基于空间降尺度的中国陆地年净生态系统生产力及其趋势的空间变化

Spatial variations of annual net ecosystem productivity and its trend over Chinese terrestrial ecosystems based on spatial downscaling

年净生态系统生产力(NEP)是一年内生态系统净固定的有机碳量,是形成碳汇的基础。量化NEP及其趋势的空间变化有助于增进生态系统对环境变化适应与响应的认知,并服务于应对碳中和的区域碳管理。基于过程模型和数据驱动模型输出的NEP栅格数据,结合多站点涡度相关观测结果,本研究获得反映NEP空间变化的最优模型模拟结果,并发展空间降尺度方法获取中国陆地高分辨率NEP数据,分析2000—2017年中国陆地NEP及其趋势的空间变化及影响因素。结果表明:相较于过程模型模拟结果,数据驱动模型模拟的NEP更能反映观测NEP的空间变化,并可基于随机森林回归结合简单扩展法实现NEP的空间降尺度。2000—2017年,空间降尺度所得中国陆地NEP总量平均值为(1.30±0.03)Pg C·a^(-1),年际之间呈现先降低后增加的趋势,在2009年出现拐点。NEP呈现自东南向西北逐渐减少的空间分布,表现为纬向递减和经向增大格局,主要受气候与生物因素的共同影响。NEP趋势呈现东高西低的空间分布,仅随经度增加呈现微弱的增大趋势,且主要由光合有效辐射和土壤有机碳含量的作用所引起。中国陆地NEP及其趋势的空间变化存在明显差异,体现出生态系统对环境变化适应与响应的明显不同。

Annual net ecosystem productivity(NEP),the amount of net carbon sequestration during a year,serves as the basis of terrestrial carbon sink.Quantifying the spatial variations of NEP and its trend would enhance our understandings on the response and adaption of ecosystems to environmental change,which also serves for the regional carbon management targeting at carbon neutrality.Based on process-based model and data-driven model simulating NEP,we selected the optimal simulating NEP mostly representing NEP spatial variations with multiple site eddy covariance measurements to develop the spatial downscaling method and generate high resolution NEP data of China,which was used to examine the spatial variations of NEP and its trend and driving factors during 2000-2017.Compared with process-based model results,data-driven model simulating NEP could mostly represent the spatial variation of site measurements.The random forest regression based on climate,soil,and biological data com-bining with the simple scaling could successfully downscale NEP to a high spatial resolution.From 2000 to 2017,the total amount of NEP in China was(1.30±0.03)Pg C·a^(-1),showing a decreasing-increasing pattern with the inflection point in 2009.Chinese NEP decreased from southeast to northwest,showing a descending latitudinal dis-tribution and an ascending longitudinal distribution,with the combined effects of climate and biotic factors.NEP trend decreased from east towards west,which was only accompanied with a slightly ascending longitudinal distribu-tion,while photosynthetically active radiation and soil organic carbon content dominated the spatial variations of NEP trend.Therefore,the spatial patterns of generated NEP obviously differed from those of NEP trend,suggesting the obvious difference between the responses and adaptions of ecosystems to environmental changes.