攸县森林生物量动态变化分析研究
发布时间：2018-11-23 13:12
【摘要】：对于森林生态系统的研究,森林生物量是森林生态系统的重要标志。森林生物量同时可以预示森林固碳能力的强弱,为评估区域森林碳平衡提供重要参数。国内外研究表明,森林生物量估算方法还停留在传统统计方法上,森林生物量空间分布与制图等方面的研究还不足,难以从空间上对森林生物量进行分析和评价,而森林生物量图可以从空间上直接估计森林生物量及来自土地利用变化的净通量,因此,这方面的研究具有重要意义。本研究以攸县为研究区,采用1999年、2004年、2009年、2014年四期Landsat遥感影像及1999年、2004年、2009年、2014年四期固定样地数据,结合三种回归模型,估算森林生物量。主要研究结果如下:(1)确定了攸县森林生物量反演的敏感因子,结合相关性分析结果,采用逐步剔除法进行筛选,最终从14个初始变量中保留7个用于森林生物量的反演,其中RGVI植被指数与森林生物量的相关性最高。(2)建立了基于Landsat影像森林生物量的最佳回归模型。采用逐步回归、logistic回归和空间地理加权回归分别构建森林生物量的遥感反演模型,结果表明空间地理加权回归模型效果较好。(3)根据空间地理加权回归模型估算了 1999年、2004年、2009年和2014年四个不同时期的攸县森林生物量。并结合GIS,对四个不同时期的攸县森林生物量进行空间统计、分析和制图。(4)研究显示,整个攸县森林生物量1-10(t/ha)级的区域面积逐年上升,在2009年达到最大,意味着其它等级森林生物量面积比例逐年下降,2014年时10-20(t/ha)级面积比例下降,说明2014年期间绿化水平略微上升。2004年时30-40(t/ha)级森林生物量下降明显,基本变化为1-10级和10-20(t/ha)级。而40-50(t/ha)级森林生物量面积比例和大于50(t/ha)级的面积比例于2004年至2014年期间增长明显,说明退耕还林政策在2004年至2014年间执行力度较大,取得了较好的森林保护效果。
[Abstract]:For the study of forest ecosystem, forest biomass is an important symbol of forest ecosystem. Forest biomass can also predict the strength of forest carbon sequestration ability and provide important parameters for assessing regional forest carbon balance. Studies at home and abroad show that the estimation method of forest biomass is still in the traditional statistical method, and the spatial distribution and mapping of forest biomass are still insufficient, so it is difficult to analyze and evaluate forest biomass in space. Forest biomass map can directly estimate forest biomass and net fluxes from land use change, so this study is of great significance. In this study, you County was used to estimate forest biomass by using four Landsat remote sensing images in 1999, 2004, 2009, 2014 and four fixed plots in 1999, 2004, 2009 and 2014, combined with three regression models. The main results are as follows: (1) the sensitive factors of forest biomass inversion in Youxian were determined. Combined with the results of correlation analysis, the stepwise culling method was used to screen, and 7 of the 14 initial variables were retained for forest biomass inversion. The correlation between RGVI vegetation index and forest biomass was the highest. (2) the best regression model of forest biomass based on Landsat image was established. The remote sensing inversion models of forest biomass were constructed by stepwise regression, logistic regression and spatial geographical weighted regression, respectively. The results show that the spatial geographical weighted regression model is effective. (3) according to the spatial geographical weighted regression model, 1999 was estimated. In 2004, 2009 and 2014, you County forest biomass in four different periods. The spatial statistics, analysis and mapping of forest biomass in Youxian in four different periods were carried out with GIS,. (4) the study showed that the area of forest biomass of 1-10 (t/ha) in Youxian increased year by year, and reached the maximum in 2009. This means that the proportion of forest biomass area in other grades has been decreasing year by year, and the proportion of forest biomass in class 10-20 (t/ha) has decreased in 2014, indicating that the green level increased slightly during 2014. In 2004, the forest biomass of grade 30-40 (t/ha) decreased obviously. The basic changes were 1-10 and 10-20 (t/ha). However, the proportion of forest biomass area of 40-50 (t/ha) class and the proportion of area larger than 50 (t/ha) level increased significantly from 2004 to 2014, which indicated that the policy of returning farmland to forest was carried out strongly from 2004 to 2014. Good effect of forest protection has been achieved.
【学位授予单位】：中南林业科技大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：S718.5
，
本文编号：2351736
[Abstract]:For the study of forest ecosystem, forest biomass is an important symbol of forest ecosystem. Forest biomass can also predict the strength of forest carbon sequestration ability and provide important parameters for assessing regional forest carbon balance. Studies at home and abroad show that the estimation method of forest biomass is still in the traditional statistical method, and the spatial distribution and mapping of forest biomass are still insufficient, so it is difficult to analyze and evaluate forest biomass in space. Forest biomass map can directly estimate forest biomass and net fluxes from land use change, so this study is of great significance. In this study, you County was used to estimate forest biomass by using four Landsat remote sensing images in 1999, 2004, 2009, 2014 and four fixed plots in 1999, 2004, 2009 and 2014, combined with three regression models. The main results are as follows: (1) the sensitive factors of forest biomass inversion in Youxian were determined. Combined with the results of correlation analysis, the stepwise culling method was used to screen, and 7 of the 14 initial variables were retained for forest biomass inversion. The correlation between RGVI vegetation index and forest biomass was the highest. (2) the best regression model of forest biomass based on Landsat image was established. The remote sensing inversion models of forest biomass were constructed by stepwise regression, logistic regression and spatial geographical weighted regression, respectively. The results show that the spatial geographical weighted regression model is effective. (3) according to the spatial geographical weighted regression model, 1999 was estimated. In 2004, 2009 and 2014, you County forest biomass in four different periods. The spatial statistics, analysis and mapping of forest biomass in Youxian in four different periods were carried out with GIS,. (4) the study showed that the area of forest biomass of 1-10 (t/ha) in Youxian increased year by year, and reached the maximum in 2009. This means that the proportion of forest biomass area in other grades has been decreasing year by year, and the proportion of forest biomass in class 10-20 (t/ha) has decreased in 2014, indicating that the green level increased slightly during 2014. In 2004, the forest biomass of grade 30-40 (t/ha) decreased obviously. The basic changes were 1-10 and 10-20 (t/ha). However, the proportion of forest biomass area of 40-50 (t/ha) class and the proportion of area larger than 50 (t/ha) level increased significantly from 2004 to 2014, which indicated that the policy of returning farmland to forest was carried out strongly from 2004 to 2014. Good effect of forest protection has been achieved.
【学位授予单位】：中南林业科技大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：S718.5
，
本文编号：2351736
本文链接：https://www.wllwen.com/shoufeilunwen/zaizhiyanjiusheng/2351736.html
