發(fā)布時(shí)間：2018-11-23 13:12

【摘要】：對(duì)于森林生態(tài)系統(tǒng)的研究,森林生物量是森林生態(tài)系統(tǒng)的重要標(biāo)志。森林生物量同時(shí)可以預(yù)示森林固碳能力的強(qiáng)弱,為評(píng)估區(qū)域森林碳平衡提供重要參數(shù)。國(guó)內(nèi)外研究表明,森林生物量估算方法還停留在傳統(tǒng)統(tǒng)計(jì)方法上,森林生物量空間分布與制圖等方面的研究還不足,難以從空間上對(duì)森林生物量進(jìn)行分析和評(píng)價(jià),而森林生物量圖可以從空間上直接估計(jì)森林生物量及來(lái)自土地利用變化的凈通量,因此,這方面的研究具有重要意義。本研究以攸縣為研究區(qū),采用1999年、2004年、2009年、2014年四期Landsat遙感影像及1999年、2004年、2009年、2014年四期固定樣地?cái)?shù)據(jù),結(jié)合三種回歸模型,估算森林生物量。主要研究結(jié)果如下:(1)確定了攸縣森林生物量反演的敏感因子,結(jié)合相關(guān)性分析結(jié)果,采用逐步剔除法進(jìn)行篩選,最終從14個(gè)初始變量中保留7個(gè)用于森林生物量的反演,其中RGVI植被指數(shù)與森林生物量的相關(guān)性最高。(2)建立了基于Landsat影像森林生物量的最佳回歸模型。采用逐步回歸、logistic回歸和空間地理加權(quán)回歸分別構(gòu)建森林生物量的遙感反演模型,結(jié)果表明空間地理加權(quán)回歸模型效果較好。(3)根據(jù)空間地理加權(quán)回歸模型估算了 1999年、2004年、2009年和2014年四個(gè)不同時(shí)期的攸縣森林生物量。并結(jié)合GIS,對(duì)四個(gè)不同時(shí)期的攸縣森林生物量進(jìn)行空間統(tǒng)計(jì)、分析和制圖。(4)研究顯示,整個(gè)攸縣森林生物量1-10(t/ha)級(jí)的區(qū)域面積逐年上升,在2009年達(dá)到最大,意味著其它等級(jí)森林生物量面積比例逐年下降,2014年時(shí)10-20(t/ha)級(jí)面積比例下降,說(shuō)明2014年期間綠化水平略微上升。2004年時(shí)30-40(t/ha)級(jí)森林生物量下降明顯,基本變化為1-10級(jí)和10-20(t/ha)級(jí)。而40-50(t/ha)級(jí)森林生物量面積比例和大于50(t/ha)級(jí)的面積比例于2004年至2014年期間增長(zhǎng)明顯,說(shuō)明退耕還林政策在2004年至2014年間執(zhí)行力度較大,取得了較好的森林保護(hù)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.
【學(xué)位授予單位】：中南林業(yè)科技大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2017
【分類號(hào)】：S718.5
，

本文編號(hào)：2351737