Research On Remote Sensing Monitoring Models For Typical Steppe Biomass At Middle Northern Slope In Tianshan Mountain

The slope center-section in North Tianshan is breeding different lawn type by terrain's superiority and characteristic, holds the extremely important status in the Xinjiang lawn animal husbandry production. The accurate prompt gain this region ground biomass as well as along with the time, spatial variation's characteristic is the key to realize highly effective and continually using lawn resources, therefore carries on the dynamic monitor and the estimation research using the satellite remote sensing technique to the lawn has the vital significance.This article uses the EOS/MODIS satellite remote sensing data, in the systematic study and summarizes the domestic and foreign lawn vegetation index and above the ground biomass remote sensing monitor research development foundation, take the slope center-section's typical prairie area of Tianshan north in Urumqi as the object to study, survey the ground biomass by remote sensing technique, has analyzed the correlational dependence between remote sensing vegetation index and biomass, and has established the dynamic estimation model of ground biomass in typical prairie area, will provide the efficient path to carry out the big area upland meadow belt estimation and the dynamic monitor for the present. This article main research conclusion is as follows:1. Vegetation index NDVI, RVI and the ground biomass exist extremely remarkable, therefore may believe that establishes remote sensing monitor model of lawn biomass by using vegetation index is feasible.2. In the research area, the analysis of vegetation index fitting trend, look at the fitting tendency with two planters by in the index relations scatter diagram's, vegetation index and ground biomass has obvious correlational dependence in the research area of meadow prairie, prairie, wilderness prairie, the vegetation index increases follow biomass.3. Through linear regression analysis and the non-linear regression analysis between two planters index and the ground biomass, obtains the meadow prairie and the RVI relevance is most remarkable, the estimation model is take RVI as an independent variable Yuan linear regression model; The prairie and wilderness prairie are most remarkable with the NDVI relevance, the estimation model is take NDVI as the independent variable conic section regression model.4. To research area vegetation index NDVI and the ground biomass seasonal variation characteristic through the dynamic monitor has carried on analysis, obtained meadow prairie, prairie, wilderness prairie dynamic estimation model respectively is a Yuan linear regression model, the conic section regression model, the power function curve model.5. Through confirmation analysis to the test point actual value and the inversion value, obtains meadow prairie, prairie, wilderness prairie dynamic model estimation precision respectively to achieve 83.06%, 90.85%, 88.06%.6. The MODIS vegetation index curve and the ground biomass curve can reflection well with lawn type vegetative season change process, this indicate that vegetation index and the ground biomass has remarkable relevance, simultaneously also explained that typical prairie area estimation model can meet the actual production needs in slope center-section of north Tianshan.