Fertilizer management for global ammonia emission reduction

Area of Science:

• Environmental Science
• Agricultural Science
• Atmospheric Chemistry

Background:

• Crop production is a major source of atmospheric ammonia (NH3), impacting air quality, health, and ecosystems.
• Current global NH3 emission estimates from croplands are uncertain due to data limitations, hindering effective mitigation strategies.

Purpose of the Study:

• To develop a machine learning model for precise, spatially explicit NH3 emission factors from global croplands.
• To quantify potential NH3 emission reductions through optimized fertilizer management.

Main Methods:

• Compiled a dataset of field observations to train a machine learning model.
• Generated crop-specific, spatially explicit NH3 emission factors at 5-arcmin resolution.
• Modeled potential emission reductions and future impacts under climate change scenarios.

Main Results:

• Global NH3 emissions from rice, wheat, and maize in 2018 were estimated at 4.3 ± 1.0 Tg N yr-1, lower than previous estimates.
• Optimized fertilizer management can reduce NH3 emissions by 38% (1.6 ± 0.4 Tg N yr-1) without changing total nitrogen input.
• Potential reductions: 47% for rice, 27% for maize, and 26% for wheat.

Conclusions:

• Machine learning provides a robust tool for estimating and mitigating NH3 emissions from agriculture.
• Targeted fertilizer management is a key strategy for reducing NH3 pollution and its associated risks.
• NH3 emissions are projected to increase under future climate change, but mitigation strategies remain effective.