نوع مقاله : مقاله پژوهشی
نویسندگان
1 گروه مهندسی آب، دانشگاه علوم کشاورزی و منابع طبیعی ساری
2 دانشگاه تهپژوهشگر پسادکتری، مؤسسه ملی تحقیقات علمی (INRS)، مرکز آب زمین محیطزیست، کبک، کانادا. ران
چکیده
کلیدواژهها
موضوعات
عنوان مقاله [English]
نویسندگان [English]
Investigating the effect of agricultural management methods on the amount of nitrate and urea in the soil has special importance. The purpose of this study is to model and investigate the interrelation of hydraulic, reactivity, and solute absorption variables of different soil depths collected in the year of (2020-2021) from the pilot rice farm located in the Sari Agricultural Sciences and Natural Resources University, Mazandaran province, with an area of 0.6 hectares. In this study, the residual moisture of the soil (θr), nitrification rate (kn), and urea hydrolysis rate (kh) variables were modeled based on four, four, and two defined scenarios, respectively, using Wavelet-Artificial Neural Network (WANN), Wavelet- Artificial Neural-Fuzzy Inference System (WANFIS), and Wavelet-Gene Expression Programming (WGEP) models. The results showed that the performance improvement percentage of WGEP models compared to WANFIS and WANFIS model compared to WANN considering the RMSE evaluation index were obtained (16.96, 41.87), (85.72, 1.00), and (20.37, 3.27) for three variables of θr, kn, and kh, respectively. the volumetric residual moisture in the soil, and the urea hydrolysis rate variable is also highly dependent on the residual moisture in the soil. Also, the results showed that the hydraulic variables, reactivity and absorption of soil solutes can be affected by the climatic conditions of the region. Therefore, providing intelligent applicable models to estimate nitrate and urea variables in soil can help managers and farmers in proper management of water and soil resources and optimal use of nitrogen fertilizer with less time and cost.
کلیدواژهها [English]