.png){kind=icon}
返回
A novel dynamic interpolation method based on both temporal and spatial correlations
.png){kind=icon}
PRE
AI
DOI：10.1007/s10489-022-03815-7.png)
.png){kind=icon}
原文求助.png){kind=icon}
分享.png){kind=icon}
收藏
摘要
En 中文
Missing data are very common in environmental monitoring activities. Traditional interpolation methods often view the problem from either spatial or temporal perspective and do not make good use of valuable related information, thereby leading to low prediction accuracy. Although many hybrid spatiotemporal interpolation methods combine temporal interpolation methods with spatial interpolation methods and show superiority over other single temporal or spatial interpolation methods, they do not finely treat each missing value in terms of its specific features. In this paper, two dynamic spatiotemporal interpolation (DST) methods, coarse-grained DST (CGDST) and fine-grained DST (FGDST) are proposed by using both temporal and spatial interpolation results. Different from other hybrid spatiotemporal interpolation methods, they create differences in the contribution of temporal and spatial interpolation results and assign them with different weights. Both CGDST and FGDST treat each missing value differently and fill it by considering the reliability of both temporal and spatial interpolation results in terms of the lengths of its column gap and row gap. CGDST treats each missing value in a continuous missing area equally and all missing values have the same lengths of column and row gaps. FGDST goes beyond CGDST and treats each missing value differently based on its temporal distance to the nearest real observed values in both forward and backward directions. To demonstrate their superiority, the experimental datasets were generated randomly to simulate missing values in real life. The results of the extensive experiments showed that FGDST improved by 2.44% similar to 8.21% in terms of symmetric mean absolute percentage error (SMAPE) compared with the best baseline method. Compared with CGDST, the improvements of FGDST range from 1.05% to 2.87% in terms of SMAPE. Moreover, their superiority is more obvious if there are more continuous missing values in the temporal dimension.
Keyword:
Spatiotemporal interpolation
Dynamic interpolation
Kriging
Air quality monitoring
Linear regression
期刊
IF:
3.5
论文数:
7.6K
被引数:
1.7W
机构
引用论文
分享收藏
Spatial interpolation of the extreme hourly precipitation at different return levels in the Haihe River basin.png)
JOURNAL OF HYDROLOGY
JOURNAL OF HYDROLOGYIF6.3
PREAI
分享收藏分享收藏
查看更多内容