关键词: 种植业, 农资投入品, 时空特征, 脱钩模型, LMDI分析, 灰色预测模型

Abstract:

Yunnan province was selected as the study area. A carbon accounting system was established for major agricultural inputs, including chemical fertilizers, pesticides, agricultural plastic films, irrigation and mechanized operations. This framework was used to systematically assess the carbon emission properties of crop production from 2005 to 2019. The Tapio decoupling model was applied to examine the evolving relationship between carbon emissions from the use of agricultural production materials and economic growth in the province’s planting industry. The logarithmic mean divisia index (LMDI) decomposition method was employed to quantitatively identify the driving factors. The GM(1,1) gray prediction model was then used to forecast carbon emission trends for the period 2020–2029. The results showed that from 2005 to 2019, carbon emissions from agricultural production material inputs in Yunnan’s crop production followed an inverted U-shaped trend. Emissions reached a peak of 398.67×10⁴t in 2017, before falling to 336.14×10⁴t in 2019. Over the same period, the intensity of carbon emissions per unit of agricultural output value declined each year, with an average annual reduction of 8.47%. Fertilizer use had been identified as the biggest contributor to carbon emissions. The relationship between carbon emissions from agricultural input use and economic growth alternated among four regimes: weak decoupling, expansive coupling, expansive negative decoupling and strong decoupling. After 2018, this relationship shifted from weak decoupling to strong decoupling. The decomposition analysis indicated that economic size, energy structure and agricultural industrial structure were positive driving forces for the growth of carbon emissions, with economic scale having the most significant effect. In contrast, energy intensity and population size were negative driving forces. After confirming that the data were suitable for the GM(1,1) gray prediction model, the forecast results showed that carbon emissions from agricultural production material inputs in Yunnan’s crop production would continue to decline from 2020 to 2029. By 2029, the total is expected to decrease to 177.81×10⁴t. Based on the findings, it is recommended to enhance the technological level of agricultural production in Yunnan, strengthen technical training and the dissemination of low-carbon concepts, optimize cropping structure, and establish sound agricultural management systems. These measures will help improve the carbon reduction efficiency of crop production and support the long-term, green and sustainable development of agriculture in the province.

Key words: Plantation industry, Agricultural production input, Spatial?temporal characteristics, Decoupling model, LMDI analysis, Grey prediction model