Research advances in passive microwave remote sensing of surface freeze-thaw state北大核心CSCD

Soil freeze-thaw cycles have important effects on surface water and energy balance，and then affect vegetation growth，soil water content，carbon cycle and terrestrial ecosystem. Passive microwave plays an important role in monitoring global and regional surface freeze-thaw processes due to its high temporal resolution，abundant data and sensitivity to soil moisture. With the launch of passive microwave sensors at home and abroad，it provides conditions for the study of permafrost interannual variation，seasonal variation，diurnal variation and long time series of near-surface soil freeze-thaw cycle. In recent years，the study of surface freeze-thaw cycle using passive microwave data has gradually increased. Based on previous studies，this paper summarizes the types of passive microwave remote sensing data and the characteristics of the bands contained in them. Expounded the principle of passive microwave monitoring data used for freezing and thawing，focus on passive microwave data in five categories in the study of freezing and thawing monitoring algorithms，including double index algorithm，the decision tree algorithm，freeze-thaw discriminant algorithm，seasonal threshold algorithm and based on the freezing L-band relative factors discriminant algorithm threshold，and analysis of 5 kinds of algorithms are compared；The freeze-thaw products based on different algorithms and passive microwave data were combed. Finally，the problems and future research directions of passive microwave remote sensing in surface freeze-thaw applications are summarized. In the acquisition of passive microwave data，it is found that the passive microwave data is missing due to the physical characteristics of the sensor，the shape and orbit of the earth，and the low resolution of passive microwave data leads to the low precision of freeze-thaw discrimination. For the problem of missing passive microwave data，it is proposed to use the average value of passive microwave data before and after two days to fill the missing brightness temperature data，or establish statistical function to complement the missing data. For the problem of low passive microwave resolution，the current development trend is to scale down based on passive microwave data and combine with multiple data products，such as ground temperature and active microwave data，or perform probability discrimination on surface freezing-thawing state in pixels，so as to better describe surface freeze-thaw state. In terms of the algorithm for discriminating surface freezing-thawing，based on the problem that dual-index algorithm，decision tree algorithm，freezing-thawing discriminant algorithm and seasonal threshold algorithm cannot accurately distinguish snow and frozen soil，this paper proposes to adopt the method of data assimilation or start from the snow radiation and frozen soil dielectric model. Optimization of the algorithm for the snow covered surface can further improve the accuracy of freeze-thaw classification. Based on existing freeze-thaw products，Although SMAP freeze-thaw products continue to be updated，SAMP satellite was launched late，and SAMP freeze-thaw products have a short time series. In the future，the time span of this algorithm for freezing-thawing products can be extended by combining L-band data provided by SMOS satellite. The problems mentioned above and the direction of further research are of great significance for improving the accuracy of freezing and thawing discrimination and improving the understanding of the variation law of freezing and thawing cycles，and also have certain research space. © 2022 Science Press (China).

Research advances in passive microwave remote sensing of surface freeze-thaw state

Soil freeze-thaw cycles have important effects on surface water and energy balance， and then affect vegetation growth， soil water content， carbon cycle and terrestrial ecosystem. Passive microwave plays an important role in monitoring global and regional surface freeze-thaw processes due to its high temporal resolution， abundant data and sensitivity to soil moisture. With the launch of passive microwave sensors at home and abroad， it provides conditions for the study of permafrost interannual variation， seasonal variation， diurnal variation and long time series of near-surface soil freeze-thaw cycle. In recent years， the study of surface freeze-thaw cycle using passive microwave data has gradually increased. Based on previous studies， this paper summarizes the types of passive microwave remote sensing data and the characteristics of the bands contained in them. Expounded the principle of passive microwave monitoring data used for freezing and thawing， focus on passive microwave data in five categories in the study of freezing and thawing monitoring algorithms， including double index algorithm， the decision tree algorithm， freeze-thaw discriminant algorithm， seasonal threshold algorithm and based on the freezing L-band relative factors discriminant algorithm threshold， and analysis of 5 kinds of algorithms are compared； The freeze-thaw products based on different algorithms and passive microwave data were combed. Finally， the problems and future research directions of passive microwave remote sensing in surface freeze-thaw applications are summarized. In the acquisition of passive microwave data， it is found that the passive microwave data is missing due to the physical characteristics of the sensor， the shape and orbit of the earth， and the low resolution of passive microwave data leads to the low precision of freeze-thaw discrimination. For the problem of missing passive microwave data， it is proposed to use the average value of passive microwave data before and after two days to fill the missing brightness temperature data， or establish statistical function to complement the missing data. For the problem of low passive microwave resolution， the current development trend is to scale down based on passive microwave data and combine with multiple data products， such as ground temperature and active microwave data， or perform probability discrimination on surface freezing-thawing state in pixels， so as to better describe surface freeze-thaw state. In terms of the algorithm for discriminating surface freezing-thawing， based on the problem that dual-index algorithm， decision tree algorithm， freezing-thawing discriminant algorithm and seasonal threshold algorithm cannot accurately distinguish snow and frozen soil， this paper proposes to adopt the method of data assimilation or start from the snow radiation and frozen soil dielectric model. Optimization of the algorithm for the snow covered surface can further improve the accuracy of freeze-thaw classification. Based on existing freeze-thaw products， Although SMAP freeze-thaw products continue to be updated， SAMP satellite was launched late， and SAMP freeze-thaw products have a short time series. In the future， the time span of this algorithm for freezing-thawing products can be extended by combining L-band data provided by SMOS satellite. The problems mentioned above and the direction of further research are of great significance for improving the accuracy of freezing and thawing discrimination and improving the understanding of the variation law of freezing and thawing cycles， and also have certain research space.