北方农牧交错区干旱特征变化及其对植被总初级生产力的影响

北方农牧交错区地处半湿润/半干旱生态脆弱过渡带,干旱是影响该区植被生产力的关键因素之一。探究干旱对植被总初级生产力的影响,对深刻理解气候变化下生态系统生产力变化响应特征及优化区域碳水循环具有重要意义。为了更好地了解水分限制区不同干旱特征对GPP影响,本研究以北方农牧交错区为例,基于长时间序列的标准化降水蒸散发指数(SPEI3,1900—2020年)和植被总初级生产力(GPP,1982—2018年)等数据,首先采用小波分析明确SPEI3与GPP强相关周期,在此基础上利用游程理论识别干旱特征,进而分析了北方农牧交错区干旱特征与GPP的变化趋势,最后厘定了不同干旱特征对GPP的影响。结果表明:(1) 1982—2018年北方农牧交错区SPEI3与GPP在半年周期和年周期存在显著相关关系,滞后效应随时间变化而变化;年际分析能够减弱滞后效应对SPEI3与GPP相关性的影响;(2) 1900—2020年北方农牧交错区干旱历时、干旱烈度和烈度峰值均呈现显著增加趋势,干旱烈度随着干旱历时和烈度峰值的增加而加剧,干旱特征高值区往往具有更强的增加趋势;(3) 1982—2018年北方农牧交错区GPP总体呈...

Spatio-temporal Changes of Drought Features and Their Impacts on the Gross Primary Production in Farming-Pastoral Ecotone of Northern China

Climate change has been projected to increase the drought duration and intensity in the future, with greater impact on the Gross Primary Production (GPP) in areas. However, despite our increasing understanding of the drought impacts on GPP in recent decades, there is still little understanding of how different drought features (i.e., duration, intensity, peak) GPP. A better understanding of the impact of different drought features on GPP under climate change can help optimize the regional carbon water cycle. In this study, we took the Farming-Pastoral Ecotone of Northern (FPEN) in China as an example and used long-term time series Standardized Precipitation Evapotranspiration Index (SPEI3) during 1900—2020 and GPP during 1982—2018. Firstly, the cross wavelet transform and wavelet coherence analysis was employed to reveal the correlation relationship between SPEI3 and GPP at different temporal scales. Then, the drought features were identified by runs theory, and the trend of drought features and GPP were analyzed using the Theil-Sen Median trend analysis with the Mann-Kandall test in FPEN. Finally, the effects and relative contributions of different drought features on GPP were determined by partial correlation analysis and multiple linear regression analysis. The results show that: (1) There was a strong correlation between SPEI3 and GPP at the scales of half year (6±1 month) and 1 year (12±2 month) in FPEN during 1982—2018, and the 1-year scale had a stronger relationship. Meanwhile the lag effect of SPEI3 and GPP changed with time. Therefore, we clarified that the interannual analysis was able to reduce the effect of lag effect on the correlation between SPEI3 and GPP; (2) The drought duration, intensity, and peak all showed a significantly increasing trend in FPEN during 1900—2020. Drought intensity increased with increasing drought duration and peak. The drought features with higher values often had a stronger increasing trend; (3) GPP showed a trend of “increase-decrease-increase”, with a significant increasing trend during 1982—2018, and the high values of GPP showed a stronger increasing trend; (4) Different drought features have different effects on GPP changes. The absolute contribution rate of drought features was in the order of intensity > duration > peak. Overall, drought features together explained 18.1% of the GPP change. Drought duration and intensity significantly inhibited GPP growth, with the relative contribution rate of -0.05 and -0.35, respectively; (5) GPP of different land covers responded differently to drought features. The drought duration and intensity has negative contribution to the GPP changes in forests/shrubs and grasslands/croplands, respectively.