东北地区湿地变化影响因素分析  

Analysis of Wetland Dynamics and Its Influencing Factors in Northeast China

作　　者：崔瀚文[1] 姜琦刚[1] 程彬[2] 林楠[3] 

机构地区：[1]吉林大学地球探测科学与技术学院,吉林长春130026 [2]长春师范学院城市与环境科学学院,吉林长春130032 [3]吉林建筑工程学院测绘与勘查工程学院,吉林长春130118

出　　处：《应用基础与工程科学学报》2013年第2期214-223,共10页Journal of Basic Science and Engineering

基　　金：中国地质调查局项目(1212010510613)

摘　　要：利用1975、2000、2007年遥感影像作为数据源,选用新型多元统计分析—偏最小二乘回归分析(PLS),研究东北地区湿地变化的影响因素.结果显示,湿地总面积在1975—2000年减少23079.1km2,在2000—2007年增加14523.3km2,但32年来依旧呈减少趋势.沼泽湿地景观梯度值域收缩,分布稀疏,面积减少,消退形势严峻,水稻田则反之.在景观梯度存在耦合性的三江平原(鹤岗—佳木斯—双鸭山—鸡西一线的东北部)、哈尔滨市东北部、松辽平原北部,沼泽湿地向水稻田转化剧烈,说明开垦沼泽湿地是水稻田增加的主要原因之一.偏最小二乘回归分析表明,粮食需求是造成沼泽湿地减少与人工湿地增加的主要原因,影响东北地区湿地变化的主要因素由自然因素向人为因素转变.除2000—2007年的人工湿地外,偏最小二乘回归分析的精度均较高.Based on remote sensing imageries in 1975,2000 and 2007, partial least-squares regression （PLS）, which is a new statistical analysis method, was selected to study the influencing factors of wetland dynamics. The results showed that the total area decreased 23079. 1km^2 in 1975- 2000 ,and increased 14523.3km2 in 2000--2007, whereas, over 32 years, it still decreased on the whole. Through the smaller mire gradient range, sparse distribution and area reduction, it can be seen that mire decreased severely. The paddy field change is on the contrary. Some districts have strong coupling between mire and paddy field landscape gradient. They are mainly distributed in Sanjiang Plain（ in the northeast of line Hegang City, Jiamusi City, Shuangyashan City, Jixi City）, northeastern Haerbin City, northern Songliao Plain, where most of the mire transfers to paddy field. It turns out that mire reclaim is the major reason why paddy field increases. PLS analysis reveals that food demand is the chief reason for decreased mire and increased human-man wetland. The principle influencing factor transfers from natural one to anthropogenic one. The prediction accuracy of PLS is high except human-made wetland in 2000-2007.

关 键 词：湿地 遥感 偏最小二乘回归(PLS) 影响因素 

分 类 号：TP79[自动化与计算机技术—检测技术与自动化装置] X87[自动化与计算机技术—控制科学与工程]