Spatial and temporal morphological changes in McHugh's Gully measured using aerial photogrammetry and geographic information systems

The study addresses the need for the quantitative analysis of gully growth which is one of the neglected dimensions of erosion studies. Gully growth is reflected in the spatial and temporal morphological changes of the gullies. In addition, the influence of causal factors on gully morphology is likewise vital in examining the growth of the gullies. The fast detection of the extent of gully erosion and the rapid acquisition of data is vital in designing measures to stabilise the gullies, control the formation of incipient gullies, and monitor the growth of gullies over time. It was envisioned that the integration of aerial photogrammetry and geographic information systems (GIS) would serve this purpose. Data extracted from aerial photographs can provide important inputs for a GIS database dedicated for monitoring gully systems and providing information to assist erosion control measures. The gully system used for the study was McHugh's Gully which is situated north of Ballarat, Victoria. The morphological characteristics considered in this study were length, width, and order of the gullies and the fractal dimension of the entire gully system. Photo-interpretation of stereoscopic pairs of large-scale aerial photographs (1: 15 840 and 1:25 000) enabled the delineation of the gully outline, classification of the vegetation cover, and measurement of the gully width. Gully outlines and other spatial features were digitised to form the thematic layers for a GIS database. Elevation data was measured from topographic maps and was used to derive the slope and aspect classes. A grid with a cell size of 100 x 100 m was generated to store the attribute data. A PC Version of ARC/INFO and IDRISI were used to manipulate both raster- and vector representation of the data. The findings indicate that the total length of the gully system increased by 0.30 km/yr and affected an additional three hectares of land each year between 1946 and 1980. The majority of the gullies have widths of less than 6.25 m although some gullies were more than 18.75 m wide. McHugh's Gully can be measured to the seventh gully order following Strahler's stream ordering scheme. The branching out of first and second order gullies profused after 1946. The complexity of the gully system was also measured using the concept of fractals. The gully system has a fractal dimension that increased from 1.26 in 1946 to 1.55 for the period 1972-1991. Results also indicate that interrelationships among the gully morphological parameters exist. Evidence show, in general, that longer gullies have shorter widths and the first and second order gullies are longer than the higher order gullies. The study confirmed that rainfall, vegetation cover, slope, and aspect of maximum slope influenced the morphology of the gullies. Gullying increases with more frequent rain events. The amount of rainfall in the McHugh's Gully area was not linearly correlated with the year but the frequency of annual, autumn, winter and spring rainfall showed strong positive correlation with the year of record. Gullies found in the dense vegetated areas were longer with narrower widths. Moreover, the non-uniformity of the topographic characteristics of the area induce the formation of incipient gullies and the growth of existing ones. Most gullies developed on 10-19 per cent slopes and the direction of the flow of sediments is towards the north-east and north-west. The dynamic nature of gully processes is associated with difficulties in modelling and simulating the growth of gullies as reflected in the contrasting patterns in gully morphology. However, the monitoring of the growth of the gullies and the fast detection of gullied areas is facilitated through the integration of aerial photogrammetry with geographic information systems. Additional research towards the improvement of the methods described in this thesis are therefore encouraged