Temporal And Spatial Changes And Prediction Analysis Of Snow In Tianshan Mountain Area Of Xinjiang

Tianshan Mountain is one of the largest mountain systems in Asia.The east and west of Tianshan Mountain in China are long,and the north and south are wide.Tianshan Mountain is the mountainous area with the most snowfall in the three major mountain systems of Xinjiang and a major area of land snow cover in China.Most of the river recharge in Xinjiang comes from the snow cover of Tianshan Mountain.How to monitor the snow cover of Tianshan Mountain quickly and effectively is an important key part to study the water resources of Xinjiang.Based on MOD10A2 snow products,meteorological data and ecosystem data,snow cover method was used to analyze snow cover products in Tianshan Mountain area.Based on multivariate statistical analysis and support vector machine(SVM)model,the prediction and error analysis of snow cover in Tianshan Mountain area are concluded as follows:(1)The annual distribution of snow cover in 2000-2018 showed a U change.From January to July,the snow cover showed a downward trend,reaching a minimum(4.41%)in late July and early August,with an area of about 12679.09 km2..Snow accumulated from the beginning of September,increasing snow cover,to January reached the maximum(51.88%),the area reached about 164505.05 km2..The snow cover area>700 times in2000-2018 accounts for about 3.53%of the total mountain area,between 500?700 times,that is,between 25 and 35 times a year,accounting for 29.43%of the total area of Tianshan Mountain area,between 200-400 times.accounting for 64.34%of the total area of Tianshan Mountain area.(2)Different years of snow cover ratio of the minimum value of the date is not consistent,snow extreme value of the occurrence of occurrence of time shift,and then show that the Xinjiang Tianshan Mountains snow melting date is delayed.Bayinbrook has the highest elevation,the lowest temperature temperature reaches-20?,and the highest is about-15?.(3)The meteorological factors of different stations vary,the temperature in Hami area is higher than the rest of the stations and the average pressure of Bayinbrook is the lowest.The highest fitting coefficients between snow cover and temperature and sunshine hours from the linear model are:R~2=0.93 and R~2=0.68.The correlation degree between meteorological factors and snow cover in 6 meteorological stations is different,and the correlation between stations is different,which is mainly affected by geographical location and altitude.From the change of ecosystem area in Tianshan Mountain area in 2000-2005,it can be seen that the farmland ecosystem increased by 0.1%,covering an area of about300 square kilometers.Grassland ecosystems decreased by 0.14 per cent in 2000-2005,covering an area of about 450 km~2;other ecosystems increased by about 200 km~2in2000-2005.(4)Based on support vector machine(SVR)for snow prediction based on different meteorological stations,it can be seen that the results of SVR model simulation of six meteorological stations are similar,the correlation coefficients are all greater than 0.95,and the root mean square error is between 0.03-0.05.According to the time path analysis of forecast results and measured data,the cycle characteristics of annual changes are shown in the whole,the winter and summer alternations are obvious in the year,and the snow cover in the whole year shows a steady trend.The errors from the timing analysis of6 meteorological stations are mainly distributed in winter and summer extreme weather.As a whole,the standard deviation between the training value of the SVR model of six meteorological stations and the residual error of the simulation of the verification value is less than 0.1,which indicates that the prediction ability of the whole training set and the verification set is stronger and more accurate.based on the design experiments to remove temperature,sunshine hours,precipitation,wind speed,relative humidity,air pressure in turn,the correlation coefficients obtained are 0.94,0.91,0.82,0.77 and 0.32 in turn.It shows that snow cover is the result of the combined influence of meteorological factors,and the influence of single meteorological factors is small.