Abstract:

To ensure the sowing in spring, it is necessary to carry out the monitoring of ground temperature and the forecasting of frozen soil thawing depth. The meteorological observation data of frozen soil from Shihezi of Xinjiang were utilized and relevant coefficient and linear regression methods were applied to analyze the frozen soil characteristics and the change rule of ground temperature and frozen soil depth in the thawing process, aiming to establish a forecasting model of frozen soil thawing in spring. The results indicate that the stable frozen soil period of Shihezi lasts from the middle ten days of November to the last ten days of March of the next year. The maximum depth of frozen soil tends to become shallow year by year, with the inclination rate of -5.4 cm/10 a; the freezing date of frozen soil is postponed, and the inclination rate is 2.0 d/10 a; the unfreezing date of frozen soil becomes earlier, and the inclination rate is -1.5 d/10 a. The thawing period of frozen soil lasts from the middle ten days of March to the first ten days of April, and the thawing rate of frozen soil is between 3.1-4.0 cm/d. The ground temperature in spring has significantly positive correlations with average temperature, thawing depth of frozen soil and cumulative positive temperature, and the corresponding forecasting models are established. Historical back substitution accuracy rate of the forecasting model of 10 cm ground temperature is over 96%, and historical back substitution accuracy rate of thawing depth forecasting model of frozen soil is over 94%. The forecasting of ground temperature and soil thawing in spring could be carried out through the models.

Key words: frozen soil in Xinjiang, ground temperature during sowing period, characteristics of frozen soil, forecasting model of frozen soil thawing, thawing of frozen soil in spring