نوع مقاله : مقاله پژوهشی
نویسندگان
1 دانشجوی کارشناسی ارشد دانشگاه شهرکرد
2 استادیار دانشگاه شهرکرد
3 استادیار دانشگاه رامین اهواز
4 استادیار دانشگاه ارومیه
5 دانشجوی سابق کارشناسی ارشد گروه خاکشناسی دانشکده کشاورزی دانشگاه رامین اهواز
چکیده
کلیدواژهها
عنوان مقاله [English]
نویسندگان [English]
Volume of vertisols changes by variation in soil water content and it causes re-arrangement of soil pores. Shrinkage curve shows the relation between soil porosity and soil moisture. Many models are presented to quantify shrinkage curve. The objective of this study was to evaluate McGarry and Malafant, Groenevelt and Grant, Peng and Horn, and Cornelis et al models for estimating shrinkage curve of some swelling soils of Dashte Zarrin, Chaharmahal-o- Bakhtiari province, Iran. Therefore, 35 soil samples were randomly selected from 0-30 cm depth of the selected site. Soil shrinkage curves were measured using pressure plate apparatus. The mentioned models were fitted to the measured shrinkage curve data. Then, performance of the models was evaluated using root of mean square error (RMSE), mean error (ME), absolute value of mean error (AME) and Akaike information criterion (AIC). Results showed that the models could describe most parts of the variation of shrinkage curve in the study soils. However, Groenevelt and Grant model with the AIC equal to -57.29, RMSE of 0.00052 cm3.cm-3, ME of -0.00122 cm3.cm-3, and AME equal to 0.0054 cm3.cm-3 was the best model. McGarry and Malafant model had the lowest performance, with AIC equal to -48.82 and RMSE, ME and AME values equal to 0.0122, 0.00342 and 0.0101 cm3.cm-3, respectively. Thus, for predicting the soils shrinkage curve of the study site, application of Groenevelt and Grant model is suggested.
کلیدواژهها [English]