摘要
基于南四湖近30 a来多源遥感影像、土地利用、环境监测和实地调查数据,选择1982年、1992年、2002年和2012年4个时间断面,从湿地组织结构、整体功能和社会经济3个方面选取指标,从无警、轻警、中警、重警和巨警5个级别划分警度,采用多级模糊综合评价法进行警度评价,并运用BP神经网络模型对未来10 a警度发展进行预测,研究结果表明:1南四湖湿地生态环境现在处于重警状态,农业化肥、农药污染和工业污染物排放是主要的警报来源;2近30 a来南四湖湿地整体生态环境处于不断恶化趋势,社会经济指标恶化趋势明显,未来10 a各项指标和整体预警度均呈下降趋势,整体水平达到中警状态,社会经济投入的增加是重要影响因子;3近10 a环境恶化速度比前20 a有所减缓,社会经济指标恶化速度较快,人类活动的负向干扰仍大于正向干扰。
Currently, as most lake areas in China are under high interference of intensifying human activities,lake ecosystem has been severely damaged and causing serious deterioration of human habitat.. As lake wetland eco-environmental early warning is the key to achieve lake monitoring, environmental management and ecological restoration and it is the further work which is based on ecosystem vulnerability and health assessment as well, a reseach of eco-environmental early warning on Nansihu Lake, China is done in this article. In this study, remote sensing date, land use and environment monitor date in 1982, 1992, 2002 and 2012 were chosen as date base. Then, an index system combine of organization structure, whole function and social economic environment was established. Every index was divided in five grades which include non-alert, light alert, middle alert, heavy alert and giant alert. The fuzzy synthetically judgment model was used as assessment.To predict the state of eco-environmental early warning in 2022, the back propagation algorithm neural network mothod is used. The model＇s learning samples are results statistics in 1982, 1992 and 2002 and the testing sample is statistics in 2012. The testing results show that the accuracy meets the requirement and it can be used to make prediction of the early warning state in 2022. The conclusions were as follows： 1） As chemical fertilizer and pesticide issue in agriculture activities and industrial emission were over the environmental capacity. Nansihu wetland was in heavy alert state in 2012. In each under grades indicators system, organization structure indicators system was in middle alert state, whole function indicators system was in heavy alert state and social economic indicators system was in huge alert state; 2） As the little natural precipitation in 2002, the whole wetland has been deteriorating while the social economic environment indexes were getting even worse during the 1980s-2000 s. The overall early warning state achieved huge alert state in 2002. In the coming 10 years, either indexes or the whole warning state is getting better because of the increasing investment in ecological restoration. 3） The degradation rate of nearly 10 years was slower than it of earlier 20 years while the social economic environment indexes always keep a high speed, which can be explained by the negative disturb form human beings activities. This research is significant to the environment management of Nansihu wetland and the result can provide scientific reference for lake management and regional sustainable development.
出处
《地理科学》
CSCD
北大核心
2016年第8期1219-1226,共8页
Scientia Geographica Sinica
基金
国家科技部支撑计划(2012BAC04B00)
国家自然科学基金项目(41671158)
教育部博士点基金项目(20132136110001)
辽宁省优秀人才支持计划(LR2013050)共同资助~~
关键词
预警
湖泊湿地
生态环境
模糊综合评价
南四湖
warning
lake wetland
eco-environment
fuzzy synthetically judgment model
Nansihu Lake
