近20 a中亚净初级生产力与实际蒸散发特征分析

中亚碳、水循环在气候变异和人为活动的影响下呈现新的时空特征。但由于观测数据稀缺，生态过程特殊，植被、土壤空间异质性强，中亚植被净初级生产力(NPP)、实际蒸散发(AET)的时空特征相关信息相对不足，且时效性不高。利用全球尺度的NPP、AET、土地覆被数据，气象站点与区域气候数据分析近20 a中亚地区NPP和AET的时空特征。结果表明：与1990年相比，2000年中亚地区农田NPP增幅小于自然植被，植被总固碳量增加了254.65 Tg C；近20 a中亚地区实际总蒸散量先增后降，农田对中亚水资源散失的贡献减小，自然植被的贡献增大，自然植被与农田面积变化决定中亚总蒸散量动态；北部农田区、东部山区及山前绿洲为NPP和AET的高值区，中西部荒漠为低值区。

Net primary productivity and actual evapotranspiration of Central Asia in recent 20 years

In the past decades,climate change and human activities have caused new tempo spatial characteristics of carbon and water cycle of Central Asia (CA). Net primary productivity (NPP) is an effective indicator of the interaction between vegetation and environmental factors,and plays a decisive role in carbon cycle. That about 90% precipitation in arid land returned to atmosphere through actual evapotranspiration (AET) made [WTBX]AET[WTBZ] the dominant factor that controls the water balance of arid region,especially in CA. However,due to special ecological processes,high spatial heterogeneities of vegetation and soil,and limited observation data on biome and climate,traditional methods such as field based inventory and ecological modeling have not been conducted to investigate the tempo spatial patterns of NPP and AET of CA. Long term series remote sensing data is independent of ground observation data and capable of detecting tempo spatial changes of NPP and AET. However,previous remote sensing based global or continental studies on NPP did not carried out specified analysis in CA and most evapotranspiration studies only focused on the impacts of environmental factors change on pan evaporation which significantly differed from AET. In this study,we employed remote sensing based NPP data from AVHRR Global Production Efficiency Model (1990 and 2000),AET data from Global terrestrial ET dataset (1990) and MOD16 Global Terrestrial Evapotranspiration Dataset (2000 and 2009),land use data from DISCover (1992-1993),MODIS LC-1 (2000) and GlobCover (2009),site observation meteorological data from Global Surface Observation Datasets (1990-2009),and reanalysis climate data from Climate Forecast System Reanalysis of NOAA's National Centers for Environmental Prediction (1990-2009) to analyze tempo spatial changes of temperature,precipitation,NPP and AET of CA in recent 20 years. The remote sensing datasets in our paper were well validated against field observation data at regional or global scales. The results indicated as follows: (1) climate change showed great tempo spatial variability in CA during the past 20a,temperature and precipitation increase rate were much faster than global or continental average level,and climate change trend was warm dry in middle and western CA during 1990-2000,cold wet in middle and southeastern during 2000-2009,and warm wet in middle and southern CA in 1990-2009; (2) from 1990-2000,carbon sequestrated by vegetation of CA increased 254.65 Tg C,among which natural vegetation increased by 621.08 Tg C and crop decreased by 366.43 Tg C; NPP of crop increased less than that of natural vegetation; crop land area is the controlling factor that determined the increase of carbon sequestration by crop; (3) total actual evapotranspiration increased in the first decade then decreased in 2000-2009,and less water of CA were lost from crop while more from natural vegetation,which were mainly resulted from the nature vegetation area increased during 1990-2000 and decreased in 2000-2009; (4) NPP and AET were higher in northern crop area,eastern mountainous region and oases while lower in central and western desert.