Research Of Frozen Soil Development Process In Luogu River Region Based On In Situ Monitoring

Permafrost development is a unique permafrost hydrological phenomenon in the cold region.Against the background of severe permafrost degradation due to climate change,its development mechanism and the influencing factors of the change process need to be studied more deeply and effectively.This paper selects the Luogu River in the cold region as the study area to investigate the mechanism of permafrost development process and its influencing factors in this region,and analyzes the climatic characteristics of the Luogu River region through field trekking and theoretical research;based on the ground temperature data obtained from in situ monitoring,uses methods such as drawing development process lines,comparative analysis and correlation analysis to analyze the characteristics of permafrost development in the Luogu River region and its correlation with influencing factors;and through genetic algorithm optimized artificial neural network model to predict the freezing depth during the freezing period and explore the relationship between the freezing depth and air temperature and shallow ground temperature.The main findings of this paper are as follows:(1)The characteristics of climate change in the Logu River region are consistent with the trend of regional permafrost development.The temperature,precipitation and evaporation in the Luogu River region have shown an increasing trend in the past 40 years,with an annual average temperature of-1.66 ℃,a multi-year average precipitation of 641.17 mm and a multi-year average total evaporation of 486.78 mm.(2)The permafrost development process of Luogu River shows the characteristics of one-way freezing and two-way melting.The whole development process is divided into freezing period,thawing period and freeze-free period.The freezing period is generally from the middle and late October to the middle and early April of the following year,the thawing period is from the beginning of April to the early August of each year,and the rest of the time is freeze-free period.The freezing time is about 300 days and the maximum freezing depth is about 290 cm.(3)The correlations of ground temperature and freezing depth to air temperature were significantly correlated in both the freezing and thawing periods of the Logu River from 2018 to 2021,and the response relationship was obvious.Among them,the reason for the difference in the response relationship between ground temperature and freezing depth on air temperature in the melting period is mainly due to the influence of snow and other cover,the monitoring site located on the Heilongjiang River,moisture migration and river flow.(4)The overall prediction effect of BP neural network prediction freezing depth model optimized by genetic algorithm is good.the maximum absolute error of GA-BP neural network prediction model is 5.93 cm and the absolute mean error is 2.55 cm.the prediction model has high accuracy and can effectively solve the complex nonlinear relationship between the freezing depth of permafrost and its influencing factors in high latitude cold areas,which can provide reference for permafrost area It can provide reference for engineering design and construction.