Study On Monitoring And Modelling Gully Erosion On The Chinese Loess Plateau

Gully erosion is a serious environmental problem and the primary source of sediment loss on the Loess Plateau of China, and seriously interferes with the agricultural area of the surrounding land, produces large amounts of sediment-inducing siltation of downstream reservoirs and even causes some catastrophic flooding and pollution. The spasmodic nature in space and time of gully development, combined with monitoring techniques limitations, make it very difficult to predict gully processes. Recently, however, increasing attention has been paid to gully erosion with the development of new technologies. Study on monitoring and modelling of gully erosion on the Chinese Loess Plateau can provide an important theoretical basis and reference for understanding and controlling gully erosion.The present study was carried out Forest Park rehabilitated from agriculture and Qiaogou catchment in the central Loess Plateau and Caijiachuan basin in the eastern Loess Plateau of China. To explore where the sediment sources came from in the catchment, QuickBird imagery and 5 m resolution digital elevation model were used to compute the sediment of check dam in small catchment, and the intensity of soil erosion on inter-valley was estimated by China Soil Loss Equation (CSLE). Factors related to activity of gully heads were analyzed using the mathematical statistic analysis based on field survey. In order to evaluate the accuracy of extracting morphological parameters of gullies including area, perimeter, and length from QuickBird imagery by manual visual interpretation,3D laser scanner was used to measure the topography of gullies within two weeks after the QuickBird images were taken, on this basis, the values of morphological parameters extracted from DEM were taken as actual values and errors of morphological parameters of gullies extracted from QuickBird imagery were determined. By using QuickBird imagery, we estimated gully head retreat rates, assessed the factors leading to gully development and modeled gully area growth rate at the catchment scale. We characterized gully morphology (i.e. volume, area, length and cross-section), identifed the factors affecting gully volume and established a model for assessing gully volume using 3-D scanner data, and gully erosion volume were estimated based on the model using parameters measured from QuickBird imagery. The main research results are as following:Where slope degree is more than 25, grassland and other land (including, industrial and mining construction land and bare land) in the inter-valley of small catchment were the main soil erosion source, and would be the key area of soil and water conservation in future. Soil erosion of the valley were 3.21 to7.51 times more significant as sediment sources than the inter-valley, and the sediment yield by valley erosion represent 65.62 to 88.04% of total sediment yield caused by water erosion.Slope aspect has a significant impact on gully head activity, and the inactive gully heads mainly located in shade slops and the active gully heads mainly located in sunny slops. The effects of the local slope and drainage area on gully head activity decreased as vegetation coverage increased in upslope drainage. Vegetation of the surface 10 m above gully head also has a clear impact on gully head erosion. Grasses, compared with measures of combining grasses and shrubbery, can play a less important role in controlling gully head erosion unless it has higher vegetation coverage.The morphological parameters of gullies were extracted by manual visual interpretation based on QuickBird imagery, and the average relative error of area, perimeter and length of gully were both about 5%. In addition, by overlaying the two valley regions extracted from QuickBird imagery and DEM, the area that the distance offsets was larger than 0.6 m remained exceed 96%. The average relative error of gully length was 5%, with corresponding average of average absolute error of 0.75 m. The larger the area and length of gully were, the smaller the errors were. The accuracy of manual visual interpretation was also influenced by the vegetation types near the gully boundary. Catchments in which land covers were native grassland had higher precision than those with shrub and grass.From 2003 to 2010, the maximum retreat rates of gully heads in the 30 investigated catchments ranged between 0.23 and 1.08 m/a, with an average of 0.51 m/a. The ratio of bank gully growth area to valley area changed from 0.49 to 9.45%, depending on land use, with average increases of 3.94,4.00 and 2.09% for the three land use types identified, i.e. mixed use, grassland and forestland, respectively. Correlation analysis indicated that the effects of topographic factors on gullies decreased as vegetation coverage increased in upslope drainage areas and that vegetation coverage exceeding 60% in upslope drainage areas can significantly control bank gully development. A model was built to predict the bank gully area growth rate (Ra, m2/a) with upslope drainage area (Ai, m2), local slope gradient (S, m/m) and the proportion of the area with vegetation coverage below 60% in upslope drainage areas (Φ0.6) at the catchment scale. The regression equation is in the form Ra= 0.1540 [(Φ0.6Ai)0.24S]3.2588.The gully volume can be computed using linear or planar parameters, and gully erosion also can be perdicted based on changes of linear or planar parameters. The high R2 value indicated that gully length (L) was a good predictor of gully volume (V) in the region; however, compared with the V-L relation, V with gully area (Ag was more pronounced, of the form V= 0.2762Ag1.3971 (R2= 0.91). Gully erosion volume, estimated based on the V-Ag relation using parameters measured from QuickBird imagery, ranged from 0.41 to 36.76 m3/a and averaged 10.32 m3/a, which is equivalent to 0.62 to 55.14 t/a, with an average of 15.48 t/a. The results show that it is feasible method to monitor gully erosion using high-resolution remote sensing image and 3-D laser scanner.