Discrimination Of Freezing And Thawing State Of Surface Soils In Northeast China Based On AMSR-E Data

Permafrost is an important part of the cryosphere and is extremely sensitive to climate change.Northeast China is the main high latitude permafrost region in China.In the past 40 years,the southern boundary of permafrost has continuously moved south,the thickness of the active layer has deepened,and the permafrost has been seriously degraded.The freeze-thaw process is an important parameter indicating the degradation of permafrost.The lag of the freezing period,the advance of the thawing period,and the freeze-thaw transformation rate have a significant impact on the ecological environment.Studies have shown that the freeze-thaw process affects soil water and heat changes,and the carbon and nitrogen cycle affects energy balance and terrestrial ecological processes.Surface soil freeze-thaw changes are important indicators of climate change,and it is important to accurately determine the surface freeze-thaw conditions.The research of surface freeze-thaw discrimination based on remote sensing data is rarely studied in Northeast China,and existing algorithms cannot be directly applied in Northeast China.In order to address this problem,this paper selects the dual-index algorithm and makes improvements.The main theories are as follows:(1)Surface temperature is an important indicator of the surface freeze-thaw discrimination algorithm.In this paper,passive microwave brightness temperature data(AMSR-E)of advanced microwave scanning radiometer is selected,and single-channel inversion method and multi-channel linearity based on emissivity estimation are selected respectively.Inversion of surface temperature in Northeast China by fitting method.Based on the original method,an algorithm improvement is proposed: the single channel inversion method is used to establish the emissivity estimation equations in the growing season and the non-growing season according to the vegetation growth cycle,and the inversion capabilities of each microwave channel cover type are investigated.Combine the most accurate inversion channels,and use the microwave polarization difference index(MPDI)as the emissivity parameter to add the multi-channel fitting equation.The results show that the obtained surface temperature is 100% usable after excluding water and ice and snow invalid pixels.The improved single-channel inversion method reduces the root mean square error from 3.58 to 4.6 to 2.0 to 3.1,and the error is less than 2K in 75% of the area.;The final root-mean-square error of the multi-channel fitting method is 2.6 ~ 3.5.It also has high accuracy and can only use the microwave brightness temperature data to obtain the surface temperature.(2)Snow cover is an important type of seasonal ground cover in Northeast China.In order to improve the applicability of the surface freeze-thaw discrimination algorithm in snow-covered areas,the relationship between microwave standard deviation and snow depth is established.The results show that the microwave standard deviation decreases with increasing snow depth when the snow depth is less than 11 cm,and increases with snow depth when the snow depth is greater than 11 cm.Based on previous research results,it is assumed that the surface of the area is frozen when the snow depth is greater than 16 cm.(3)Select 5.5K,270 K,and 16 cm as the microwave standard deviation,threshold values for surface temperature,and snow depth,respectively.In Matlab,the frozen area is marked as 1 and the unfrozen area is marked as 0.Using the measured surface temperature of 23 meteorological stations in the study area as the verification data,the accuracy rate of discrimination in a one-month data sample is greater than 87%.The results show that the improved surface freeze-thaw discrimination algorithm has higher accuracy and practicability,and achieves the expected research purpose.