亚热带地区马尾松林碳储量的遥感估算——以长汀河田盆地为例

全球气候的变化已使得人类日益关注森林生态系统的碳储量变化.以福建省长汀县河田盆地为例,开展马尾松林碳储量估算模型的研究.通过2010年的野外样地调查获得了马尾松林的实测数据,并将其与同年的ALOS遥感影像对应样地的植被光谱信息进行比较.通过研究5种遥感植被指数与马尾松林碳储量之间的相关关系,从中选取了基于归一化植被指数(NDVI)的研究区最佳马尾松林碳储量反演模型.精度分析表明,该模型平均相对误差为-1.95％,均方根误差为3.01 t/hm2,因此可以有效地用于反演研究区的马尾松林碳储量.利用该模型反演出河田盆地2010年马尾松林的总碳储量为114.58×104 t,碳密度为34.92 t./hm2.

Remote-sensing estimate of the carbon storage of subtropical Pinus massoniana forest in the Hetian Basin of County Changting, China

The global climate change has led to an increasing concern about the dynamics of the carbon storages of the forest ecosystem. In the past a few decades, remote sensing technology has been frequently applied in measuring forests'' carbon storage on various scales. Nevertheless, little work has been done in the estimation of the carbon storage of the masson''s pine, which is a widespread pine species in central and southern China. Therefore, this paper aims to develop a model based on remote sensing technology to estimate the carbon storage of Pinus massoniana forest using the case of the Hetian Basin in County Changting, Fujian province, southeastern China. We have carried out field measurements with 50 sampling sites in November 2010, in order to acquire basic data of Pinus massoniana forest in the study area. Each sampling site has a size of 20×20m to match the pixel size of remote sensing imagery. The filed-acquired data were correlated with the corresponding vegetation spectral information derived from a near-synchronized Advanced Land Observing Satellite (ALOS) image. To examine whether the image needs to be radiometrically corrected before it can be used for the task, the ALOS image was radiometrically corrected with the ICM and IACM models respectively. The difference between the two models lies in the latter corrects not only solar illumination and terrain effects but also atmospheric effects. Five vegetation indices were then derived from the ICM- and IACM-corrected images, as well as the original DN-based image. This is to determine which index would be most suitable for estimating the carbon storage of Pinus massoniana forest in the area. The five indices used are the Normalized Difference Vegetation Index (NDVI), the Difference Vegetation Index (DVI), the Perpendicular Vegetation Index (PVI), the Soil Adjusted Vegetation Index (SAVI), and the Soil Adjusted Ratio Vegetation Index (SARVI). By studying the agreement between the field-measured data and the data of the five selected vegetation indices derived from the ALOS image using regression analysis, the IACM-corrected NDVI data with an exponential regression model appeared to have the highest degree of agreement with the filed data and thus was utilized to calculate the carbon storage of Pinus massoniana forest in the Hetian Basin area. Accuracy assessment revealed that the model-estimated data were strongly correlated with field-measured data, suggested by a R² of 0.979, a root mean square error of 3.01t/hm², and a relative error of -1.95%. The estimated data show a slight underestimate by 2% when compared with the measured data. This suggests that the remote-sensing based model can be effectively used for estimating the carbon storage of the Pinus massoniana forest in the study area. Nevertheless, an atmosphere correction for the remote sensing image should be carried out before it can be put in use, because this study has confirmed that the IACM-corrected data, which has been radiometrically corrected for atmosphere effects, can significantly improve the precision of the estimated results. Based on the retrieved estimate model, the carbon storage of Pinus massoniana forest in the Hetian Basin in 2010 was revealed, which was 114.58×10⁴t in total, with a density of 34.92t/hm².