کارایی روش‎ الکتره ترای در ارزیابی تناسب اراضی منطقه چالدران برای کشت سیب زمینی

Abstract
Background and objectives: Land suitability evaluation is inevitable for soil protection. Selection of an appropriate land suitabiliyty method seems to be necessary due to the limitation of the FAO method as a conventional method. In this regard, multi-criteria desision methods (MCDM) such as electrothermics, can be helpful due the considering the weights of effective factors and experts oponions in land suitability evaluation.
Materials and methods: In this research, Square Root Method and ELECTRE Tri multiple criteria method have bean also used for suitability evaluation in an area about of 5000 ha of Chalderan region located between 47° 19′- 47° 23′ East longitude and 39° 01′-39° 08′ North latitude for potato cultivation. In general 60 soil profiles were dug, described and sampled and after soil samples analysis, soil texture, gravels, calsium carbonate, gypsum, EC, ESP, slope, flooding, drainage and region climate in different profile were produced by weighted coefficients. Criteria weights via pair-wise comparison matrix method and ELECTRE Tri profiles and thresholds limits determined using requirements table, soil and land properties data. In two selected land suitability method, land production potential was calculated by multiplication of the soil indices and radiation thermal production potential.
Results: soils of study region has been classified on Entisols, Inceptisols and Aridisols based on 12th Keys to soil taxonomy. Statistics comparison of allocated grades mean to suitability classes’ reveald that results of two methods were significantly different and ELECTRE Tri method had the best grade. Also mean comparing of soil and land properties amount via Duncan test showed significant difference between suitability classes. The ELECTRE Tri methods was the best model from the viewpoint of statistical indices including R2 (0.87) and RMSE (241.2). Also, Root mean square method has a R2 (0.79) and RMSD (197.6). Predicted potential production by Root mean squre method and ELECTRE Tri method was 9330 to 30200 and 10140 to 32350, respectively. The GMER index also indicated lowestimation of two selected land evaluation method. ELECTRE Tri suitability evaluation that implies high accuracy of this method due to determination of transition suitable limits, appropriate values of thresholds and criteria weights as well as usage of fuzzy logic principals on calculation processes.
Conclusion: ELECTRE Tri method, integrated into expert opinion, is better suited to the problem of land suitability for agriculture. The time saving during the development of the land suitability evaluation for the agriculture of the potato was considerable. The coherence of the obtained rsults with observed yield confirms the system effectiveness.