Hojjat Emami

The main challenge in the agricultural sector is to increase crop production with minimal water consumption. In this regard, various high efficient irrigation systems have received particular attention. This paper investigates drip (tape) irrigation (DTI). It estimates water productivity (WP), application efficiency (AE) and corn yield in the Miandoab region located in the southeast of Lake Urmia using swarm intelligence methods. The studies were performed in-field monitoring by using a hybrid approach based on an extreme machine learning method (ELM) and tree growth optimization algorithm (TGO) to estimate corn yield, WP, and AE criteria. TGO-ELM method was evaluated on a dataset including irrigation-fertilizer, climate, and soil characteristics of corn fields from 2020 to 2021. The DTI method was compared with the standard furrow irrigation (FI). The field monitoring results showed that the DTI increased corn yield, WP, and AE by 22.2%, 0.58 kg/m3, and 70%, respectively, compared to the FI method. The modeling results showed that model M9 gave a more optimistic estimate of the corn crop, WP and AE by applying the parameters of irrigation levels, rainfall, and soil moisture as model inputs. Also, TGO-ELM with optimal values ​​of R2 = 0.988, RMSE = 0.005, NSE = 0.981, and MAPE = 0.812 had a good performance compared to similar intelligent methods. The results confirmed that TGO-ELM can provide a more accurate estimation of corn yield compared to other similar methods.