| The study,with Jinzhai Tianma Natural Reserve as a survey region,selected 32 sample plots in five forest types on the basis of data concerning secondary researches on the forest resources of Jinzhai County,of which,8 plots are coniferous forest,4 are mixed broadleaf-conifer forest,4 are evergreen and broadleaved deciduous forest,12 are broadleaved deciduous forest and 4 evergreen broad-leaf forest.The terrain factors of sample plots and the GPS coordinates were recorded,and five quadrats with size of50 cm x 50 cm within the sample plot were established along the diagonal to collect undecomposed and semi-decomposed litters.In addition,5 soil profiles of 80 cm x80cm x 60 cm were arranged in the shape of “S” and the soil,which was collected with cutting-ring method,were analyzed in the laboratory to identify the soil physical properties and water conservation capacity.The author selected 6 sample plots of various forest types with similar site factors(altitude,gradient,exposure etc.)from such 32 plots to compare and analyze the water conservation capacity,in order to understand water conservation features of litters and soil layers and identify the water conservation capacity of various forest types.At the same time,the digital image processing technique were applied to the Rapid Eye high-resolution remote sensing images of Tianma Natural Reserve,such images then were categorized by forest types.After extraction of characterization factors of remote sensing images,with water conservation of 32 sample plots as dependent variables and corresponding remote sensing factors and terrain factors as independent variables,an water conservation capacity estimation model with unit of pixel was established.The research results reveal that:(1)The gross standing crop of different forest types is between 3.30 to 5.70t/hm~2,and the order can be arranged as follows: coniferous forest> mixed broadleaf-conifer forest> evergreen and broadleaved deciduous forest> castanea seguinii broadleaved deciduous forest>evergreen broad-leaf forest>lacquer tree broadleaved deciduous forest.In general,it is concluded that the litter standing crop of coniferous forest is higher than that of the mixed broadleaf-conifer forest while the latter is higher than that of the broad-leaf forest.The maximum water-holding capacity of litters is between9.41-18.37t/hm~2,and such capacity of litter layer is 2.4-3.8 times of its standing crop since human activity has a great influence on the maximum water-holding capacity offorest litter.(2)In respect to physical properties of soil,the mean value of unit weight is vary between 0.948-1.167 g/cm3,and total porosity of soil is between 33.647-53.997%,the capillary porosity is between 23.451-38.722% and the non-capillary porosity is between 7.430-19.051%.The soil bulk density of the broad leaf forest is lower than that of the coniferous forest and soil of the former is more loose than the latter.The total porosity and capillary porosity of soil are decreasing with increasing soil depth,but there is no significant variation pattern for non-capillary porosity.(3)The gross value of saturated water storage of different forest types is between2333.545 to 2885.571t/hm~2 with order as follows: evergreen and broadleaved deciduous forest(2885.571t/hm~2)> castanea seguinii broadleaved deciduous forest(2790.750t/hm~2)>evergreen broad-leaf forest(2748.561t/hm~2)> mixed broadleaf-conifer forest(2461.676t/hm~2)> lacquer tree broadleaved deciduous forest(2440.694 t/hm~2)> coniferous forest(2333.545t/hm~2).The evergreen and broadleaved deciduous forest is of the highest water-holding capacity while the coniferous forest is of the lowest.(4)The total water storage of different types of forest is between 2349.780 to2898.222 t/hm~2 with order as follows: evergreen and broadleaved deciduous forest(2898.222t/hm~2)> castanea seguinii broadleaved deciduous forest(2809.116t/hm~2)>evergreen broad-leaf forest(2759.926 t/hm~2)> mixed broadleaf-conifer forest(2477.943t/hm~2)> lacquer tree broadleaved deciduous forest(2450.101 t/hm~2)>coniferous forest(2349.780 t/hm~2).The general law is concluded as broad-leaf forest >coniferous forest,mixed forest> pure forest.(5)After pre-processing of remote sensing images,the remote sensing image interpretation signs were set up.With application of the maximum likelihood method,such images were categorized and the mixed matrix method were then utilized to evaluate the accuracy of categorized results,which show that the overall accuracy of categorization is 80.04% and the Kappa efficient is 0.749,which meet the standard required in categorization of the whole survey region.(6)The non-capillary water storage data obtained through ground investigation were taken as the dependent variable and each band value of remote sensing image,exponential value of vegetation and terrain factors from actual measurement as the independent variable,with application of the multiple regression analysis method and combination with SPSS18.0 software,a model for such variables was established.After applying the backward regression method to screen independent variables,a resulting equation of linear regression is:Y=-69.159-3208.24NDVI+578.488CRM+0.36B-0.427R+0.218RE-4.174SLOPE of which,the multiple correlation coefficient R is 0.791,the determination coefficient R~2 is 0.626 and the adjusted determination coefficient R~2 is 0.536.There is a relatively strong correlation between the dependent variable and the independent variable,and high degree of model fitting.Through comparison and analysis of measured value and evaluated value of sample plots,with confidence coefficient of 0.05 and 0.01,the estimation accuracy of non-capillary water storage are 90.748% and 83.777% respectively,both are higher than 80%.As a result,the accuracy of such estimation model meets the requirement.(7)With application of estimating model and on the basis of area of arbor forest obtained through remote sensing images,it is estimated that the total water conservation quantity in the research area is 6.335 * 10~7 t. |
PDF Full Text Request