朱颖墨,窦小东,王瑞芳,解明恩,黄玮,李蒙. 2021. 气候变化对云南省小粒咖啡适生区的影响. 气象学报,79(5):878-887. DOI: 10.11676/qxxb2021.049
引用本文: 朱颖墨,窦小东,王瑞芳,解明恩,黄玮,李蒙. 2021. 气候变化对云南省小粒咖啡适生区的影响. 气象学报,79(5):878-887. DOI: 10.11676/qxxb2021.049
Zhu Yingmo, Dou Xiaodong, Wang Ruifang, Xie Ming'en, Huang Wei, Li Meng. 2021. Climate change impact on the region suitable for Coffea arabica growth in Yunnan province. Acta Meteorologica Sinica, 79(5):878-887. DOI: 10.11676/qxxb2021.049
Citation: Zhu Yingmo, Dou Xiaodong, Wang Ruifang, Xie Ming'en, Huang Wei, Li Meng. 2021. Climate change impact on the region suitable for Coffea arabica growth in Yunnan province. Acta Meteorologica Sinica, 79(5):878-887. DOI: 10.11676/qxxb2021.049

气候变化对云南省小粒咖啡适生区的影响

Climate change impact on the region suitable for Coffea arabica growth in Yunnan province

  • 摘要: 为揭示气候变化对云南省小粒咖啡适生区的影响,基于最大熵(MaxEnt)模型,结合小粒咖啡物种分布数据、环境变量数据,构建云南省小粒咖啡适生区评估及预测模型,对当前气候条件下小粒咖啡在云南省的适生区进行评估,并对未来气候条件下,小粒咖啡在云南省的适生区进行预测,再对预测结果进行对比分析。结果显示:(1)构建的最大熵模型能够较精确地用于小粒咖啡在云南省适生区的评估和预测,当前气候条件下,评估模型的训练集与测试集的AUC (Area under ROC Curve)值均为0.941,达到评估结果为极好的标准。(2)影响云南省小粒咖啡种植的主导环境因子依次为11月平均最高气温、7月降雨量、海拔高度、2月平均最低气温、10月降雨量、坡度和最冷月最低气温,共占总贡献率的91.4%。(3)当前气候条件下,小粒咖啡的适生区主要分布在滇西、滇西南以及滇南的保山、德宏、普洱、临沧、西双版纳等地区,总适生区约为116300 km2,占云南省国土面积的29.51%,且总体上,高适生区外围分布中适生区,中适生区外围分布低适生区。RCP4.5、RCP8.5情景下,小粒咖啡总适生区的面积分别约为98300、69700 km2,分别占云南省国土面积的24.95%、17.69%,两种排放情景下小粒咖啡总适生区面积分别减少了18000、46600 km2,国土面积占比分别减少了4.56%、11.82%,且总适生区的质心均由东南向西北方向移动,与RCP4.5情景相比,RCP8.5情景的移动距离更远。(4)未来气候变化将会导致小粒咖啡在云南省的总适生区面积减小,总适生区的质心位置向海拔更高与纬度更高的方向移动,且高碳排放情景下这种变化幅度更大。

     

    Abstract: In order to reveal the climate change impact on the region suitable for Coffea arabica growth in Yunnan province, based on the Maximum entropy (MaxEnt) model combined with the species distribution data and environmental variable data of Coffea arabica, an assessment and prediction model is constructed to evaluate the region suitable for Coffea arabica growth in Yunnan province under current climatic condition and predict the distribution of regions suitable for Coffea arabica growth under future climate condition. The prediction is analyzed. The results are as follows. (1) The constructed MaxEnt model can be used for the assessment and prediction of regions suitable for Coffea arabica growth in Yunnan province. Under current climate condition, the AUC (Area under ROC Curve) values of the training set and the test set of the evaluation model both are 0.941, and the evaluation result meets the standard of "Excellence". (2) The dominant environmental factors affecting the cultivation of Coffea arabica in Yunnan province are the average maximum temperature in November, rainfall in July, altitude, average minimum temperature in February, rainfall in October, mountain slope, and minimum temperature in the coldest month. The contribution rate is 91.4%. (3) Under current climatic condition, the regions suitable for Coffea arabica growth are mainly distributed in western Yunnan, southwestern Yunnan and Baoshan, Dehong, Pu'er, Lincang, Xishuangbanna and other areas in southern Yunnan. The total size of areas suitable for Coffea arabica growth is about 116300 km2, which accounts for 29.51% of the land area in Yunnan. In general, there are moderately suitable areas around the periphery of highly suitable regions, and less suitable areas are distributed around the periphery of moderately suitable areas. Under the RCP4.5 and RCP8.5 scenarios, the total areas suitable for Coffea arabica growth are about 98300 km2 and 69700 km2, accounting for 24.95% and 17.69% of the land area of Yunnan province, respectively. The areas of suitable regions would decrease by 18000 km2 and 46600 km2, and their proportions of land area would decrease by 4.56% and 11.82%, respectively. The center of the overall suitable region would migrate from southeast to northwest. Compared with that under the RCP4.5 scenario, the migration under the RCP8.5 scenario extends further northwest. (4) Future climate change will lead to decreases in the total area suitable for Coffea arabica growth in Yunnan province. The center of the overall suitable region would shift to higher altitudes and higher latitudes, and the magnitude of such changes is larger under higher carbon emission scenarios.

     

/

返回文章
返回