摘要
以洪泽湖为例,研究多代测高卫星不同版本测高资料的数据编辑、地球物理改正、波形重定方法;利用误差改正后的数据计算洪泽湖2010年1月至2018年12月的年度、季度与月度水位异常时间序列,分析得到洪泽湖水位的变化规律及水位异常时间;利用BP神经网络模型结合气温、降水、卫星测高数据,预测洪泽湖2019年部分月份的水位,并与测高所得水位进行比较。研究结果表明,利用卫星测高对内陆湖泊水位变化进行监测具有一定的可行性,利用神经网络模型对内陆湖泊水位进行预测具有一定的准确性。
Taking Hongze Lake as an example,the method of data editing,geophysical correction,and waveform resetting of altimetry data from different versions of multi-generation altimetry satellites is studied.It calculates the annual,quarterly,and monthly water level anomaly time series of Hongze Lake from January 2010 to December 2018 using the data after error correction.It uses BP neural network model combined with temperature,precipitation,and satellite altimetry data to predict the water level of Hongze Lake in some months of 2019 and compares it with the water level obtained by altimetry.The research results show that it is feasible to monitor the water level of inland lakes by using satellite altimetry,and it is accurate to predict the water level of inland lakes by using neural network models.
作者
梁子亮
解琨
付超
王春
LIANG Ziliang;XIE Kun;FU Chao;WANG Chun(Provincial Geomatics Center of Jiangsu,Nanjing 210013,China)
出处
《测绘与空间地理信息》
2021年第8期172-175,共4页
Geomatics & Spatial Information Technology
关键词
卫星测高
洪泽湖
BP神经网络
水位预测
水位异常
satellite altimetry
Hongze Lake
BP neural network
water level prediction
water level abnormality