GPS signal-to-noise ratio snow depth inversion method assisted by support vector machine

Abstract

The invention discloses a GPS signal-to-noise ratio snow depth inversion method assisted by a support vector machine, and the method comprises the steps: firstly carrying out the preprocessing of GPSdata through a quadratic term fitting method, and obtaining the direct reflection signals of all GPS satellites in batch; analyzing the direct reflection signal to separate out satellite altitude angle information and signal-to-noise ratio data of 5-25 degrees, and performing trend term and unit conversion on signal-to-noise ratio elimination to obtain low altitude angle linear signal-to-noise ratio residual data; performing Lomb-Scargle spectrum analysis on the low altitude angle linear signal-to-noise ratio residual error data to calculate a spectrum frequency, solving a difference value between a vertical distance of a receiver and an antenna height, and performing formula conversion to obtain snow depth inversion values of a plurality of satellites; and finally, screening out snow depth inversion values of the two groups of satellites by integrating GPS satellite observation quality and conditions, performing supervised learning and training on the snow depth inversion results of the two groups of satellites screened out by utilizing a support vector machine, and outputting a snow depth prediction value after machine learning.

Description

technical field [0001] The invention relates to the field of computer machine learning and the use of a global positioning system (Global Positioning System, GPS), in particular to a support vector machine-assisted GPS signal-to-noise ratio snow depth inversion method. Background technique [0002] As a key storage element of the hydrological cycle, snow cover not only affects the climate of the global cryosphere, but also plays a crucial role in the surface energy balance and global climate change feedbacks. Therefore, research on changes in snow cover is important for assessing regional and global energy and water cycle changes. [0003] The current conventional snow detection method is mainly to measure snow depth through manual field inspection and sampling, while the remote sensing technology developed rapidly in recent years—Global Positioning System-Interferometric Reflectometry (GPS-IR) technology relies on high-precision, The advantages of low cost, low power consu...

AI Technical Summary

Problems solved by technology

[0007] The present invention proposes a GPS signal-to-noise ratio snow depth inversion method assisted by a support vector machine to solve the problem of being easily affected by the observation quality of a single GPS satellite and jumping and jumping in the current international use of GPS-IR single-satellite inversion of snow depth. The key technical problem that the inversion results greatly deviate from the measured snow depth caused by the deviation problem

Images