Forest Transformation In The Loess Plateau Based On Biomass Research On The Mechanism Of Time-space Differentiation And Influence

Forest ecosystem plays an important role in improving the real ecological environment,maintaining the global carbon balance,protecting biodiversity,soil and water conservation,water conservation,windbreak and sand fixation.In this paper,the land use data of Zhidan County from 1980 to 2015 and the remote sensing data from 1990 to 2015 are used to analyze the land use status and the temporal and spatial distribution of forest land transfer in Zhidan County.Based on the biomass data and the centroid position of the second class forest resources survey in Zhidan County of the Loess Plateau in 2005,the corresponding band reflectance,texture,vegetation index,topography,meteorology and other information are extracted with multiple values Using multivariate stepwise regression method and k-nearest neighbor algorithm(KNN)to model the forest biomass inversion in Zhidan County,the results of the study were interpolated by inverse distance weight method(IDW)to obtain the distribution map of forest biomass grade in Zhidan County.This paper puts forward an appropriate estimation method of forest quality transformation in the Loess Plateau by remote sensing inversion,identifies the growth law between the forest area and the forest quality in the process of forest transformation,summarizes the influencing factors of forest transformation,and puts forward policy suggestions according to local conditions.The results show that:(1)According to the land use data of six phases in Zhidan County from 1980 to 2015,the forest land area of Zhidan County is increasing continuously,from 49904.37 hm~2 in 1980 to 57811.95 hm~2.In terms of spatial distribution,from 1980 to2000,the outflow of forest land mostly distributed in the southern forest area of Zhidan County,mainly converting forest land into cultivated land,grassland,water area,urban and rural construction land.From 2000 to 2015,the direction of woodland outflow is mainly reflected in the expansion of urban construction land and grassland conversion,mainly in the southern border area of Zhidan County.The cultivated land is mainly distributed in the north of Zhidan County.From 1980 to 2000,the area of forest land decreased slowly.From 2000to 2015,the area of forest land increased gradually,and then gradually increased to a stable stage.The change of forest land in Zhidan County accords with the kurenz curve of forest environment,so the forest transformation has taken place in Zhidan County.(2)Forest biomass was retrieved from Landsat image in Zhidan County.Band math,vegetation index,rainfall and so on were significantly correlated with forest biomass;The texture information extracted by 9×9 window has the best fitting effect for forest biomass model,and the second-order moment(SM)and synergy(HO)have a good correlation with forest biomass;there is no obvious difference between the regression results of Mahalanobis distance(MD)and Euclidean distance(ED)of KNN algorithm regression model,and RMSE gradually decreases with the increase of K.when k value is 10,the decrease trend of root mean square error of the model is slow.The root mean square error of forest biomass estimated by stepwise regression method is 14.10m~3/hm~2,R~2=0.558,bias=0.5866;the root mean square error of forest biomass estimated by KNN algorithm is 14.05m~3/hm~2,R2=0.647,bias=0.0466,KNN The inversion effect of the model is better than that of the stepwise regression model.According to the estimated value of the optimal model,the distribution map of forest biomass grade in Zhidan County is generated,which is consistent with the actual situation of returning farmland to forest area.KNN method is an efficient and reliable method for remote sensing estimation of forest volume and inversion of forest parameters.The results can provide a practical tool for monitoring the ecological and environmental benefits of the project.(3)Based on the spatial econometric model,the change of forest biomass in Zhidan County was analyzed.The results showed that the global Moran’s I value of forest biomass in 1990-2015 in Zhidan County was greater than 0,showing positive spatial autocorrelation,P value was less than 0.01,Z value was greater than 0.01 1.96,indicating that the biomass distribution is not independent,and the spatial correlation of forest biomass is very significant.This shows that the distribution of local forest biomass presents a spatial aggregation,that is,the forest biomass in the surrounding areas with higher biomass is also higher,while the forest biomass in the surrounding areas with lower biomass is also lower.The spatial lag model(SLM)is used to measure the influence of natural geography,social economy,population and forestry policy factors on Forest Transformation in Zhidan County.Under the condition of 5%significance level,the top indicators are annual rainfall,population density,distance to main road,policy and engineering investment,which shows that abundant rainfall is the influence factor under natural environment The main driving forces of forest biomass change in Zhidan County are the direct causes of forest biomass change,human activities,road accessibility and policy input.With the development of economy,it brings about the transfer of labor force and population.At the same time,the expansion of human activities continues to encroach on the space of ecological environment,which has a positive and negative impact on the quality of ecological environment,forest land area,forest resources,etc.(4)From the point of view of the increase and decrease of forest quantity and quality,the forest area of Zhidan County has been growing continuously,and the forest transformation has been completed in an all-round way.The growth rate of forest biomass per unit area in Zhidan County lagged behind that of forest land area for one period.From1990 to 2005,the growth rate gradually increased.By 2005,the increase rate of forest area reached the peak.From 1990 to 2010,the growth rate of forest biomass per unit area in Zhidan County increased.In 2010,the growth rate of forest biomass reached its peak,and in2010 to 2015,the growth rate decreased.With the long-term growth of forest biomass,the increase of forest biomass has lagged behind the change of forest area.In the process of forest transformation,the transformation of forest quality lags behind the transformation of forest area for one period.