Effect of Microtopography on Spatial and Temporal Rill Initiation using Close Range Photogrammetry at Angereb Watershed North of Lake Tana, Ethiopia Introduction Rill initiation is geomorphologically important as sediment transport rates increase rapidly once rill incision occurs. It is controlled by slope steepness, contributing area and surface microtopography. Rill initiation has been predicted using inverse relation between unit contributing area and critical slope gradient. However, research results indicated the difficulty to find threshold area without taking into account the influence of surface microtopography (Desmet and Govers, 1997; Zapp and Nearing, 2005). The objective of the study was to identify and determine critical rill initiation areas on different soil textures under conventional tillage induced soil surface roughness in the highlands of Ethiopia. Materials and Methods Field experiments were carried out during rainy seasons (July-August) 2007 and 2008 at Angereb watershed ( ’ to ’ E and ’ to ’ N) in Ethiopia. It was conducted on five freshly tilled bare plots (3.7 m wide and 14.4 m long) with different roughness size and orientation (Fig. 1). Soil textures were clay, clay loam, loam, sandy clay loam and sandy loam. Clay loam and sandy clay loam plots had tillage oriented roughness. A calibrated Canon EOS 1Ds camera at 2.4 m height was used to take 14 overlapped photographs at scale of 1:130 (Fig. 2) after cumulative rainfall of 46, 116, 174 and 410 mm for clay and sandy clay loam soils in 2008 and after cumulative rainfall of 42 and 100 mm for clay loam, loam and sandy loam soils in DEMs were computed with 2 cm grid resolution using LPS in ERDAS IMAGINE software for the photogrammetric analysis. String Leg position Ground control points Sediment trough Tanks Rill network was generated from DEM. At head cut of sample rills that matched to digitized rills with defined minimum cross-section; unit rill contributing area, A c (m 2 m -1 ), slope gradient, S cr (m m -1 ), and random roughness, RR (m) were computed. Non-linear regression of the form Scr =a*A c b *RR c was used to predict threshold area (a) and relative parameters (b and c). Critical rill prone areas were identified where S cr *A c -b *RR -c exceeds the threshold area (a). ASA-CSSA-SSSA Annual Meeting, 1-5 November 2009, Pittsburgh Fig.2 Photography field setup Fig.1 Tillage induced surface roughness

Fig. 3 Areas prone to rilling (white shaded areas) after 42 and 100 mm cumulative rainfall for clay loam and loam soils. 42 mm100 mm 42 mm G. D. Gessesse 1, H. Fuchs 2, R. Mansberger 2, D. Rieke-Zapp 3, and A. Klik 1 ASA-CSSA-SSSA Annual Meeting, 1-5 November 2009, Pittsburgh Results 1) Threshold area for critical rill initiation Threshold area (a) was 0.20 to 0.50 from average rill contributing area of 1 m 2 and local slope gradient of % (Table 1). At initial events, rougher surfaces (C, SCL) gave low threshold area than smooth surfaces (CL, L, SL). Continuous increase in the threshold area was observed on rough surfaces as a result of gradual change to smooth surface. Due to rill formation, roughness was increased from the initial on smooth surfaces and led to reduce threshold area. Unit contributing area had minor effect on the prediction of rill initiation. The exponent b was less than 0.1 compared to 0.4 to 0.6 obtained in other studies (Vandeal, et al, 1996;Desmet and Govers, 1997) at different experimental conditions. Table 1. Threshold area for rill initiation ( a), relative unit contributing area (b), relative random roughness (c) parameters and percent of rill prone areas ++ for sandy clay loam soil the Derived DEM after 100 mm rainfall was affected by grass cover

The nature and size of initial surface roughness have a strong influence on the spatial and temporal variation in the threshold area and relative parameters of random roughness, and on rill network development. Rill sensitive areas are related to high surface roughness and/or to poorly consolidated plough ridges. Introducing surface microtopography for rill initiation significantly improved the prediction of rill initiation, and it can be applied to irregular surface topographies and able to improve the capacity of empirical erosion models towards estimation of representative rill erosion. Conclusions Fig.4 Areas prone to rilling (white shaded areas) after rainfall of 46, 116, 174 and 410 mm for clay and sandy clay loam 2) Prediction of areas prone to rilling Clay, clay loam and sandy clay loam soils show high percentage of rill prone areas (Fig. 4). Rough surface were more prone to initiate many rills (low threshold area) but merged into few larger rills drained out of the plot than smooth surface. In general, rill prone areas decreased with an increase in slope length, but, independent of slope length under oriented roughness for example for clay loam and sandy clay loam soils (Fig. 3 and 4). Rills start to occur on lower side of plough ridges with low bulk density and form knick points (Fig. 5) References: Desmet, P.J.J. and Govers, G., Two-dimensional modelling of the within-field variation in rill and gully geometry and location related to topography. Catena 29: Rieke-Zapp, D. H., and Nearing, M. A., Slope shape effects on erosion: A Laboratory Study. Soil Sci. Soc. Am. J. 69: Vandael, K., Poesen, J., Govers, G., and van Wesemael, B., 1996b. Geomorphic threshold conditions for ephemeral gully incision. Geomorphology 16: Acknowledgement: The study was financially supported by the Austrian Exchange Service (ÖAD). Fig.5 Rill initiation and development on clay site 1) Institute of Hydraulics and Rural Water Management, University of Natural Resources and Applied Life Sciences (BOKU), A-1190 Vienna, Austria 2) Institute of Surveying, Remote Sensing, and Land Information, BOKU 3) Institute of Geological Sciences, University of Bern, Switzerland Contact: ASA-CSSA-SSSA Annual Meeting, 1-5 November 2009, Pittsburgh