Institute of Geography
University of Tübingen
Integrated Assessment of Gully Erosion Processes, Using Multispectral Remote Sensing, Stochastic Modelling, and GIS-based Morphotectonic Analysis; A Case Study in the Southwest of Iran.
Water erosion and especially gully erosion is one of the most effective phenomena that leads to decreasing soil productivity and pollution of water resources in many part of the world. Gully erosion as one of the most intensive land degradation processes especially in large parts of Iran, is the main threat for agriculture and range land. According to previous research lithology, vegetation density, topography as well as land use and land cover changes are effective drivers for soil loss in general and gully erosion in particular. In the first part of this thesis the susceptibility of the Mazayjan basin (MZJ) in the Southwest of Iran, where gully erosion is the main environmental threat, has been assessed. Therefore, a detailed terrain analysis and a stochastic modeling approach using mechanical statistics have been used. Among the terrain indices utilized in the prediction, the most important ones are: convergence index, plan curvature and slope. Gully erosion is the predominant type of water erosion and contributes to the sediment dynamics in a significant way in this catchment. Hence, in the second part of this thesis, GIS and satellite image analysis techniques were applied to derive input information for the numeric models to estimate the rate of soil erosion and deposition in the MZJ basin. Since the different types of water erosion such as rill, sheet and gully erosion are quite common in the MZJ basin, the Unit Stream Power-based Erosion Deposition (USPED) Model was integrated with a combined Stream Power Index (SPI) and Flow Accumulation (FA) approach. According to the integrated final map of erosion and deposition processes modeled with USPED including the gully erosion processes derived by SPI, round about 17.5 % of area is stable or characterized by very low erosion or deposition classes. Very high erosion values cover 28.2 % of the area, whereas 19.2 % of the area is related to deposition processes.
Tectonic activities in this study area that is part of the Zagros Mountains (ZM) significantly contributed to the formation of the existing drainage systems and hence, to landscape evolution. Neo-tectonics (Upper Quaternary) in form of earthquakes and associated uplifting, fracturing and faulting are still active in large parts of this area. In the third part of this study, we focused especially on the assessment of the vulnerability of the geologic formations to gullying and the effects of neo-tectonic processes inducing gully erosion using different data sources (ASTER GDEM., topography map information and aerial photographs) and resolutions (30 m, 10 m and 5 m) of digital elevation models (DEMs). An investigation on the location of gully features like headcuts and stream profile knickpoints reveals that the areas highly sensitive to gullying are related to stream sections showing uplifting and faulting. The TecDEM software was used to identify knickpoints showing that abrupt changes in the river profiles are located in the central part (colluvial/alluvial deposits) of the catchment. Hence, the location of knickpoints indicates tectonic activity in turn changing the drainage network along the longitudinal profile. The results illustrate that severe gully erosion is related to these tectonic processes, especially in the Southwest of the MZJ basin.
In the last part of this thesis, the spatial distribution of gully susceptibilities was modeled with a GIS-based statistical mechanics approach. Therefore, we used bands ratios from ASTER multispectral images taking into account the lithologic characteristics of the study area most likely influencing the spatial distribution of gully erosion. The results show that the multispectral analysis of the ASTER data yield valuable results in terms of mineral differentiation in the ZM area and hence, can be utilized as a useful tool for lithological mapping. In this study we used a statistical mechanics approach to assess the relation between existing gully locations and the combinations of predictor variables consisting in topographic indices and ASTER band rations. The spatial prediction shows that gullies have a high probability in areas with high amounts of salt, gypsum and silty deposits especially in the plain part of the study area. The proposed methodology allows conducting a proper gully erosion assessment in order to identify the priority areas for soil conservation and land use management in the Southwestern parts of Iran.