Sand transport under increased lateral jetting of raindrops induced by wind G. Erpul a, ⁎, D. Gabriels b , W.M. Cornelis b , H. Samray a , T. Guzelordu a a Department of Soil Science, Faculty of Agriculture, University of Ankara, 06110 Diskapi-Ankara, Turkey b Department of Soil Management and Soil Care, Ghent University, Coupure Links 653, B 9000 Ghent, Belgium abstract article info Article history: Received 6 May 2008 Received in revised form 16 August 2008 Accepted 22 August 2008 Available online 5 September 2008 Keywords: Sand transport Rain incidence angle Raindrop lateral jetting Compressive stress Shear stress Wind tunnel experiments for ‘Raindrop Detachment and Wind-Driven Transport’ (RD–WDT) process were conducted under improved lateral jetting induced by wind velocities of 6.4, 10, and 12 m s - 1 at nozzle operating pressures of 75, 100, and 150 kPa. Wind-driven rainfalls were also incident on the windward and leeward slopes of 4° and 9° to have a broad variation in the angle of incidence. The objective of this experimental set-up was to distinguish the roles of both impact components of obliquely striking wind- driven raindrops on RD and wind on WDT. Raindrop impact components and reference horizontal wind were quantified by normal (E tz ) and horizontal (E tx ) kinetic energy fluxes and wind shear velocity (u ⁎ ), respectively, to physically model the process of RD–WDT. The results showed, at each level of u ⁎ , differential sand transport rates by RD–WDT (q m(RD–WDT) ) occurred depending on the magnitude of raindrop impact components, and q m(RD–WDT) increased as the relative contribution of E tz increased. Although E tx was more correlated with q m(RD–WDT) than E tz , the extreme increases in E tx at the expense of E tz brought about no increases but decreases in q m(RD–WDT) . An RD–WDT model was built under the process of examining the discrete effects of E tz and E tx on RD together with u ⁎ and resulted in a better coefficient of determination (R 2 = 0.89) than only total kinetic energy (E t ) did alone with u ⁎ (R 2 =0.84). In this study, E tx was strongly related to u ⁎ and not to E tz , which was the principal difference from the previous rainsplash studies, which relied on the compensatory lateral jet development by the compressive pressure build-up at the raindrop– soil interface. Including E tx in the RD–WDT model both separated the distinct role of each raindrop impact component in RD and improved the performance of u ⁎ in WDT by better distinguishing its interaction with E tx , which was not explicitly separated in previous models of RD–WDT. © 2008 Elsevier B.V. All rights reserved. 1. Introduction Kinnell (1999, 2005) reviewed the modes of raindrop impact- induced erosion processes and prediction. Of those, the mode of “Raindrop Detachment and Splash Transport” (RD–ST) takes place prior to the development of runoff, and clearly, in this case there is no transport of detached particles other than downslope gradient in the system. There are many studies of explaining the relationship between slope gradient and RD–ST (Savat and Poesen, 1981; Poesen and Savat, 1981; Moeyersons, 1983; Poesen, 1985; Riezebos and Epema, 1985; Wright, 1986, 1987), and recently, Furbish et al. (2007) gives the details of the transfer of momentum from raindrops to sand particles that contribute downslope transport of those for RD–ST. They described this momentum-transfer process under controlled labora- tory conditions using high-speed images of drop impacts on sand targets. The conclusion they arrived at was that, as slope gradient increased, more sand particles moved downslope and move farther downslope than upslope. The greatest radial distance that a sand grain moved was around 20 cm or less in their study. Because of this, RD–ST is generally described as a transport-limited process, particularly when it functions on large areas (Kinnell, 1999, 2005). On the other hand, a transport process in which soil particles are detached by raindrop impact and afterward carried by wind instead of slope gradient is known as ‘Splash-Saltation Transport’ (SST) or described as ‘Raindrop Detachment and Wind-Driven Transport’ (RD– WDT) (De Ploey, 1980; Jungerius et al.,1981; Rutin, 1983; Jungerius and Dekker, 1990; De Lima et al., 1992; Erpul et al., 2002, 2004; Cornelis et al., 2004a,b). This system operates until runoff onset during wind-driven rains, but its transport (WDT) is independent of slope gradient and aspect (Erpul et al., 2004) as wind is the transporting agent. Unlike RD–ST, RD–WDT can significantly contribute to the total sediment transport from interrill areas (Erpul, 2001; Erpul et al., 2003a) and can transport particles as far as 7 m (Erpul et al., 2002, 2004) compared to the limited travel distance of RD–ST (Furbish et al., 2007). Depending on the prevailing wind direction, either uphill or downhill transport of sediments can occur by RD–WDT (Erpul et al., 2002; Visser and Sterk, 2007). Additionally, in contrast to the transport-limited system of RD–ST, RD–WDT could be a detachment- limited system because RD significantly varies with the variations in raindrop trajectory and frequency due to the changes in the angle of raindrop incidence (Erpul et al., 2003b). Geomorphology 104 (2009) 191–202 ⁎ Corresponding author. Tel.: +90 312 5961796; fax: +90 312 317 8465. E-mail address: erpul@agri.ankara.edu.tr (G. Erpul). 0169-555X/$ – see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.geomorph.2008.08.012 Contents lists available at ScienceDirect Geomorphology journal homepage: www.elsevier.com/locate/geomorph