| As the typical arid and semi-arid area in China,the vegetation of Xinjiang is especially sensitive to the impact of climate change and human activities.Grassland is the most important vegetation type in Xinjiang,which area is the third largest by total grassland area in China.In the context of global change and regional responses research,carbon sink function of grassland has gradually attracted the attention of the general public.The objectives of the paper are to study the spatial and temporal variability of the Xinjiang vegetation and grassland productivity and its climatic response characteristics,and to distinguish the relative role of the two diving factors-climate change and human activities and provide new techniques for estimation of soil carbon stock in Xinjiang grassland.In the paper,the grassland hyperspectral characteristics and prediction of soil organic carbon,and the spatial-temporal features,impacts of some influence factors(such as land use and cover change(LUCC),climate variation)and its responses of vegetation and grassland productivity in Xinjiang were studied based on MODIS data and other climatic data by using CASA model.In addition,BIOME-BGC model was also employed to evaluate the spatial and temporal pattern of NPP and NEP in three typical grasslands in Xinjiang.The response characteristics of ecosystem productivity(NPP,NEP)of the three typical grasslands were also analyzed from the perspective of the source-sink relationships and climate change scenarios in future.Meanwhile,the relative role of climate change and human activities on vegetation productivity degradation were evaluated and distinguished qualitatively and quantitatively based on the analysis change trend of NPP.The main contents and results are as follows:(1)The arid desert meadow,artemisia desert meadow,and lowland mountain meadow are three main grassland types of Xinjiang Uygur Autonomus Region of China.Using ground object spectrometer,reflectance spectral characteristics of the three main grassland types were measured and analyzed for assisting the extraction and dynamic monitoring of grassland.The results showed that the canopy spectral reflectance of arid desert meadow was smaller than that of artemisia desert meadow and lowland mountain meadow in the visible waves except Ceratocarpus arenarius Linn,while in the near infrared waves,that of Ceratocarpus arenarius Linn,Peganum harmala L and Haloxylon ammodendron was obviously bigger than that of artemisia desert meadow and lowland mountain meadow.Because of the difference between vegetation type and different internal structure of leaves,the differences of spectral reflectance between different vegetations that belong to the same type were significant in the visible waves and the near infrared waves.The value of red edge position of Haloxylon ammodendron of arid desert meadow was bigger than those of artemisia desert meadow and lowland mountain meadow.The value of D?red and Sred of Peganum harmala L of artemisia desert meadow was bigger than those of arid desert meadow and lowland mountain meadow,those of Carex liparocarpos was the smallest.Six vegetation indexes are analyzed.PRI,OSAVI and MCARI are the biggest for arid desert meadow vegetation and the smallest for artemisia desert meadow vegetation.NDVI is the highest for lowland mountain meadow and the lowest for arid desert meadow.In addition,GNDVI is the highest for lowland mountain meadow and the lowest for artemisia desert meadow.(2)The spatial distribution of NPP in Xinjiang grassland was constrained by the regional hydrothermal conditions,grassland types generally occurred by following alpine and sub-alpine meadow,plain grassland,meadows,desert grassland and alpine and sub-alpine grassland from north to south.The grassland NPP reduced from 395 gC m-2a-1 to near 0 gC m-2 a-1,10-year average total NPP of Xinjiang grassland was 56.47 Tg C.Compared with other grassland types,mean NPP of meadow was the highest(155.29gCm2a-1),while mean NPP of desert was the lowest(57.68 gCm a-1);the overall level of mean NPP was meadow>alpine and sub-alpine meadow>plain grassland>alpine and subalpine grassland>desert grassland.The overall level of grassland NPP in Xinjiang was rather low,where alpine and sub-alpine meadow,plain grassland and meadow were belonged to lower productiveecosystems;while desert grassland and alpine and sub-alpine grassland were belonged to the lowest productive ecosystems.On the main grasslands in Xinjiang,the NPP from June to August occupied 63.17%of the whole year.The monthly NPP in different grassland existed different change features,but all reached its peak in July.Compared to other types,NPP of alpine and sub-alpine meadow increased by the fastest speed in the first seven months,while the increasing speed of alpine and sub-alpine grassland was the slowest.The decreasing speed of meadow NPP was the fastest than other grassland types in the last five months,while that of desert grassland was the slowest.With the exception of meadow,the mean NPP of other four grassland types all showed a downward trend from 2001 to 2010 in Xinjiang.Total NPP of grassland in Xinjiang reached the maximum(60.21 TgC a-1)in 2007,and dropped to the minimum(53.41 TgC a-1)in 2006.Over the past 10 years,the grassland NPP in Xinjiang had a greater annual fluctuation,and tended to further reduce.(3)Spatial-temporal variations features of grassland NPP from 2001 to 2010 in Xinjiang and its response to climate change were studied and analysed.Total NPP in Xinjiang grassland indicated a decreased trend over the past 10 years,with a mean decreasing rate of 0.225 TgC a-1.During the 10 years,mean NPP of Xinjiang grassland ranged from 100.05 to 112.78 gCm-2 a-1,with a 10-year average total NPP of 56.47 TgC a-1 and a 10-year average mean NPP of 105.79 gCm-2 a-1.NPP of alpine and sub-alpine meadow had a decline rate of 0.59,and accounted for the region's grassland area 18.29%.NPP of plain grassland had a decline rate of 1.17,accounting for 22.45%of the whole grassland area in Xinjiang.As the biggest area of all grassland types,desert grassland had a decline rate of 0.20.NPP of meadow increased with the slope value of 0.23.The area where NPP had the most obviously decreased mainly distributed in western Tianshan and some parts of Yining,with the change rate of-20?-10 gCm-2a-1.There were significant differences among NPP change rates of different grassland types in spatial distribution.The inter-annual relative variability of grassland's NPP was smaller in most parts of Xinjiang,but greater in some eastern grassland area near Qinghai province.In general,grassland's annual NPP had an obvious correlation with annual precipitation.The relationship was the most significant in central-eastern Tianshan Mountains and the northern part of Kunlun Mountains between grassland's NPP and precipitation,with the correlation coefficients of 0.45-0.80.Except very few area,the correlation between annual NPP and temperature in most parts of Xinjiang grassland was not obvious than that of precipitation.Grassland's NPP exhibited a stronger negative correlation with the average annual temperature in southern areas of Altai mountain and north of the Kunlun Mountains,with the correlation coefficient of-0.96?-0.45.It could be caused by these asynchronous hydro-thermal factors,as well as the vegetation growth was more dependent on precipitation than temperature.(4)NPP of vegetation ecosystem in Xinjiang from 2001 to 2010 was estimated using improved light use efficiency model based on MODIS-NDVI,land use classification data and meteorological data.And the relationship between NPP and climate in different temporal unites and scales were analyzed based on correlation coefficients of the two-group elements.Results indicated that the average annual NPP ranged from 59.29 to 65.98 gCm-2 a-1 during 1982-2006,with an average of 62.10 gCm-2a-1.Spatially,NPP distribution exhibited a growing trend from south to north in Xinjiang.There was a significant correlation between NPP and altitude.The spatial distribution of Xinjiang vegetation is strongly dependent on precipitation,while NPP was negative correlation with the temperature.During 2001-2010,total NPP in Xinjiang ranged from 96.28 to 107.14 TgCa-1,with an average of 100.84 TgCa-1.The coefficient of NPP inter-annual variation was higher in north-central and southwest than that in Tarim and Junggar basin.The coefficient of NPP variation was below 0.5 in most parts of Xinjiang during the 10 years,accounting for 96.01%of the whole area.During 2001-2010,the area which the rate of NPP reduction was between 0-10 gCm-2a-1 accounted for 45.41%of the Xinjiang region;while the area which the rate of increase in NPP was between 0-10 gCm-2a-1 accounted for 53.92%of the total area in Xinjiang.The relationship between total yearly NPP and mean annual temperature is not significant(R=-0.318,p>0.05,n=10),however,a significant positive correlation between total yearly NPP and annual precipitation was found(R=0.69,p<0.05,n=10).The precipitation dominated the growth of the vegetation in Xinjiang from 2001 to 2010,the increase in precipitation will significantly promote the growth of vegetation and accumulation of NPP,while the rise of temperature will produce negative effects on the vegetation growth.The correlation coefficients between NPP and precipitation in all vegetation types were higher than that of temperature.In all the vegetation types,the correlation coefficient between NPP and precipitation in desert was the highest than the other vegetation types,followed by grass.Divided by the sensitivity to climatic factors,the region which accounted for 31.06%area was precipitation type,accounted for 12.65%area was temperature type;approximately 12.53%of the area in Xinjiang was compound type,approximately 43.76%of the area was insensitive type.(5)Since the Chinese government initiated its economic reform in 1978,rapid economic development has spurred land use and land cover change(LULCC)in China,resulted in many ecological problems,including land degradation and desertification.To address these increasingly serious ecological crisis,the government launched a series of ecological restoration programs which lead to significant LULCC,resulting in a profound impact on the terrestrial ecosystem.This study uses net primary productivity(NPP)as an important indicator of the arid and semi-arid ecosystem's productivity to estimate the impacts of the LULCC driven by ecological restoration programs in Xinjiang from 2001 to 2009.The modeling method is based upon the Carnegie Ames Stanford Approach(CASA)terrestrial carbon model and uses Moderate-resolution Imaging Spectroradiometer(MODIS)remote sensing data and meteorological data for modeling simulation.The results demonstrate that the forest area of Xinjiang has the most net increase of 9,093 km2 in the study period,compared to other land cover types.The most dominant land cover changes during 2001-2009 were from grassland to forest and mutual transformation between grassland and desert.Total NPP of whole area increased by 252.51 Gg C during the study period.The most obvious increase of total NPP was observed in forest,which has a net increase of 1,782.88 GgCyr-1.It can be concluded that NPP increasing mainly result from forest expansion.During 2001-2009,the mean NPP in forest,grassland and desert had a slight decrease,whereas the cropland and crop/natural vegetation mosaic land mean NPP increased fractionally.By using the climate in 2001 to simulate the NPP of Xinjiang in 2009,we explored the influences of land use and cover changes and climate change on regional NPP.Compared to climate change,human activities produced an obvious positive effect in the increase of total NPP,especially for forest land.As a result,ecological restoration programs produced positive impacts on forest expansion and carbon sequestration in Xinjiang.(6)Climate change and human activities are the two primary driving factors in the process of land degradation,their roles have gradually become a hot spot in the desertification study.In this study,the Miami,Thornthwaite Memorial and Carnegie Ames Stanford Approach(CASA)models and Moderate-resolution Imaging Spectroradiometer(MODIS)remote sensing data and meteorological data were applied to simulate regional potential net primary productivity(NPP)and actual NPP from 2001 to 2010 in North Xinjiang.This paper assessed the relative roles of climate change and human activities in land degradation and vegetation restoration using potential NPP and human appropriation of NPP(HANPP)which represents the difference between potential NPP and actual NPP.HANPP was used as the indicator reflecting the roles of human activities in the process of land degradation or vegetation restoration.The potential NPP and actual NPP were stimulated by the climatic models and CASA model,respectively.The results showed that:(1)human activities was the dominant factor that induced land degradation,accounting for 61.85%(172,228.5km)of the total degradation,whereas 24.47%(68,146.5km2)of the total degradation resulted from climate change.In contrast,56.42%(61,514.5km2)of the total vegetation restoration was dominated by human activities and 33.33%(36,338km2)was caused by climate change.(2)Human activities played a key role in vegetation restoration in the central-southern and western areas of North Xinjiang,and in land degradation in the eastern and northwestern and northeastern areas.Contrastively,climate change dominated vegetation restoration in the southeastern areas of the study region,and land degradation in the southwestern and central-southern areas.Overall,human activities was the dominant factor responsible for the land degradation and vegetation restoration in the study region from 2001 to 2010.(3)The vegetation restoration of grassland and forest dominated by human activities in the study region were mainly attributed to vegetation restoration projects such as afforestation.These results demonstrated that these ecological restoration projects were effective on alleviating land degradation and promoting vegetation restoration in the southern areas of North Xinjiang.(4)This paper provided a preliminary assessment method to identify and assess the roles of climate change and human activities in land degradation and vegetation restoration in arid and semi-arid areas,as well as in various vegetation types.(7)The soil organic carbon(SOC)estimation of grassland in the northern Tianshan Mountains was carried out through the combination of soil spectroscopy and multivariate stepwise linear regression.Twelve types of transformations were applied to the soil reflectance(R)to remove the noise and to linearize the correlation between reflectance and soil organic carbon content.Based on the spectral reflectance and its derivatives,hyperspectral models can be built using correlation analysis and multivariable statistical methods.The results show that the main response range of soil organic carbon is between 350-750 nm,with a maximal spectral response at 355 nm.Correlation analysis indicated that SOC has stronger correlation with the second derivative than with the original reflectance and other transformations data.The two models developed with laboratory spectra gave good predictions of SOC,with root mean square error(RMSE)<5.0.The use of the full visible near-infrared spectral range gave better SOC predictions than using visible separately.By comparing two models,we found that the multivariate stepwise linear regression of second derivate model(Model A)is optimal for estimating SOM content,with a determination coefficient of 0.894 and root mean squared errors(RMSE)of 0.322.The results of this research study indicated that,for the grassland regions,combining soil spectroscopy and mathematical statistical methods does favor accurate prediction of SOC.(8)The improved BIOME-BGC model was used to estimate the NPP and NEP of three typical grasslands in Xinjiang from 2001 to 2010.The results indicated that the mean NPP and NEP of lowland mountain meadow were 122.65 and 8.36 gCm-2a-1,respectively,and that of dry desert grassland were 134.64 and 8.79 gCm-2a-1,and that of artemisia desert meadow 134.20 and 9.26 gCm-2a-1,respectively.Correlation analysis showed that the coefficient between the NPP,NEP of the three typical grasslands and annual temperature was unconspicuous,however,the correlation between precipitation and productivity index(NPP,NEP)was siginificant.The precipitation was dominated factors to control the NPP and NEP accumulation.The change tendency of NPP and NEP were predicted by using BIOME-BGC model under future climate change scenarios.Scenario analysis showed that double CO2 would partly benefit for NPP and NEP.When CO2 concentration fixed,NPP responded positively to precipitation change only.NEP responded negatively to temperature change and positively to precipitation change.When CO2 concentration doubled and climate changed,NPP responded positively to precipitation changed and temperature rise,while NEP responded negatively to precipitation and temperature change.The relationship between carbon source and sink of different types of grasslands in Xinjiang was distinct,and the features of carbon cycle under future climate change scenarios were predicted in this research.As for the innovation of this text,first,the researches on the integration of spatial-temporal pattern,productivity(NPP and NEP)and relative influence factors(LUCC and climate variation)and so on were carried out by combining with the climate characteristic of arid and semiarid regions in Xinjiang,and provided fundamental scientific basis for research of regional response to global change in arid and semiarid regions.Second,potential NPP and human the occupation of NPP(HANPP)were employed to assess the relative role of climate change and human activities on land degradation and vegetation recovery.The relative role of climate change and human activities on vegetation degradation assessment were evaluated from qualitative and quantitative point of view,respectively.The dominated driving factors of land degradation or vegetation recovery in Xinjiang were distinguished and cleared,whose influenced area were also estimated.Third,based on hyperspectral remote sensing technology for spectral characteristics analysis of soil organic carbon in Xinjiang grassland,and combined with the methods of mathematical statistics,two hyperspectral prediction models of soil organic carbon had been established.The most suitable soil organic carbon prediction model was further determined and the accuracy of the prediction were also tested.For the grassland regions,combining soil spectroscopy and mathematical statistical methods does favor accurate prediction of SOC.It also provided a new technique for the estimation of grassland's soil organic carbon pool.In short,the spectral features and SOC prediction of grassland,vegetation productivity and its climatic response,LUCC's impacts,driving factors and so on were studied and estimated by using hyperspectral technology,remote sensing data and mathematical models on the paltform of some geo-spatial softwares,respectively. |
PDF Full Text Request