Introduction
Improving the sustainability of agricultural systems requires the optimisation of inputs use, especially chemical fertilisers and pesticides, which are major contributors to energy consumption and environmental degradation. Variable Rate Technology (VRT), as a precision agriculture strategy, allows site-specific management of inputs based on spatial variability in soil fertility, weed distribution, and crop requirements. Although VRT has shown promise in enhancing resource-use efficiency, its integrated effects on energy indicators and environmental burdens in wheat production under irrigated conditions remain insufficiently investigated. This study assessed the impact of VRT on energy performance and environmental emissions in comparison with conventional uniform application, using a real six-hectare winter wheat field in Karaj, Iran.
Materials and Methods
A comprehensive field-scale simulation of VRT was conducted for nitrogen, phosphorus, potassium fertilisers, herbicides, and insecticides. Spatial maps of soil fertility and weed distribution were generated using UAV-based remote sensing and ground sampling. Two scenarios were examined: (1) VRT-based variable application of chemical inputs and (2) conventional uniform application. Energy inputs and other outputs were calculated based on standard coefficients and categorised as direct, indirect, renewable, and non-renewable. Environmental impacts, including Global Warming Potential (GWP) and pollutant emissions to air, water, and soil, were quantified using the ReCiPe 2016 Midpoint (H) method. All results were compared for the production of 34,800 kg of wheat.
Results and Discussion
Energy Indicators
The total energy input under VRT (131,631.57 MJ) was lower than that of consumed under conventional management (160,318.55 MJ). Direct and indirect energy uses declined by 12.64 MJ and 19.31 MJ, respectively, in the VRT system. VRT improved all energy indicators i.e., energy ratio increased to 6.82 (21.79% higher than the conventional method), energy productivity rose to 0.264 kg MJ⁻¹, and energy intensity decreased to 3.78 MJ kg⁻¹. Net energy under VRT reached 765,705.76 MJ, exceeding the conventional value. Major reductions were attributed to substantial decreases in herbicide use (80.40%) and potassium fertiliser (77%), driven by UAV-derived weed distribution maps and soil fertility maps.
Environmental Impacts
The GWP of the VRT scenario was 17,691.21 kg CO₂-eq, representing approximately a 20% reduction compared to the 22,202.74 kg CO₂-eq emitted under conventional application. Nearly half of the GWP originated from direct field emissions, followed by nitrogen fertiliser use. Optimisation of nitrogen rates and reduced field-level emissions were the primary contributors to the decrease. Pollutant emissions to the atmosphere also declined significantly: CO₂ by 16.2%, N₂O by 21.9%, and NH₃ by 22.5%. Waterborne pollutants were reduced, with nitrate declining by 22.4% and phosphorus by 38.9%. Heavy metal emissions to soil also decreased, with Pb reduced by 22.4% and Zn by 35.0%, while elements with naturally low accumulation (Fe, B, Mn, Mo) remained unchanged. These improvements align with previous international findings demonstrating the effectiveness of UAV-assisted VRT in reducing chemical inputs, enhancing energy efficiency, and minimising environmental pollution across diverse cropping systems.
Conclusion
The results demonstrate that implementing VRT for applaying chemical inputs in wheat production substantially reduces both energy consumption and environmental impacts. Compared with the conventional uniform method, VRT improved all energy indicators, decreased total energy inputs, and increased net energy output. Environmentally, VRT reduced GWP by roughly 20%, lowered key atmospheric pollutants, and substantially decreased nutrient leaching and heavy metal accumulation in soil. Overall, VRT proves to be a highly effective strategy for achieving sustainable wheat production through optimised input management, enhanced energy efficiency, and minimised ecological burdens.
Acknowledgement 
The authors gratefully acknowledge the support of the agricultural research team and field specialists involved in data collection, UAV operations, and soil and crop analyses throughout the study.