Extraction Of Structured Vegetation Factor Index In Regional Soil Erosion Based On Remote Sensing

Vegetation factor is regarded as one of the key parameters of soil erosion prediction model and the most important basis of soil and water conservation evaluation. A scientific and reasonable structured vegetation factor index on regional scale by remote sensing technology should be proposed and obtained, which is a theory research urgently needed in soil and water loss monitoring and ecological environment construction.This paper selected Yanhe river basin as a case study. To achieve the goal that estimating land surface parameters more accurately, Landsat5 TM was selected and processed by geometric correction to match with DEM, calibration and atmospheric correction to convert to surface reflectance. Combined with vegetation observation in field, the relationship between land surface parameters and observed vegetation fractional coverage was discussed, and then a structured vegetation factor model was proposed in study area. In succession, some experiment data were used to verify the precision of the model and a further analysis was accomplished. Based on the upper methods, the following conclusions were made:1 According to the main factor in vegetation soil and water conservation and remote sensing mechanism, a index system for vegetation factor in regional soil erosion studies is established with remote sensing data. First, based on the summaries in researches on vegetation resources characteristics and soil and water conservation mechanism, it can be got that the vertical structure of vegetation is the most important factor in affecting the soil and water conservation benefits. The traditional vegetation fractional coverage can not reflect the vertical structure of vegetation and the impact of different layers on soil erosion. So, a more reasonable index system should contain two important aspects: one is green vegetation fractional coverage index, which will represent the forest, shrub and grassland coverage information; the other is yellow vegetation fractional coverage index, which can represent the coverage information of crop residue, such as 'aged' residue, drying grassland etc.2 A structured vegetation factor is proposed with remote sensing technology in Yanhe river Basin. Combined remote sensing images with soil line, a kinds of vegetation indices are derived. Doing correlation analyses between the vegetation indices and the observed sample data in field, the results show that PVI(Perpendicolar Vegetation Index) is the optimal indicator for green vegetation coverage index, and NDTI(Normalized Difference Tillage Index) can stand for the yellow vegetation index. Through multivariate regression analysis on PVI, NDTI and structured vegetation index(Cs) evaluated by observed vegetation coverage data, an regression equation of them is put forward: Cs = 0.778 * NDTI + 0.756 * PVI.3 Verification and analysis on structured vegetation index. Comparing the observed data of test samples to its evaluation values, we find that the structured vegetation index obtained with remote sensing data and technique can reflect the actual capability of vegetation soil and water conservation more accurately and reasonably. The value of structured vegetation index in Yanhe river basin is mainly concentrated in 0.07-0.35.The study area has a poor vegetation coverage in the mass. The order of structured vegetation index in different land use types is forest>grassland>crop> settlement place>water>undefined land use types. Aspect also has influence on Cs, but it only has a obvious influence on forest coverages. Vegetation coverage of north slopes is better quantity than east and west slopes in quantity, and the south slopes is the worst.