Study On The Surface Of The Forest Fuel Estimation Model In The North Slope Of Tianshan Mountain Based On3S Technology

Forest fuel is a material base of burning,Within a certain range, the higher the fuelloads, the bigger the fire risk, In the event of fire, fire intensity is also bigger, spread fastand difficult to put out in a timely manner.Studying the Forest spatial distribution pattern and fuel loads, not only be able topredict the occurrence of forest fires, but also to provide a theoretical basis for theprevention of forest fires. it is important to operatiing and managing forests and thedevelopment of fire prevention measuresTo estimate the amount of forest fuel loads through remote sensing,Factor for remotesensing and topographical factors such as participation, where extraction is essential part ofthe analysis of the variables,But as the uncertainty factors in selecting, making theextraction of effective factors and their mutual combinations of difficult，The author usethe SPSS17.0statistical software,Study on and application of stepwise regression analysisin mathematics in a variety of remote sensing factor in forest site factors andvariables,Filter out of forest fuel load with a significant role effective parameterfactor,Conducive to the development of interventions, based on information provided bythe departments of forest fire prevention in order to enhance remote sensing estimationaccuracy of forest fuel load.In this study, by taking the northern slope of Tianshan surface litter layer of forest fuelas the research object, major survey of litter layer thickness, type and weight of the soilfauna surveys, collecting survey data and samples combustible stand factors, experimentalanalysis.Using stepwise regression analysis, correlation analysis was carried out on the surfacefuel loads and stand factors, and establish the fuel loads and the forecast model of remotesensing factors, forest site factors. By extracting and consistent measurement of surfacespectral data of NDVI MODIS algorithm, combined with the measured fuel load, fuel loadof building surface spectral model, and through building surface spectral data relationshipbetween NDVI and MODIS NDVI models.The eventual establishment of combustibles based on MODIS NDVI remote sensingestimation model for loads, for estimating fuel loads, this leads to the followingconclusions： （1） Through the analysis of surface litter layer thickness and unit area on the quality ofsoil animals and types come in gradually reducing the quantity and quality of the soil faunafrom west to east. Decomposition in the soil, the less the thicker layer of litter.But the iliforest litter, soil invertebrate animals too much, so much litter is broken, there is a firefrequency. Hami forest litter less, less soil invertebrates, so less is litter decomposition, firefrequency is relatively more.（2） Surface can burning real contains volume and forest site type of slope, andaltitude, and soil animal of quality, and dead branches from the leaves layer of thicknessand MODIS NDVI data for correlation analysis and established gradually regressionequation concluded that can burning real contains volume by multiple factor of effect, asingle factor on can burning real contains volume of effect is one-sided of, in research canburning real contains volume Shi to combined multiple factor to for research analysis.（3） The thickness of the north slope of tianshan mountain forest soil litter,generallybetween0²3cm, A0average annual accumulation of litter layer soil is2.11t/hm²（dryweight）, the thickness of the surface fuel forest changes often around5cm, but the standgrowth lush forest land also has10-12cm thick.（4） Selected forest site factors and remote sensing information, including eightindependent variables to establish pixel units of the northern slope of Tianshan forest fuelload estimation model. The estimation model is y=372.869-0.121×x₃+1.156×x₄+2.662×x₈，Residual correlation analysis by as much as5%of the tested,It can effectivelyforecast the north slope of tianshan mountain forest fuel loads.