Method for repairing missing of underground water level monitoring value

Abstract

The invention discloses a method for repairing missing of underground water level monitoring values, which is based on an underground water level space-time sequence characteristic analysis result, carries out research around a space-time missing value hybrid repairing algorithm principle, and abstracts a missing repairing hybrid model considering space-time correlation. By analyzing the characteristics of a geostatistics method and a machine learning algorithm, a ubiquitous kriging spatial interpolation method and a time missing repair algorithm suitable for underground water level monitoringdata are selected, and according to a space-time element extension theory, the single space and time missing repairing algorithm is fused to construct the space-time missing data repairing hybrid model.

Description

technical field [0001] The invention relates to the technique of groundwater level monitoring, in particular to a repair method for missing water level monitoring values. Background technique [0002] Water is the source of life and the basis for the survival and development of animals and plants. For a long time, as an important water supply resource, groundwater has been plagued by problems such as overexploitation and pollution, resulting in frequent occurrence of a series of environmental geological problems such as groundwater level drop, land subsidence, and freshwater salinization. Long-term monitoring of groundwater level is an important basis for understanding the status quo of groundwater development and utilization and planning for sustainable development of water resources. As a typical time-space sequence data, groundwater monitoring water level data has the characteristics of nonlinearity, instability, and strong time-space correlation. It is the basis for gro...

AI Technical Summary

Problems solved by technology

However, the existing traditional restoration methods for missing groundwater level data are poor in efficiency and accuracy. Restoration methods based on spatio-temporal geostatistics and machine learning often only focus on time or spatial elements when repairing spatio-temporal sequence data, and have no effect on spatio-temporal correlation. The groundwater level data has not only obvious time series characteristics, but also regular spatial characteristics. The water level data changes with time and location, which is a typical time-space series data.

The existence of spatiotemporal missing data directly leads to the reduction of the spatiotemporal information of the research samples, reduces the integrity and reliability of the data set, and seriously affects the scientific rationality of people's analysis and modeling of the spatiotemporal dynamic characteristics of the research object.

Images