A hybrid extreme learning machine approach for modeling the effectiveness of irrigation methods on greenhouse gas emissions

Indiscriminate use of water resources in the agricultural sector, decrease in precipitation, and increase in greenhouse gases (GHG) are the most influential factors in creating a crisis in meeting the ecological needs of Lake Urmia. In present study, the amount of GHG emissions due to changing irrigation practices was estimated to reduce water consumption. This investigation was carried out in-field using a hybrid approach called the anti-coronavirus optimization (ACVO) algorithm and extreme learning machine (ELM) to estimate carbon dioxide (CO2) levels. The ACVO-ELM approach was considered on a field dataset, applying irrigation-fertilizer, tillage methods, and crop varieties of selected farms located in the Lake Urmia basin from 2020 to 2021. Five various input combinations were introduced to evaluate CO2 emissions. The capability of the ACVO-ELM approach was computed with the statistical indicators of the Nash-Sutcliffe model efficiency coefficient (NSE), root mean square error (RMSE), and root mean square relative error (RRMSE). Field monitoring outcomes demonstrated that by changing irrigation methods, the amount of CO2 emissions in cultivating wheat, tomato, and sugar beet crops was 1766 KgCO(2)e ha(-1), 2917 KgCO(2)e ha(-1), and 3933 KgCO(2)e ha(-1), respectively. Comparing the treatment to the control, the amount of CO2 is dropping by 25%. Modeling results displayed that the ACVO-ELM approach with the average RMSE = 0.009, NSE = 0.975, and RRMSE = 0.075 has good accuracy in estimating CO2 emissions. Model L4 provides a more optimistic estimate of CO2 emissions by applying parameters of irrigation-fertilizer, and tillage methods as model inputs.