Spatial-temporal Simulation Of Vegetation Carbon Sink And Its Influential Factors Based On Casa And Gsmsr Model In Shaanxi Province

In order to estimate vegetation carbon sink of Shaanxi Province quantitatively, assessing carbon budgets and responses to natural, cultural elements. Based on the CASA(Carnegie Ames Stanford Approach) model, this study estimates the monthly vegetation NPP(Net Primary Productivity) of Shaanxi Province in 2003~2012. Then stimating the same period soil respiration(Rs) of Shaanxi Province based on the GSMSR(Geostatistical Model of Soil Respiration) model, which is suitable to regional scale. Relational equation between Rs and Rh is built by using the measured datas. Then NEP(Net Ecosystem Productivity) can be gotten, which is used to assess the carbon sink/source conditions of Shaanxi province as a quantitative indicators. And analyzing the influence of temperature, precipitation, solar radiation, GDP, population and other factors on vegetation carbon source/sink. This study is composed of four parts, main conclusions are below.1 Analysis of temperature and precipitation rasterization based on optimized parameters in Shaanxi ProvinceDuring the experimental comparison of 4 categories 11 different methods, the temperature of Shaanxi Province using “Regression + Residual IDW(Inverse Distance Weighting)” to rasterize, whose accuracy is highest. And its MAE and RMSE are lowest, which are 0.498, 0.775 respectively. Its goodness of fit R2 is 0.9548 with the true value, but the process is more complicated. For the precipitation, the OK(Ordinary Kriging) method is the most accurate and operated simplly, R2 is 0.9478.2 Spatial-temporal analysis of vegetation net primary productivity in Shaanxi Province based on CASA model1) The vegetation NPP of Shaanxi Province from 2003 to 2012 shows a significant growth trendency at a rate of 3.9406 g C/( m2·a). The total annual NPP increases from 84.44 TgC in 2003 to 91.98 TgC in 2012, indicating the improvement of vegetation cover and ecological environment.The NPP changes obviously in one year, which is as high as 52.43% in summer. Spring and autumn are lower, accounted for 23.93%, 19.64%. And winter is lowest, which is 4.00 %.2) Annual variation of NPP for different vegetation types are comparatively stable. The mean values are sorted below: evergreen broadleaf forest > deciduous broadleaf forest > mixed forest > grassland > cultivated land > shrub > evergreen coniferous forest > permanent wetland > bare land and sparse vegetation. The vegetation show an increasing tendency except the permanent wetland. The crop land improves the fastest(P < 0.01), having an annual average increment of 5.89gC/(m2?a). Bare land and sparse vegetation, mixed forest, shrub, grassland improve more than 2gC/(m2·a)(P<0.05).3) NPP spatial distribution is relative great with a trend of the south NPP is higher than the north one in generally, which expressing latitude zonality. 78.53% of the area have a growing trend in NPP, while 24.47% of which increases significant/ extremely significant. 21.47% of the area has a decreasing trend in NPP, only 2.27% of which is significant/ extremely significant, which is mainly distributed in the central region of Shaanxi and around area of Xi’an. The results show that the vegetation growing conditions in Shaanxi Province is improving integrally, but deteriorating partly.3 Spatial-temporal simulation of vegetation carbon sink based on GSMSR model in Shaanxi Province1) The vegetation NEP of Shaanxi Province from 2003 to 2012 shows a significant growth trendency at a rate of 4.411 gC/(m2?a), expressing the capacity of carbon sink is increasing, particularly in northern Shaanxi. The total annual NEP increases from 18.24 TgC in 2003 to 27.60 TgC in 2012, indicating the capacity of carbon sequestration in ecosystem is enhanced.2) Monthly average NEP of different vegetation types in Shaanxi Province shows obvious "Peak-Valley" characteristic, which also has stratification among different vegetation types. NEP of evergreen broadleaf forest and deciduous broadleaved forest are high, and wetlands, bare land are low in generally. Apart from the bare land and sparse vegetation is carbon source whole the year, the carbon source/ sink properties of other vegetation types are significant seasonal, which are carbon sinks in summer half year and carbon sources in winter half year.3) NEP spatial distribution is relative great with a trend of the south NEP is higher than the north one in generally, which expressing latitude zonality. 81.25% of the area have a growing trend in NEP. And bounded by Yan’ an, the NEP in most parts of north Loess Plateau is lower than-25 gC/(m2·a), and the south of that is higher than 0 gC/(m2?a). Highest NEP areas are spread in Qinling Mountains, ziwu mountains and huanglong mountain, etc. And lowest NEP areas are spread in Taibai mountain, Xi’ an, and around area of the Great Wall.4 Analysis of factos affecting vegetation carbon sink in Shaanxi Province1) From the perspective of single factor analysis, temperature, precipitation, solar radiation and NDVI have significant influence on the NEP, having more than moderate positive correlation with NEP(P < 0.01). The correlation coefficients with NEP are sorted from large to small as below: temperature > NDVI > solar radiation > precipitation. It’ s highest that the coefficient of determination between NEP and temperature, whose R2 is 0.9362. Therefore, the regression equation can be used to estimate NEP approximately.2) There exists Threshold Effect in the influence of temperature, precipitation, solar radiation, NDVI on NEP. For example, when the monthly average temperature is between-5~9℃, there is a stable relationship between temperature and NEP, expressing weak negative correlation; when the temperature is between 9~22℃, it’s positively correlated with NEP, and linear growth trend is obvious; when the temperature is between 22~26℃, the positive trend slowes and becomes stable.