| Canopy interception of forest is a very important part of forest hydrology proccess. The first layer that rainwater reach is the canopy of forest. one of the main hydrology character of canopy is the interception of rainwater. On the one hand the interception prevents rainwater flushing the surface directly, slows the time rainwater hit the surface, which makes the litter layer and the soil layer able to fully absorb the rainwater and reduce the soil erosion. On the other hand, the rainwater that was intercepted by the canopy will evaporate from the leaves and get into the local water cycle of the forest eco-system, and play a role in water content.The study area of this thesis is the Jinggangshan National Nature Reserve, about 95% of the land of which is covered by forest. The climate type of the study area is subtropical monsoon climate with annual precipitation 1400mm around. As the Nature Reserve is an important source place of water of Ganjiang river, the interception study of forest canopy could be the base of some science study such as water balance, water resources evaluation et.al, and will provide scientific help for the protection and management of forest and water resources and the environment.Model built by existing study is applied in the thesis for forest canopy interception simulation, the disadvantage of which is improved before using. The model’s undetermined parameters is calibrated by field data, and it was programed. For area scale simulation of the interception model, remote sensing model for calculating LAI is built. Canopy interception simulating has been excuted using programed model and by analyzing the simulation result, the features of forest canopy interception at area scale are demonstrated in different time scales.Aston canopy interception model is used and improved:a) considering the impact of water capacity of trees and other metelogical factors rainfall events are distinguished as two types, which are that can fill the water capacity of trees and that can not, perpectively, and that can make the model suit for any rainfall event; b) the dryness of the tress is taken into account, which could avoid the assumption that the tree is totally dry before a rainfall event; c) the correlation between water capacity and ecological factors i.e. LAI of trees is is built for simulation at area spatial scale. Besides those metelogical factors that could be obtained from conventional metelogical measurement, the parameters, which need be calibrated by experinments,of the improved model are LAI, eveperation of wet trees e and the water capacity of trees 11m.The leaf area index is the function of cover density of forest canopy in study area. Ten samples of LAI and c are obtained by using the LAI measuremt instrument of LP-80 and camera, respectively in field surveying. By analyzing the ten samples of c with corresponding LAI, the equation for LAI and c is calibrated as LAI= 6.258c1.669, the determining coefficient of which is 0.867.The parameters of Penman-Monteith model for calculating the evaperatin of wet trees are all from conventional metelogical measurement, and study of this paper shows that P-M model suit for the study region. The evaporation of 2013a is calculated in stdy region with the temporal step of 0.5h, the result of which is closed to the the mearserment data that is annual averaged,with an offset of 3% in total at Poyan lake basin. And monthly evaporation of both regions are closed except for March and September.Key parameters Icm* in the improved interception model could be estimated using retrieval models of RS. Field data of interception experiments shows that the amount of rain water flow through tree stem accounts for only 1.76%-13.91% of the atmosphere rainfall. This study promoted more the interception model promoted by Aston, and its undetermined parameters the capacity for holding water 11m is caliberated using LAI, which is Icm*= 0.2963exp(0.5595LAI). When evaluating the precision of the calibrated model, the coefficient of efficiency of model in simulating individual rainfall interception at experiment field No.7 is 0.71 higher than that of empirical models and the root-mean-square error is 0.87 mm lower than that of empirical models, which means the caliberated premoted Aston model is better than empirical models. The coefficients of efficiency in simulating forest interception amount in different experiment fields at different experiment periods vary from 0.61 to 0.89 and the root-mean-square errors are from 2.11 to 4.40 mm The evaluating result shows that the calibrated model could meet the need of application.The result of canopy simulation by computer shows that at area spatial scale and annual temporal scale the interception ratios of different forest types show little difference. The canopy interception models are programmed using Python, which have been packeged as a Toolbox in ArcGIS software. Raster data for canopy interception amount/ratio in each rain, each month, each year and the accumulation amount from March 26th,2012 to September 20th,2014 simulated by the calibrated model is simulated by the program and analyzed in this thesis. The analysis at area scale of accumulation and annual interception shows that the canopy interception ratio of forest has little difference(21%±2%) between each year. The total amount of interception of the study area by canopy in simulation period is 675.35mm·m-2, the interception ratio is 20.34%. Study area’s annual interception amount is 285.43 mm·m-2, interception ratio is 20.47%. The interception amount of different forest types vary from 283.23 mm·m-2 to 310.17mm·m-2, and the interception ratios of them are from 20.31% to 22.25%.The monthly interception analysis at area scale shows that there is big difference of monthly canopy interception ratio, which is negtively correlated with monthly precipitation(R2= 0.44). The monthly canopy interception amount is dramatically positively correlated with the monthly precipitation(R2= 0.8±0.05), and it could be estimated using monthly rainfall directly, the coefficients of determination of which vary from 0.799 to 0.806 and the mean residual of squares is from 8.42 to 9.00 mm. The monthly canopy interception amount of rainfall could be estimated from monthly precipitation and the days with the precipitation more than 0.6 mm, which have determine coeffiecences vary from 0.911 to 0.914 and the mean residual of squares is from5.69-6.11 mm. It shows that methods using parameters both monthly precipitation and the days with daily precipitation more than 0.6 mm to estimated monthly canopy interception amount of forest at area scale is better than the methods only using montly precipitation. |
PDF Full Text Request