The Dynamic Remote Sensing Monitoring Of Vegetation Eco-Water In Jiuzhaigou

In the terrestrial ecosystem,vegetation plays a crucial role.The growth and development of vegetation are restricted by terrestrial hydrological environment,while the growth of vegetation,in turn,plays an important regulating role in hydrological environment.They complement and restrict each other.Eco-Water is the most important water storage body in the ecological environment,which can regulate,cache and allocate rainfall.The study of Eco-Water is a crucial part of quantitative research of water resources.The storage of vegetation Eco-Water is relatively particular,so it is difficult to quantify and extract storage quantity using conventional surveying and mapping methods.With the development of remote sensing technology,as a powerful technical support,the quantitative research of remote sensing information will provide a new method and idea for the quantitation research of Eco-Water.At present,domestic and foreign quantitative remote sensing studies are dominated by precision agriculture and ecological environment monitoring,etc.The concept of Eco-Water was coined by our team.The existing studies on Eco-Water are contributed by our team exclusively,too.After years of exploration and investigation by our team members,great progress has been achieved.Most of the studies on vegetation Eco-Water content carried out by our project team in the early stage were based on the measured spectral data of vegetation.Most of the spectral collections were about leaves,while the aboveground part of vegetation was not quantified as a whole.As plants mainly consist of dry matters and water,in the present study,the vegetation Eco-Water content above the sample plot was calculated using biomass and measured water content.According to the vegetation Eco-Water content above the sample plot and the area of sample plot,the vegetation Eco-Water content per unit area was obtained.Finally,according to the relationship between the vegetation Eco-Water content per unit area and the vegetation index,a model was set up and the vegetation Eco-Water content was inverted.There are different types and materials of vegetation,so the water contents vary.In different growing seasons,the vegetation Eco-Water contents differ.The Eco-Water content of vegetation will be affected by the characteristics and life cycle of vegetation and seasonal changes.So far,there is no quantitative research on vegetation Eco-Water in different months,so it is of great significance to perform a dynamic monitoring on vegetation Eco-Water and probe into the temporal and spatial changes of vegetation Eco-Water.This paper selected remote sensing images on April9,2018,July 14,2018,November 16,2017 and January 22,2019 in order to carry out a quantitative remote sensing inversion of vegetation Eco-Water contents in these four months.The main research contents and findings of this paper are as follows:（1）The vegetation type information in the researched area was extracted using Maximum likelihood methodThe author extracted vegetation type information in the researched area,according to the features,texture structures,tones and other auxiliary information of ground objects in the remote sensing images,combined with the results of field survey and verification.Using three classification methods,namely,the maximum likelihood method,support vector machine method and neural network method,the vegetation was classified.Among them,the maximum likelihood method had the highest overall accuracy（90.5401%）and the kappa coefficient was 0.8853.The classification results using maximum likelihood method were selected as extraction results of vegetation types in the researched area.The researched area can be divided into coniferous forest,broad-leaved forest,mixed forests,shrubs,meadows,waters,bare land and others.While other types of vegetation mainly included residential land and roads.The extraction results showed that broad-leaved forest existed at an altitude of2000-3500m,coniferous forest at 2000-2500m and mixed forests at 2500-2800m.The vegetation 3500m above the researched area was dominated by alpine meadows.The slopes were abrupt,the land was barren and the vegetation was stunted.（2）A regression model was set up according to the relationship between vegetation Eco-Water content per unit area and the vegetation index MSI 7The present study was a quantitative study of vegetation Eco-Water content above the researched area,and the vegetation Eco-Water content aboveground was calculated by biomass.Based on the previous findings achieved by our team,a remote sensing inversion model was set up based on the relationship between the vegetation index and the vegetation Eco-Water content per unit area.Compared with other models established based on the relationship between vegetation Eco-Water content and different vegetation indexes,the regression model established with MSI 7（Moisture Stress Index）had the best inversion effect.Using this model,the author successfully inverted the vegetation Eco-Water content in the research area.（3）A dynamic monitoring of vegetation Eco-Water contents in different monthsVegetation Eco-Water content changes with season and growth status.In different seasons,the contents vary.In this paper,the author inverted the vegetation Eco-Water contents at different times using remote sensing images in four different months,and finished a multi-temporal quantitative study on vegetation Eco-Water.The inversion results showed that the vegetation Eco-Water content order in these four months was July>April>November>January.The average Eco-Water contents were315.66 Mg ha^-1,173.19 Mg ha^-1,157.78 Mg ha^-1 and 144.37 Mg ha^-1.The storage quantities of vegetation Eco-Water were 15417131.28 Mg,8727004.27 Mg,14048252.19 Mg and 12881028.00 Mg respectively.（4）A classified statistics of the storage quantity of vegetation Eco-WaterAccording to the extraction results of vegetation types in the researched area,as well as the remote sensing inversion results of vegetation Eco-Water,the author counted the storage quantities of vegetation Eco-Water by vegetation type.The storage quantity of Eco-Water in broad-leaved forest accounted for the largest proportion in the total storage quantity of Eco-Water,followed by mixed forests,shrubs and coniferous forest.The same vegetation type may have different storage quantities of Eco-Water in different months.These different storage quantities accounted for different proportions in the total storage quantity.