Research On Inversion Model Of Fractional Vegetation Cover And It's Dynamic Change In Yanqing District

The paper firstly used the TM image and actual sample data from Yanqing District in 2004 as the data source,and used the pixel bisection model method,the improved linear spectral hybrid model and the neural network model to extract the fractional vegetation cover,compared and analyzed the accuracy of the three models.Then used the model with the highest precision among the three models to calculate the fractional vegetation cover in the three phases of 2004,2009 and 2014 in Yanqing District,and analyzed the Yanqing District from the perspective of the overall,regional,elevation gradient,slope gradient and aspect.Dynamic change of fractional vegetation cover.Finally,a driving force factor database was established to drive the driver's spatialization.The driving force was quantitatively analyzed using GIS and Logistic model.The research results of this paper can provide scientific basis for the land use and ecological environment construction in the study area.At the same time,it enriches the correlation method between the dynamic change of fractional vegetation cover and driving force factors.The main research results of the paper are as follows:(1)Using three fractional vegetation cover inversion models to calculate the fractional vegetation cover in Yanqing District in 2004.After accuracy verification,the accuracy of the pixel binary model is 87.76%,the accuracy of the improved linear spectrum hybrid model is 91.24%,and the neural network model The accuracy is 80.02%,and the overall accuracy result is an improved linear spectral hybrid model>pixel bisection model>neural network model.The accuracy of the improved linear spectral mixture model in each fractional vegetation cover model gradient is greater than 80%,indicating that combining the linear spectral hybrid model with NDVI improves the model accuracy and the model improvement is relatively successful.(2)Using the improved linear spectral hybrid model to calculate the fractional vegetation cover in Yanqing District during the three periods,the average value in 2004 was 0.51,the 2009 average was 0.58,and the 2014 average was 0.63.The fractional vegetation cover in Yanqing District has generally risen during the 10 years,showing the transition from low coverage to medium coverage and high coverage.Correlation analysis of fractional vegetation cover changes with elevation and slope gradients showed that the correlation coefficient between fractional vegetation cover and elevation was 0.83 in 2004,0.91 in 2009,and 0.87 in 2014.It can be seen that fractional vegetation cover and elevation have A certain positive correlation;similarly,at a slope of 0-45°,the correlation coefficient between fractional vegetation cover and slope was 0.89 in 2004,0.88 in 2009,and 0.91 in 2014,indicating a certain positive correlation at this time;For a gradient of 60°,the correlation coefficient in 2004 was 0.79,0.84 in 2009,and 0.82 in 2014,indicating a certain negative correlation at this time.The mean distribution of fractional vegetation cover on the aspect of the slope is:sunny slope>half sunny slope>half shady slope>shady slope>no slope.In the analysis of vegetation cover change in different land types,the mean value of fractional vegetation cover in forest land is the largest among all land types.In 2004-2009,the change in fractional vegetation cover of cultivated land was the largest,increasing by 0.24,2009-2014.During the period,the average change in fractional vegetation cover for construction land was the largest,increasing by 0.11.(3)After HL index and significant level test,the spatial simulation of fractional vegetation cover change by Logistic model is better.The result shows that the driving force factor for increasing fractional vegetation cover in 2004-2009 is the distance from the road in the order of weight.Returning farmland to forests>Distance from villages>Distance from rivers>Elevation>Slope>Population density,2009-2014 The driving force factors for increasing fractional vegetation cover are:Distance from roads>Distance from towns>Altitude>Slope>Per Capita GDP>Population Density;driving forces for decreasing fractional vegetation cover from 2004 to 2009 are Weighted by distance from town>Elevation>Slope>Population density>Per capita GDP>Distance to road,fractional vegetation cover 2009-2014 The order of decreasing driving force factors is also the Distance from the town>Altitude>Slope>Population density>Per capita GDP>Distance from the road.