Methodology Study Of Frozen Soil Monitoring And Simulation Based On Remote Sensing And Numerical Model

As an essential component of the cryosphere,frozen soil is of high sensitivity and poses strong feedback effect to climate changes.Under the background of global warming,spatial-temporal changes of seasonal frozen soil and permafrost have become an important issue closely related to regional environment and human activity as well as daily life.The acquisition of spatial-temporal frozen soil processes is the fundamental need for study of land surface hydrological processes,global climate change,reconstruction of historical climates,establishment of future climate scenarios and forecasting future environmental changes.The advancement of remote sensing technology and better understanding of frozen soil physics provide a valuable means to frozen soil and permafrost studies,which have fascilitate the great progresses of relevant studies on frozen soil dynamic monitoring and numerical model simulations in recent decades.However,due to some limitations,frozen soil dynamic monitoring and numerical model simulations are still in an exploratory stage,the large uncertainty involved in many issues yet remained as a big challenge to relevant studies.Further investigations are needed to improve the description and expression of frozen soil dynamics,and new technology for systematically monitoring frozen soil dynamics in cold regions is the key to break through some of the bottlenecks.For example,the types and structures of frozen soil are spatially and temporally varied,and these variations definitely pose different influence on regional hydro-thermal exchanges in froze soil process with different levels and modes.Till now,most of the frozen soil monitoring by remote sensing did not fully take the specialties of different types of frozen soil into account,and lacked of a systematic and unified technical system for frozen soil dynamic monitoring.Numerical model development on simulations of frozen soil hydro-thermal processes was mainly focused on model integration in basin scales with specific research object or task,the frozen soil water and heat transfer processes were simply parameterized or even lacking.Therefore,remotely sensed information with new technology in association with the distributed new generation of numerical models is urgently needed to fulfillment the requirement for the fine spatial and high temporal resolution simulations on frozen soil dynamics.In the context of the foregoing descroped background,present study was carried out mainly focused on the following research contents: 1)Discriminating the surface soil freeze and thaw status and monitoring the distributon changes of permafrost with the improved Dual-index Algorithm(DIA)by using passive microwave remote sensing technology;2)Establishing a distributed numerical model(FFIMS model)for simulation of frozen soil processes and dynamics based on water and heat transferring mechanism;3)Integrating the remotely sensed information on surface frozen soil status into the developed distributed numerical model(FFIMS model)and coupling the frozen soil process with the distributed watershed hydrological process model(ESSI-3 model).The technical method system was comprehensively applied in frozen soil dynamic and hydrological process studies in Northeast China.The main conclusions and innovative acheievements obtained in this study can be summarized as follows:(1)An improved DIA algorithm by taking soil water characteristic parameters into account on the basis of DIA algorithm was proposed to improve the discrimination accuracy on soil freeze and thaw status of the ground surface.The improved DIA algorithm was applied to discriminate surface soil freeze and thaw status in an experimental area over Northeastern China,the average accuracy of about91.6% was achieved on average,and spatial distribution of permafrost with different types in each year over the study period was classified with classification error less than 3% compared with the officially documented permafrost maps.Statistic analysis on the mapping results of permafrost of the study region suggested that the southern permafrost boundary in the study area has generally moved northward approximately25-75 km since 2000.Overall,the permafrost has been degradated over the study region obviously over the study period.(2)In terms of low spatial resolution of passive microwave remote sensing in discrimination of surface soil freeze/thaw status,a spatial downscaling method based on spectrum analysis by taking characteristics of spatial heterogeneity and scale effect of sensonal frozen soil into account was developed in present study.In the frequency domain of the remotely sensed microwave band image acquired on frozen days,PSD and spatial frequency are highly similar between original low and high spatial resolution images in the same area at same time.With this relationship the amplitude information of the unresolved high spatial resolution image can be approximately estimated from the original low spatial resolution image.The accuracy of frozen days estimated with the proposed method,compared with those estimated with original low spatial resolution RES and GWR image,was improved significantly.It is worthwhile to mention that this spectrum analysis based spatial downscaling method doesn't only consider the spatial heterogeneity,but also reflect the large-scale characteristics in the process of spatial downscaling.(3)A fully distributed frozen soil processes integrated modeling system(FFIMS model)based on coupled water-heat transferring theory was developed in final stage of this study.The simulations with the FFIMS model indicated that the model accurately described the temporal variations and spatial distribution of the hydro-thermal parameters of frozen soil over the study area.The improved remotely sensed surface frozen soil information by using the proposed monitoring techniques significantly improve the accuracy of the FFIMS model simulations,especially in the simulations of surface soil freezing or thawing alternative period.Simulation results by ESSI-3 model coupled with FFIMS suggested that hydro-thermal processes affect watershed hydrology almost throughout the overall process of water circulations.However,hydro-thermal process poses limitated effections on the total runoff of watershed hydrology.Neverthless,frozen soil process was found significantly predominate the watershed hydrology especially in seasonal thawing/frozen alternative period.For model simulation of this period,the average Nash efficiency coefficient can be increased from nearly 0 toabout 0.67,which implied that it is necessary to consider the influence of frozen soil process on hydrological processes in cold region watershed hydrological simulations.In summary,present study made a breakthrough in several key scientific issues or bottlenecks related to frozen soil studies,i.e.the improvement on the original DIA algorithm in discrimination of surface soil freeze/thraw status by remote sensing techniques;development of a spatial downscaling method based on spectrum analysis by taking characteristics of spatial heterogeneity and scale effect of sensonal frozen soil into account;development of a fully distributed frozen soil processes integrated modeling system(FFIMS model)based on coupled water-heat transferring theory and coupled it merically into a distributed hydrological processes model.The present study has enhanced research capability of reseachers in many relevant fields.