| As a basic parameter in the process of forming,transforming and consuming of surface water resources,soil moisture has a very important influence on vegetation growth,agricultural production,and healthy operation of regional ecosystems.The Aksu River Basin,which located in the arid and semi-arid regions of western China,has a dry year-round climate with few precipitation but high evaporation,Agricultural production relies almost entirely on pumping irrigation in rivers and underground wells.The"natural-social”dual nature of the surface water circulation process in The Aksu River Basin is particularly evident,and the spatial and temporal variation of soil moisture is complicated,which resulting to a inefficient use of water resources and the shortage of water resources there is very serious.In recent years,the intensity of land development and utilization in Akesu River Basin has increased significantly.Meanwhile,with the increasing of agricultural water demand,rivers and groundwater levels in the basin are continuously declining instead.Not only has agricultural production been affected,but also the oasis ecosystem is gradually deteriorating.So it is urgent that a wide range,rapid acquisition and dynamic monitoring of soil moisture in The Aksu River Basin should be carried out to explore the water and heat exchange process of "air-surface",make a guidence on agricultural production and promote the rational distribution and efficient use of water resources in the basin.It is also important for the restoration of desert vegetation and the protection of the ecological environment of The Aksu River Basin.Based on above,this article takes The Aksu River Basin as the research area,According to the new demand of modern agricultural information acquisition and remote sensing monitoring,two high-resolution remote sensing images GF-1 WFV and Landsat8 OLI combined with the measured data of soil moisture at different depths in the study area were used as data sources to construct the soil moisture inversion model with PDI,MPDI and VAPDI respectively.Then using the measured data to verify and evaluate the accuracy of each inversion model,And at the same time,uncertainty analysis of inversion model from various aspects was implemented.On this basis,an inversion model with a high precision index was selected to retrieve the spatial distribution map of surface soil moisture at sampling time in the study area by two kinds of remote sensing images.After a comparative analysis of the inversion results,four temporal GF-1 WFV remote sensing images were choosed to dynamically monitor and analyze the surface soil moisture in The Akesu River Basin finally.The conclusions obtained in this study are as follows:(1)Three kinds of remote sensing drought indices PDI,MPDI and VAPDI,based on GF-1 WFV and Landsat8 OLI remote sensing images,have a certain linear negative correlation with the measured values of soil moisture in 0-10cm,10-20cm and 20-30cm depth layers in the study area,It indicates that it is feasible to build the soil moisture retrieval model by using the drought index method of high resolution optical remote sensing.However,the correlation between drought indices and soil moisture of 0-10cm depth layer is the strongest,the average R2 is up to 0.67 and the accuracy of the inversion models are higher,which shows that drought index based on near infrared and red light reflectivity of optical remote sensing image is more sensitive to the information of soil moisture near the surface layer,but the accuracy of soil moisture retrieving in deep is slightly lower in the study area.(2)The accuracy of VAPDI and MPDI inversion models of GF-1 WFV and Landsat8 OLI remote sensing images is higher than that of PDI,from the comparison of R2,MAE,MRE and RMSE of each soil moisture inversion model,the VAPDI soil moisture inversion model has the highest precision;The uncertainty of soil moisture inversion model is mainly affected by the parameters and factors such as the number of modeling samples,sample heterogeneity,vegetation cover level and vegetation index,among which the vegetation cover level and the selection of vegetation index have the greatest influence on the model uncertainty.(3)The spatial distribution pattern of the surface soil moisture in the study area retrieved by VAPDI soil moisture inversion model of GF-1 WFV and Landsat8 OLI remote sensing image can directly reflect the difference of soil moisture level and difference between different areas in the watershed.It indicates that the use of high-resolution optical remote sensing image VAPDI for retrieving large scale surface soil moisture has some certain utilities.The overall pattern of spatial distribution of surface soil moisture in the study area is consistent based on two remote sensing images,But the results retrived by GF-1 WFV remote sensing image with higher spatial resolution are more fine than that of Landsat8 OLI,and the spatial distribution of soil moisture is more hierarchical,which can reflect the soil moisture of similar plots.It is proved that the comprehensive inversion effect of GF-1 WFV remote sensing image is better than that of Landsat8 OLI.(4)The dynamic monitoring results derived from VAPDI soil moisture retrieval model basaed on four temporal GF-1 WFV remote sensing images show that the temporal and spatial variation of surface soil moisture in the study area was obvious,but the overall spatial pattern was relatively stable from June to August 2016.The middle and upper reaches of Wushi,Wensu,southern Awati,northern Akesu,Eastern Akesu and alar city have high and stable soil moisture;But the soil moisture in the northern Wushi County,Wensu county and the eastern Akesu City,the western and southern Awati County,the southern,the Ke Ping county and the Shah Ya county are low,and the temporal and spatial changes are relatively strong.(5)The spatial distribution of surface soil moisture in Akesu River Basin is mainly influenced by artificial irrigation,land use types and vegetation coverage.So the agricultural water requirement can be improved by using drip irrigation to improve the efficiency of water resources utilization in the upper reaches of farmland;In non agricultural production areas with strong temporal and spatial variation in surface soil moisture on the edge and lower reaches of the basin,the cultivation and maintenance of artificial shelterbelts and ecologic forests should be strengthened to prevent further land desertification and the degradation of natural woodland and grassland,so as to ensure the sustainable development of agriculture and the health operation of the oasis ecosystem in the Akesu River Basin. |
PDF Full Text Request