Welcome to Chinese Journal of Ecology! Today is Share:

cje

Previous Articles     Next Articles

Soil erosion evaluation and landscape pattern analysis of loess hilly gully area based on RUSLE-SMA: A case study of Caijiamiao watershed in Qingcheng, Gansu Province of Northwest China.

PAN Jing-hu, WEN Yan**   

  1. (College of Geography and Environment Science, Northwest Normal University, Lanzhou 730070, China)
  • Online:2013-02-10 Published:2013-02-10

Abstract: By using a linear spectral mixture analysis (SMA) model, the vegetation coverage in the Caijiamiao watershed in Qingcheng County of Gansu was extracted from the Landsat TM images, and the factors affecting the vegetation coverage were obtained. With the help of the revised universal soil loss equation (RUSLE), the amount of the soil erosion in the watershed in 2003 and 2010 was estimated. The relationships between the soil erosion and the land-use type and topographic factors were analyzed, and the changes in the soil erosion landscape pattern were studied. In the 7 years from 2003 to 2010, the soil erosion amount in the watershed increased from 3.61×106 to 4.48×106 t·a-1, with an increment of 24%. The mean annual soil erosion intensity increased from 8590.23 t·km-2 to 10652.01 t·km-2. The soil erosion varied greatly with land use type, being more serious on unutilized land, grassland, and garden land, and on the areas with a slope greater than 15°, elevation below 1395 m, and aspect west. Overall, the erosion landscape in the study area was highly fragmented, the total number of patches reduced, fragmentation degree lowered, landscape heterogeneity decreased, and landscape shape tended to be regular. Except the slightly eroded landscapes, the landscape types tended to be simplified.