关键词: 碳排放, 青稞（Hordeum vulgare L. var. nudum Hook. f.）种植业, IPCC清单估算, 非线性次回归神经网络, 多步预测, 青藏高原, 青海省

Abstract: In order to study the measurement of carbon emissions from the planting industry on the Qinghai-Tibet Plateau, the carbon emissions, carbon emission intensity and carbon cost of the highland barley （Hordeum vulgare L. var. nudum Hook. f.） planting industry in Qinghai Province were measured. Through the analysis of the total carbon emissions and agricultural carbon emissions in Qinghai Province from 1997 to 2019, it was found that the total carbon emissions in Qinghai Province were still high, and agricultural carbon emission had always been at a high level. Based on the agricultural carbon emission data of Qinghai Province from 1997 to 2019, a nonlinear sub-regression neural network model was built by using GUI toolbox in Matlab software, and a neural network timing tool was used to make multi-step prediction of agricultural carbon emission of Qinghai Province from 2020 to 2026. The results showed that from 2020 to 2026, agricultural carbon emissions in Qinghai Province would still show a trend of high fluctuation growth, and it was urgent to carry out “peak carbon dioxide emissions and carbon neutrality” emission reduction. After analyzing the sources of carbon emission from highland barley planting industry, six measuring indexes of agricultural diesel, artificial, chemical fertilizer, pesticide, agricultural film and N₂O emission had been established. Then, an IPCC inventory estimation model was established based on the data of the established six indicators from 2015 to 2020, and the total carbon emission, carbon emission intensity and carbon cost of highland barley planting industry in Qinghai Province from 2015 to 2020 were measured. The results showed that, the total carbon emission of highland barley planting industry in Qinghai Province from 2015 to 2020 did not increase significantly with the increase of highland barley planting area, while the carbon emission intensity and carbon cost of highland barley planting industry decreased with the increase of highland barley planting area.

Key words: carbon emissions, highland barley （Hordeum vulgare L. var. nudum Hook. f.）planting industry, IPCC inventory estimation, nonlinear sub-regression neural network, multi-step prediction, Qinghai-Tibet Plateau, Qinghai Province