农业科学与技术:英文版
農業科學與技術:英文版
농업과학여기술:영문판
Agricultural Science & Technology
2011年
6期
926-930
,共5页
张学儒%张镱锂%刘林山%张继平%高俊刚
張學儒%張鐿鋰%劉林山%張繼平%高俊剛
장학유%장의리%류림산%장계평%고준강
亚高山针叶林%Logistic回归%GIS%青藏高原东部山区
亞高山針葉林%Logistic迴歸%GIS%青藏高原東部山區
아고산침협림%Logistic회귀%GIS%청장고원동부산구
Subalpine Coniferous Forests%Logistic Regression%GIS%East Mountain Area of Tibetan Plateau
以位长江上游的大渡河流域的壤塘县为研究区。基于TM遥感影像,采用面向对象与目视解译相结合方法,提取1989年和2009年两个时期亚高山针叶林的空间分布,并对两期数据叠加分析,查找出其演替发生的类型和空间位置,作为因变量样本。同时,以GIS技术为支撑,基于DEM和专题数据,提取高程、坡度、坡向正弦、坡向余弦、地表曲率、距居民点距离、距耕地距离和距道路距离8个自变量,使之与亚高山针叶林空间演替样本进行Logistic回归拟合,计算各自变量的贡献率和演替发生概率在空间上的变化特征。结果表明:亚高山针叶林向草甸演替的过程中,距居民点距离和距耕地距离两个变量贡献较大,距居民点距离减小1个单位,其转变为草甸的概率将增大1.15倍。在亚高山针叶林向落叶阔叶灌丛演替的过程中,坡向正弦和距居民点距离贡献较大,其发生概率随坡向朝北程度的增大而增大,增大1个单位,概率将增大1.2倍。对亚高山针叶林空间演替特征的定量分析可以为亚高山针叶林的保护与恢复提供科学依据。
以位長江上遊的大渡河流域的壤塘縣為研究區。基于TM遙感影像,採用麵嚮對象與目視解譯相結閤方法,提取1989年和2009年兩箇時期亞高山針葉林的空間分佈,併對兩期數據疊加分析,查找齣其縯替髮生的類型和空間位置,作為因變量樣本。同時,以GIS技術為支撐,基于DEM和專題數據,提取高程、坡度、坡嚮正絃、坡嚮餘絃、地錶麯率、距居民點距離、距耕地距離和距道路距離8箇自變量,使之與亞高山針葉林空間縯替樣本進行Logistic迴歸擬閤,計算各自變量的貢獻率和縯替髮生概率在空間上的變化特徵。結果錶明:亞高山針葉林嚮草甸縯替的過程中,距居民點距離和距耕地距離兩箇變量貢獻較大,距居民點距離減小1箇單位,其轉變為草甸的概率將增大1.15倍。在亞高山針葉林嚮落葉闊葉灌叢縯替的過程中,坡嚮正絃和距居民點距離貢獻較大,其髮生概率隨坡嚮朝北程度的增大而增大,增大1箇單位,概率將增大1.2倍。對亞高山針葉林空間縯替特徵的定量分析可以為亞高山針葉林的保護與恢複提供科學依據。
이위장강상유적대도하류역적양당현위연구구。기우TM요감영상,채용면향대상여목시해역상결합방법,제취1989년화2009년량개시기아고산침협림적공간분포,병대량기수거첩가분석,사조출기연체발생적류형화공간위치,작위인변량양본。동시,이GIS기술위지탱,기우DEM화전제수거,제취고정、파도、파향정현、파향여현、지표곡솔、거거민점거리、거경지거리화거도로거리8개자변량,사지여아고산침협림공간연체양본진행Logistic회귀의합,계산각자변량적공헌솔화연체발생개솔재공간상적변화특정。결과표명:아고산침협림향초전연체적과정중,거거민점거리화거경지거리량개변량공헌교대,거거민점거리감소1개단위,기전변위초전적개솔장증대1.15배。재아고산침협림향락협활협관총연체적과정중,파향정현화거거민점거리공헌교대,기발생개솔수파향조북정도적증대이증대,증대1개단위,개솔장증대1.2배。대아고산침협림공간연체특정적정량분석가이위아고산침협림적보호여회복제공과학의거。
The study area lies in the Dadu River drainage area in upstream Yangtze River.The spatial distribution of subalpine coniferous forests in 1989 and 2009 was extracted by means of a combined method of object orientation and visual interpretation,and then the overlaying analysis of these data was conducted.The type and spatial location of succession were discovered and served as the sample of dependant variable.Meanwhile,supported by GIS technology and based on DEM and thematic data,the eight variables including altitude,slope,sin and cosin of aspect,curvity of land surface,and distance to residential area,cultivated land and road were extracted served as the sample of spatial succession of subalpine coniferous forests to fit Logistic Regression,and then the contribution of each independent variable as well as the spatial property of the occurrence probability of succession was calculated.The results suggested that,during the succession of subalpine coniferous forests to meadow,the closer to the residential area and cultivated land,the greater the contribution to succession is.In particular,when the distance to the residential area decreases by one unit,the probability for its conversion to meadow will be increased by 1.15 times.During the succession of subalpine coniferous forests to deciduous-broadleaved shrubs,the sin of aspect and distance to residential area contribute more,and the probability of succession increases with increasing degree of northwardness,i.e.when the degree of northwardness increases by one unit,the probability will be increased by 1.2 times.The quantitative analysis of spatial succession property of subalpine coniferous forests will supply scientific basis to the protection and restoration of subalpine coniferous forests.