Recently, Li Fangbai, a researcher at the Institute of Ecological Environment and soil of the School of Sciences of Guangdong Province, cooperated with Zhang Xin, associate professor of Henan Normal University to use machine learning to reveal the synergy mechanism and potential of biological carbon to slow down the cadmium pollution and methane emissions.Related results were published in the “Environmental Science and Technology” in the form of sub -cover articles.
The current journal associate cover.Interviewee confession
Daeta provides food for more than half of the population in the world, and its sustainable development is a issue that is concerned about global attention.However, paddy field heavy metal pollution has become an environmental problem that threatens food security around the world.Cadmium is the most widely existed heavy metal in rice field soil, and the bioccoly accumulation factor in rice is higher than other grains.At the same time, a large amount of emissions of the greenhouse gas (mainly methane) in the rice fields can lead to climate change, and then destroy global rice productivity. Therefore, the emissions of rice field methane further exacerbated the challenge of rice safety.Faced with this universal rice dilemma, biological carbon technology has shown its potential.However, due to the diversity of biological carbon and the lack of mature assessment tools, the multi -target repair of biocision in rice field soil is still a prominent problem.
The Li Fangbai team developed a multi -tasking deep learning (MTDL) model based on the soft parameter sharing algorithm, and explored the potential of biological charcoal to repair the cadmium pollution coordination of rice fields.This model can accurately predict the synergy efficiency of biological charcoal, so as to use biocham coordination to alleviate cadmium pollution and methane emissions in the paddy field soil.In addition, the model also quantitatively the key factor in determining the synergy of cadmium pollution and methane decline in biological carbon characteristics, which will provide manufacturing guidelines and standards for the formulation of ideal biological carbon used for rice soil repair in practice.
Multi -task deep learning reveals biological charcoal coordination to slow down paddy field cadmium pollution and methane emissions.Interviewee confession
In the context of sustainable development goals, rice soil repair requires multiple goals such as climate change, carbon footprint, soil quality improvement, and food production safety. The MTDL model can be further expanded and included in more independent data sets to achieve realizationCollaboration between multiple goals.
This study provides a example of biological charcoal sustainable treatment of soil issues, and the research results have greatly expanded the potential of artificial intelligence in sustainable soil repair.
The above research has been supported by the National Natural Science Foundation of China, the national key research and development plan, the Guangdong Provincial Key R & D Plan, and the Guangdong Science and Technology Plan.