Comparison of gully erosion estimates using airborne and ground-based LiDAR on Santa Cruz Island, California Ryan L. Perroy a, ⁎, Bodo Bookhagen b , Gregory P. Asner c , Oliver A. Chadwick b a Department of Geography and Earth Science, University of Wisconsin-La Crosse, 1725 State Street, La Crosse, WI 5460, USA b Department of Geography, 1832 Ellison Hall, UC Santa Barbara, Santa Barbara, CA 93106-4060, USA c Department of Global Ecology, Carnegie Institution, 260 Panama Street, Stanford, CA 94305, USA abstract article info Article history: Received 11 October 2009 Received in revised form 9 January 2010 Accepted 14 January 2010 Available online 22 January 2010 Keywords: Erosion Gully LiDAR Volume estimates Mediterranean climate Gully erosion removes comparatively large volumes of soil from small areas. It is often difficult to quantify the loss of soil because the footprint of individual gullies is too small to be captured by most generally available digital elevation models (DEMs), such as the USGS National Elevation Dataset. Airborne LiDAR (Light Detection and Ranging) has the potential to provide the required data density, but an even newer class of ground-based sensors may provide better local resolution at lower cost. In this study, we compared digital elevation models produced by airborne and ground-based LiDAR systems with ground-based geomorphic and geodetic survey data to determine their utility in quantifying volumetric soil loss due to gully erosion in a heavily degraded watershed (7.55 × 10 -2 km 2 ), on southwestern Santa Cruz Island in southern California. Volumetric estimates of the eroded sediment were produced by comparing the LiDAR-derived DEMs of the gully system to a modeled pre-erosion surface. Average point densities were significantly higher for the ground-based LiDAR system and provided more detailed information; however, its limited scanning footprint and side-looking orientation presented serious challenges in collecting continuous data from deeply incised gullies, making the airborne system preferable for this type of investigation and likely for most applications where heavy topographic shadowing is prevalent. © 2010 Elsevier B.V. All rights reserved. 1. Introduction Gully erosion, the removal of soil from narrow channels via the accumulation of surface runoff, tends to produce more sediment loss than other forms of soil erosion such as overland flow or rilling (Wasson et al., 2002; Poesen et al., 2003; De Vente et al., 2005; Huon et al., 2005; Wells et al., 2009). Gullies are generally defined by their channel depth, which for permanent gullies can range from 0.5 to 30 m (Soil Science Society of America, 2001). They are also one of the most destructive forms of erosion, destroying soil, undermining infrastructure, damaging agricul- tural fields, altering transportation corridors, and lowering water tables (Valentin et al., 2005). Furthermore, their damage is difficult to reverse. Gully erosion dramatically affects sediment budgets and flux rates, and influences stream dynamics as evidenced from data on hydrographs (Kelsey, 1980; Costa and Bacellar, 2007). In addition, in some areas gully erosion is directly linked to changing climatic conditions (Nearing et al., 2004; Chaplot et al., 2005; Nunes et al., 2008, 2009). Quantifying gully erosion will assist in understanding gully formation and spatiotemporal evolution. Although gullies are visually striking, their small spatial extent generally renders them undetectable in most generally available topographic datasets. The goal of this work was to test the viability of applying airborne and ground-based LiDAR technology to quantify gully erosion on southwestern Santa Cruz Island in southern California. Limited by the minimum spot spacing of the airborne LiDAR dataset, our study only considered gullies with a width of >2 m. Previous attempts to quantify gully erosion have usually involved labor-intensive field measurements, such as field tapes, micro profilers, theodolite or total station, and differential GPS methods (Lawler, 1993, Casalí et al., 2006; Moody and Kinner, 2006; Nyssen et al., 2006; Rustomji, 2006; Wu et al., 2008). Erosion pins have also been used to measure gully wall retreat (Ireland et al., 1939; Brumbaugh, 1983). In addition to being spatially limited in scope, these methods are all time consuming, tedious, labor intensive, and expensive. By contrast, high-resolution LiDAR datasets offer the potential to efficiently measure gully volumes at the landscape scale. Until recently, the resolution of available topographic digital elevation models (DEMs) (National Elevation Dataset, USGS topographic maps, Shuttle Radar Topographic Mission), has not been fine enough to capture small features such as gullies. An exception are DEMs produced via digital photogrammetric analysis, which have allowed measurements of gully-scale erosion (Betts and DeRose, 1999; Martinez- Casanovas, 2003; Martinez-Casanovas et al., 2004). Recent work on three- dimensional gully monitoring using low altitude, unmanned platforms has produced extremely high-resolution (5 and 7.5 cm pixel size) DEMs (Marzolff and Poesen, 2009). Although photogrammetry has made impressive advances in the past few years, the growing proliferation Geomorphology 118 (2010) 288–300 ⁎ Corresponding author. Tel.: +1 608 785 8334; fax: +1 608 785 8332. E-mail address: perroy.ryan@uwlax.edu (R.L. Perroy). 0169-555X/$ – see front matter © 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.geomorph.2010.01.009 Contents lists available at ScienceDirect Geomorphology journal homepage: www.elsevier.com/locate/geomorph