摘要
以相关资料和文献中1950-2009年的数据为依据,利用相关分析、灰色预测方法和BP神经网络模型,定量分析了三江平原湿地面积变化的影响因素,并对未来20年三江平原的湿地面积进行预测。结果表明,耕地面积、年均气温、人口数量、径流深度、相对湿度、年降水量、政策因子与湿地面积变化关系密切,可以作为BP神经网络模型的预测变量。BP神经网络模型预测2010s三江平原的湿地面积为58.58万hm2,2020s为45.86万hm2,湿地面积在未来一段时间还有减少的趋势,但减少趋势明显减缓。
The global wetlands area has decreased during the recent decades.Wetlands changes and trend prediction become the research hotspot now.Based on the correlation analysis,Grey prediction method and BP neural networks,we did a quantitative analysis at the factors that affect the changing of Sanjiang Plain wetland and forecasted of area of wet land in the following 20 years.The result indicated that the changing of wetland had close relationship with cultivated area,temperature,population quantity,runoff depth,relative humidity,rainfall and policy,which could be used as the forecasting evidence of the BP neural network model.The BP neural network forecasted that the area of Sanjiang Plain wetland was 585,800 hm^2 in 2010s and will be 458,600 hm^2 in 2020s.The area showed the reducing tendency in the future and the tendency will slow down.
出处
《干旱区资源与环境》
CSSCI
CSCD
北大核心
2012年第10期53-56,共4页
Journal of Arid Land Resources and Environment
基金
国家自然科学基金项目(41071037)
吉林师范大学研究生创新科研计划项目(201111)资助
