Performance evaluation of corrected land indices to determine the Potential of Maize production using FAO Method

Maize is one the major utilization type in Gobadloo region where placed in East Azarbaijan porivince, North-West of Iran, so performance of land suitability evaluation and land production potential prediction are very important for knowing environmental limitations and planning proper cultivation. FAO guidelines on the land evaluation system were widely used for the land suitability, so soil morphological and analytical data were carried out for 16 land units. Then, based on AEZ model, radiation thermal production potential for Maize was estimated and then soil indices which indicate the extent of soil limitations effectiveness on production reduction, was calculated by the square root formulas. Finally land production potential was calculated by multiplication of the soil indices and radiation thermal production potential. The results reaveald that parametric methods (square root and storie formulas) uncorrected land indices had lower values than which what it was expected in real conditions. For solving this problem land indices were corrected and the results improved land suitability classes. Coefficient of correlation values between land production potential and observed yield were 0.79, 0.84, 0.86 and 0.9 for uncorrected storie, uncorrected root mean square, corrected storie and root mean square models respectively. Based on the results, can conclude that Mean Absolute Error is able to predict yield better than that other methods because of higher regression coefficient and lower error.