninov have a high rank. in the case of two indicators, the combination of aridity index and vci in eas t azerbaijan province and the combination of angs trömand ivanov in wes t azerbaijan province have a better rating. in the case of three indicators, in wes t azerbaijan province, ivanov-silyaninov and pei indices have better performance, and in eas t azerbaijan province, aridity-vci and pei indices have better performance. in the case of four indicators, in eas t azerbaijan province, the combination of long-de martonne-aridity and vci indices, and in wes t azerbaijan province, koppen 2-3-angs trömand ivanov has better performance.conclusiondetermining the effective climate index in each region for crop yield forecas ting is a powerful tool for decision-making in crop management and improvement. the results showed that using a combination of indicators is more accurate than using a single indicator. the reason for this problem can be s tated that in the combined mode, corrections are made in terms of the used data and terms of the mathematical s tructure. the effect of the type of indicator in each climate on the crop yield is different. eas t and wes t azarbaijan provinces are very similar in terms of climate, but indices along with the ivanov index in wes t azarbaijan province and the aridity index in eas t azarbaijan province have a high impact on wheat yield. this problem shows that each region has its index in the simulation of climatic processes governing the crop yield. the use of other multi-criteria decision-making approaches or their integrated mode can be one of the sugges tions of this research. by using a suitable and accurate index, the forecas ting of the crop yield becomes closer to the actual values.