Measurement and prediction of sediment production from unpaved roads,St John,US Virgin Islands

Excess delivery of land‐based sediments is an important control on the overall condition of nearshore coral reef ecosystems. Unpaved roads have been identified as a dominant sediment source on St John in the US Virgin Islands. An improved understanding of road sediment production rates is needed to guide future development and erosion control efforts. The main objectives of this study were to: (1) measure sediment production rates at the road segment scale; (2) evaluate the importance of precipitation, slope, contributing area, traffic, and grading on road sediment production; (3) develop an empirical road erosion predictive model; and (4) compare our measured erosion rates to other published data. Sediment production from 21 road segments was monitored with sediment traps from July 1998 to November 2001. The selected road segments had varying slopes, contributing areas, and traffic loads. Precipitation was measured by four recording rain gauges. Sediment production was related to total precipitation and road segment slope. After normalizing by precipitation and slope, the mean sediment production rate for roads that had been graded within the last two years was 0·96 kg m^?2 cm^?1 m m^?1 or approximately 11 kg m^?2 a^?1 for a typical road with a 10 per cent slope and an annual rainfall of 115 cm a^?1. The mean erosion rate for ungraded roads was 42 per cent lower, or 0·56 kg m^?2 cm^?1 m m^?1. The normalized mean sediment production rate for road segments that had been abandoned for over fifteen years was only about 10 per cent of the mean value for ungraded roads. Sediment production was not related to traffic loads. Multiple regression analysis led to the development of an empirical model based on precipitation, slope to the 1·5 power, and a categorical grading variable. The measured and predicted erosion rates indicate that roads are capable of increasing hillslope‐scale sediment production rates by up to four orders of magnitude relative to undisturbed conditions. The values from St John are at the high end of reported road erosion rates, a finding that is consistent with the high rainfall erosivities and steep slopes of many of the unpaved roads on St John. Other than paving, the most practical methods to reduce current erosion rates are to minimize the frequency of grading and improve road drainage. Copyright © 2005 John Wiley & Sons, Ltd.     

Runoff and road erosion at the plot and road segment scales,St John,US Virgin Islands

Previous studies have identified unpaved roads as the primary source of erosion on St John in the US Virgin Islands, but these studies estimated road erosion rates only as annual averages based primarily on road rill measurements. The goal of this project was to quantify the effect of unpaved roads on runoff and sediment production on St John, and to better understand the key controlling factors. To this end runoff and sediment yields were measured from July 1996 to March 1997 from three plots on naturally vegetated hillslopes, four plots on unpaved road surfaces and two cutslope plots. Sediment yields were also measured from seven road segments with contributing areas ranging from 90 to 700 m². With respect to the vegetated plots, only the two largest storm events generated runoff and there was no measurable sediment yield. Runoff from the road surface plots generally occurred when storm precipitation exceeded 6 mm. Sediment yields from the four road surface plots ranged from 0·9 to 15 kg m⁻² a⁻¹, and sediment concentrations were typically 20–80 kg m⁻³. Differences in runoff between the two cutslope plots were consistent with the difference in upslope contributing area. A sprinkler experiment confirmed that cross‐slope roads intercept shallow subsurface stormflow and convert this into surface runoff. At the road segment scale the estimated sediment yields were 0·1 to 7·4 kg m⁻² a⁻¹. Road surface runoff was best predicted by storm precipitation, while sediment yields for at least three of the four road surface plots were significantly correlated with storm rainfall, storm intensity and storm runoff. Sediment yields at the road segment scale were best predicted by road surface area, and sediment yields per unit area were most strongly correlated with road segment slope. The one road segment subjected to heavy traffic and more frequent regrading produced more than twice as much sediment per unit area than comparable segments with no truck traffic. Particle‐size analyses indicate a preferential erosion of fine particles from the road surface and a rapid surface coarsening of new roads. Published in 2001 by John Wiley & Sons, Ltd.     

Erosion prediction on unpaved mountain roads in northern Thailand: validation of dynamic erodibility modelling using KINEROS2

The event‐ and physics‐based KINEROS2 runoff/erosion model for predicting overland flow generation and sediment production was applied to unpaved mountain roads. Field rainfall simulations conducted in northern Thailand provided independent data for model calibration and validation. Validation shows that KINEROS2 can be parameterized to simulate total discharge, sediment transport and sediment concentration on small‐scale road plots, for a range of slopes, during simulated rainfall events. The KINEROS2 model, however, did not accurately predict time‐dependent changes in sediment output and concentration. In particular, early flush peaks and the temporal decay in sediment output were not predicted, owing to the inability of KINEROS2 to model removal of a surface sediment layer of finite depth. After 15–20 min, sediment transport declines as the supply of loose superficial material becomes depleted. Modelled erosion response was improved by allowing road erodibility to vary during an event. Changing the model values of erosion detachment parameters in response to changes in surface sediment availability improved model accuracy of predicted sediment transport by 30–40%. A predictive relationship between road erodibility ‘states’ and road surface sediment depth is presented. This relationship allows implementation of the dynamic erodibility (DE) method to events where pre‐storm sediment depth can be estimated (e.g., from traffic usage variables). Copyright © 2001 John Wiley & Sons, Ltd.     

Catchment‐scale contribution of forest roads to stream exports of sediment,phosphorus and nitrogen

The relative contribution of forest roads to total catchment exports of suspended sediment, phosphorus, and nitrogen was estimated for a 13 451 ha forested catchment in southeastern Australia. Instrumentation was installed for 1 year to quantify total in‐stream exports of suspended sediment, phosphorus, and nitrogen. In addition, a total of 101 road–stream crossings were mapped and characterized in detail within the catchment to identify the properties of the road section where the road network and the stream network intersect. Sediment and nutrient generation rates from different forest road types within the catchment were quantified using permanent instrumentation and rainfall simulation. Sediment and nutrient generation rates, mapped stream crossing information, traffic data and annual rainfall data were used to estimate annual loads of sediment, phosphorus, and nitrogen from each stream crossing in the catchment. The annual sum of these loads was compared with the measured total catchment exports to estimate the proportional contribution of loads from roads within the catchment. The results indicated that 3·15 ha of near‐stream unsealed road surface with an average slope of 8·4% delivered an estimated 50 t of the 1142 t of total suspended sediment exported from the catchment, or about 4·4% of the total sediment load from the forest. Stream discharge over this period was 69 573 Ml. The unsealed road network delivered an estimated maximum of 22 kg of the 1244 kg of total phosphorus from the catchment, or less than 1·8% of the total load from the forest. The average sediment and phosphorous load per crossing was estimated at 0·5 t (standard deviation 1·0 t) and 0·22 kg (standard deviation 0·30 kg) respectively. The lower proportional contribution of total phosphorus resulted from a low ratio of total phosphorus to total suspended sediment for the road‐derived sediment. The unsealed road network delivered approximately 33 kg of the 20 163 kg of total nitrogen, about 0·16% of the total load of nitrogen from the forest. The data indicate that, in this catchment, improvement of stream crossings would yield only small benefits in terms of net catchment exports of total suspended sediment and total phosphorus, and no benefit in terms of total nitrogen. These results are for a catchment with minimal road‐related mass movement, and extrapolation of these findings to the broader forested estate requires further research. Copyright © 2007 John Wiley & Sons, Ltd.     

Assessing dominant factors affecting soil erosion using a portable rainfall simulator

Investigating the causes of soil erosion is difficult in natural conditions owing to the presence of other factors. Without simplifying the experimental conditions, studying soil behavior with its numerous parameters while considering factors such as vegetation cover, topography, and rainfall is difficult and in most conditions impossible. The application of simulation approaches is therefore necessary to simplify the prototype. In this research, the effects of physical soil factors such as texture and antecedent soil moisture, along with land slope and vegetation cover were evaluated in the Taleghan watershed, Iran, using a rainfall simulator and soil erosion plots. For this purpose, a 89 × 120 cm rainfall simulator producing 24.5 and 32 mm/h rainfall intensities of 30 min duration, as a common condition of the study area, was used at 144 locations over soil erosion plots with dimensions of 95 × 125 cm. Plots had slope classes of 12-20 and 20-30 %, different soil textures, different antecedent soil moistures, and medium to poor vegetation cover conditions. It was found that for 24.5 and 32 mm/h rainfall intensities, the sediment yield had high correlations of-0.771 and -0.796 with vegetation cover and slight correlations of 0.045 and 0.029 with land slope respectively. Regression equations for predicting the sediment yield were also developed for different conditions.     

Physically based modelling of sediment generation and transport under a large rainfall simulator

A series of large rainfall simulator experiments was conducted in 2002 and 2003 on a small plot located in an experimental catchment in the North Island of New Zealand. These experiments measured both runoff and sediment transport under carefully controlled conditions. A physically based hydrological modelling system (SHETRAN) was then applied to reproduce the observed hydrographs and sedigraphs. SHETRAN uses physically based equations to represent flow and sediment transport, and two erodibility coefficients to model detachment of soil particles by raindrop erosion and overland flow erosion. The rate of raindrop erosion also depended on the amount of bare ground under the simulator; this was estimated before each experiment. These erodibility coefficients were calibrated systematically for summer and winter experiments separately, and lower values were obtained for the summer experiments. Earlier studies using small rainfall simulators in the vicinity of the plot also found the soil to be less erodible in summer and autumn. Limited validation of model parameters was carried out using results from a series of autumn experiments. The modelled suspended sediment load was also sensitive to parameters controlling the generation of runoff from the rainfall simulator plot; therefore, we found that accurate runoff predictions were important for the sediment predictions, especially from the experiments where the pasture cover was good and overland flow erosion was the dominant mechanism. The rainfall simulator experiments showed that the mass of suspended sediment increased post‐grazing, and according to the model this was due to raindrop detachment. The results indicated that grazing cattle or sheep on steeply sloping hill‐country paddocks should be carefully managed, especially in winter, to limit the transport of suspended sediment into watercourses. Copyright © 2006 John Wiley & Sons, Ltd.     

Physically-based distributed soil erosion and sediment yield model (DREAM) for simulating individual storm events

Abstract

A relatively simple process-oriented, physically-based distributed (PBD) hydrological model, the distributed runoff and erosion assessment model (DREAM), is described, and a validation study conducted in the semi-forested watershed of Pathri Rao, in the Garhwal Himalayas, India, is reported. DREAM takes account of watershed heterogeneity as reflected by land use, soil type, topography and rainfall, measured in the field or estimated through remote sensing, and generates estimates of runoff and sediment yield in spatial and temporal domains. The model is based on simultaneous solution of flow dynamics, based on kinematic wave theory, followed by solution of soil erosion dynamics. As the storm rainfall proceeds, the process of overland flow generation is dependent on the interception storage and infiltration rates. The components of the soil erosion model have been modified to provide better prediction of sediment flow rates and sediment yields. The validation study conducted to test the performance of the model in simulating soil erosion and sediment yield during different storm events monitored in the study watershed showed that the model outputs are satisfactory. Details of a sensitivity analysis, model calibration and the statistical evaluation of the results obtained are also presented and discussed. It is noteworthy that the distributed nature of the model combined with the use of geographical information system (GIS) techniques permits the computation and representation of the spatial distribution of sediment yield for simulated storm events, and a map of the spatial distribution of sediment yield for a simulated storm event is presented to highlight this capability.

Citation Ramsankaran, R., Kothyari, U.C., Ghosh, S.K., Malcherek, A., and Murugesan, K., 2013. Physically-based distributed soil erosion and sediment yield model (DREAM) for simulating individual storm events. Hydrological Sciences Journal, 58 (4), 872–891.     

Spatial simulation of forest road effects on hydrology and soil erosion after a wildfire

Post-fire catchment and water utility managers throughout the world use predictive models to estimate potential erosion risks to aid in evaluating downstream impacts of increased runoff and erosion, and to target critical areas within a fire for applying mitigation practices. Erosion prediction can be complicated by forest road networks. Using novel GIS technology and soil erosion modelling, this study evaluated the effect of roads on surface runoff, erosion and sediment yields following a wildfire and determined that the predictive models were providing reasonable results. The GeoWEPP model was used to simulate onsite erosion and offsite sediment delivery before and after fire disturbance using a 2-m resolution DEM as the terrain layer. Erosion rates in excess of 4 Mg ha⁻¹ year⁻¹ were predicted mainly from steep moderate and high severity burn areas. Roads influenced surface runoff flow path distributions and sub-catchment delineations, affecting the spatial distribution of sediment detachment and transport. Roads tended to reduce estimated erosion on slopes below the roads but increases in erosion rates were estimated for road fillslopes. Estimated deposition amounts on roads and in sediment basins were similar to measured amounts. The results confirm that road prisms, culverts and road ditches influence sedimentation processes after wildfire, and they present opportunities to detain eroded sediments.     

Application of the WEPP model to determine sources of run‐off and sediment in a forested watershed

This study investigates critical run‐off and sediment production sources in a forested Kasilian watershed located in northern Iran. The Water Erosion Prediction Project (WEPP) watershed model was set up to simulate the run‐off and sediment yields. WEPP was calibrated and validated against measured rainfall–run‐off–sediment data. Results showed that simulated run‐off and sediment yields of the watershed were in agreement with the measured data for the calibration and validation periods. While low and medium values of run‐off and sediment yields were adequately simulated by the WEPP model, high run‐off and sediment yield values were underestimated. Performance of the model was evaluated as very good and satisfactory during the calibration and validation stages, respectively. Total soil erosion and sediment load of the study watershed during the study period were determined to be 10 108 t yr^?1 and 8735 t yr^?1, respectively. The northern areas of the watershed with dry farming were identified as the critical erosion prone zones. To prioritize the subwatersheds based on their contribution to the run‐off and sediment production at the watershed's main outlet, unit response approach (URA) was applied. In this regard, subwatersheds close to the main outlet were found to have the highest contribution to sediment yield of the whole watershed. Results indicated that depending on the objective of land and water conservation practices, particularly, for controlling sediment yield at the main outlet, critical areas for implementing the best management practices may be identified through conjunctive application of WEPP and URA. Copyright © 2014 John Wiley & Sons, Ltd.     

The effect of wildfire on runoff and erosion in native Eucalyptus forest

Wildfires raise concerns over the risk of accelerated erosion as a result of increased overland flow and decreased protection of the soil by litter and ground vegetation cover. We investigated these issues following the 1994 fires that burnt large areas of native Eucalyptus forest surrounding Sydney, Australia. A review of previous studies identifies the fire and rainfall conditions that are likely to lead to increased runoff and accelerated erosion. We then compare runoff and erosion between burnt and unburnt sites for 10 months after the 1994 fires. At the scale of hillslope plots, the 1994 fire increased runoff by enhancing soil hydrophobicity, and greatly increased sediment transport, mainly through the reduced ground cover, which lowered substantially the threshold for initial sediment movement. However, both runoff and sediment transport were very localized, resulting in little runoff or sediment yield after the fire at the hillslope catchment scale. We identify that after moderately intense fires, rainfall events of greater than one year recurrence interval are required to generate substantial runoff and sediment yield. Such events did not occur during the monitoring period. Past work shows that mild burns have little effect on erosion, and it is only after the most extreme fires that erosion is produced from small, frequent storms. © 1998 John Wiley & Sons, Ltd.     

Assessment of runoff and sediment yield in the Miyun Reservoir catchment by using SWAT model

The Soil and Water Assessment Tool (SWAT) was used to simulate the transport of runoff and sediment into the Miyun Reservoir, Beijing in this study. The main objective was to validate the performance of SWAT and the feasibility of using this model as a simulator of runoff and sediment transport processes at a catchment scale in arid and semi‐arid area in North China, and related processes affecting water quantity and soil erosion in the catchment were simulated. The investigation was conducted using a 6‐year historical streamflow and sediment record from 1986 to 1991; the data from 1986 to 1988 was used for calibration and that from 1989 to 1991 for validation. The SWAT generally performs well and could accurately simulate both daily and monthly runoff and sediment yield. The simulated daily and monthly runoff matched the observed values satisfactorily, with a Nash‐Sutcliffe coefficient of greater than 0·6, 0·9 and a coefficient of determination 0·75, 0·9 at two outlet stations (Xiahui and Zhangjiafen stations) during calibration. These values were 0·6, 0·85 and 0·6, 0·9 during validation. For sediment simulation, the efficiency is lower than that for runoff. Even so, the Nash‐Sutcliffe coefficient and coefficient of determination were greater than 0·48 and 0·6 for monthly sediment yield during calibration, and these values were greater than 0·84 and 0·95 during validation. Sensitivity analysis shows that sensitive parameters for the simulation of discharge and sediment yield include curve number, base flow alpha factor, soil evaporation compensation factor, soil available water capacity, soil profile depth, surface flow lag time and channel re‐entrained linear parameter, etc. Copyright © 2009 John Wiley & Sons, Ltd.     

Modelling road surface sediment production using a vector geographic information system

Field investigations indicate that unpaved roads are the largest sediment source on St John, US Virgin Islands. Cross-sectional measurements of eroded road surfaces were used to establish an empirical relationship to predict annual road surface erosion as a function of road gradient and contributing drainage area. A model (ROADMOD) for estimating and mapping average annual sediment production from a road network was developed by combining this empirical relationship with a series of network algorithms to analyse road data stored in a vector geographic information system. ROADMOD was used to estimate road surface erosion in two St John catchments with very different road densities but similar land cover, topography and soils. Unpaved roads were found to increase sediment production in the more densely roaded catchment by a factor of three to eight, and in the less-roaded catchment by a factor of 1·3–2·0. Turbidity measurements in the receiving bays of these two catchments are consistent with model predictions and observed sediment delivery processes. Although this model was developed specifically for St John, it can easily be adapted to other locations by substituting a locally derived predictive equation for road erosion. Model assumptions, limitations and potential improvements are discussed. © 1998 John Wiley & Sons, Ltd.     

Sediment transport,redistribution and storage on logged forest hillslopes in south‐eastern Australia

Post‐logging changes in catchment sediment yield have traditionally been attributed to increases in hillslope erosion and delivery rates as a result of forest harvesting activities. Linking hillslope erosion to catchment yield in forestry environments remains difficult, however, primarily because of the scarcity of data on the nature of hillslope sediment storage and delivery processes. A large rainfall simulator (350 m²) was used to apply rainstorms to a logged hillslope containing a snig track (skid trail) and a general logging or harvesting area (GHA) on 10 forest compartments in south‐eastern Australia. The experiments confirmed that the compacted, disturbed surfaces, such as roads and tracks, are the dominant sources of sediment in forestry areas. Sediment transport rates were limited by available sediment supply on both the snig track and the GHA, introducing important implications for the modelling of these surfaces using sediment transport capacity theories. Sediment delivery from the snig track to the adjacent GHA, via a cross‐bank (drainage diversion), was strongly influenced by the percentage fine fraction in the eroded sediment. Preferential deposition of coarse aggregates was measured at erosion control structures and on the adjacent GHA. Over 50% of fine‐grained material were deposited on the hillslope over a relatively short, flow path length of <5 m, highlighting the effectiveness of runoff diversion as a practice in reducing sediment flux. The transfer of water and sediment from disturbed to less disturbed parts of the landscape, and the associated potential for sediment storage, needs to be considered as part of any catchment impact assessment. Copyright © 1999 John Wiley & Sons, Ltd.     

Long-term erosion on granitic roadcuts based on exposed tree roots

Exposed roots were used to estimate soil and bedrock erosion on the cut slopes of a 45-year-old road constructed in granitic soils of the Idaho Batholith. The original roadcut surface was defined by projecting a straight line from the toe of the cut past the end of the exposed root to the intersection of a straight line projected along the surface of the hillslope. A cross-sectioning technique was then used to determine erosion to the present roadcut surface. A total of 41 exposed root sites were used to estimate erosion on a 1350 m-long section of road. Average erosion was 1·0 and 1·1 cm/year for soil and bedrock respectively. Buttressing by tree roots caused lower erosion rates for soil as compared to bedrock. Both soil and bedrock erosion rates showed statistically significant correlations with the gradients of the original cut slope. The bedrock erosion data provide a reasonable estimate of the disintegration rate of exposed granitic bedrock exhibiting the weathering and fracturing properties common to this area. The road is located in a study watershed where long-term sediment yield data are available. Sediment data from adjacent study watersheds with no roads were compared to sediment data from the roaded watershed to estimate the long-term increase in sediment yield caused by the road. The increase amounts to about 2·4 m³/year. This figure, compared to the average annual on-site road erosion, provides an erosion to sediment delivery ratio of less than 10 per cent. Based on study results, road construction and maintenance practices are suggested for helping reduce roadcut erosion.     

The effect of soil type,meteorological forcing and slope gradient on the simulation of internal erosion processes at the local scale

Numerical simulation experiments of water erosion at the local scale (20 × 5 m) using a process‐based model [Plot Soil Erosion Model_2D (PSEM_2D)] were carried out to test the effects of various environmental factors (soil type, meteorological forcing and slope gradient) on the runoff and erosion response and to determine the dominant processes that control the sediment yield at various slope lengths. The selected environmental factors corresponded to conditions for which the model had been fully tested beforehand. The use of a Green and Ampt model for infiltration explained the dominant role played by rainfall intensity in the runoff response. Sediment yield at the outlet of the simulated area was correlated positively with rainfall intensity and slope gradient, but was less sensitive to soil type. The relationship between sediment yield (soil loss per unit area) and slope length was greatly influenced by all environmental factors, but there was a general tendency towards higher sediment yield when the slope was longer. Contribution of rainfall erosion to gross erosion was dominant for all surfaces with slope lengths ranging from 4 to 20 m. The highest sediment yields corresponded to cases where flow erosion was activated. An increase in slope gradient resulted in flow detachment starting upstream. Sediment exported at the outlet of the simulated area came predominantly from the zone located near the outlet. The microrelief helped in the development of a rill network that controlled both the ratio between rainfall and flow erosion and the relationship between sediment yield and slope length. Copyright © 2010 John Wiley & Sons, Ltd.     

A quantitative study of sediment delivery and stream pollution from different forest road types

The objective of this paper is to quantify, and enable the prediction of, sediment delivery and water pollution impacts from a spectrum of forest roads. Ten 100–200 m long sections of forest road were selected to incorporate a wide range of the key physical site factors that are likely to affect the rate of sediment generation. Each road section was permanently instrumented for 1 year to measure rainfall and runoff continuously. Suspended load, bedload, and traffic were integrated measurements over 2‐ to 3‐week site‐service intervals. Total annual sediment load (normalized for slope) varied about 25‐fold, from 216 mg m^?2 per millimetre of rain for a high‐quality gravel surfaced road with minimal traffic to 5373 mg m^?2 per millimetre of rain for an unsurfaced road on an erodible subsoil with moderate light‐vehicle traffic. For the seven gravel‐surfaced roads in this study, truck traffic (axles/week) explained 97% of the variation in annual sediment delivery (per unit of rainfall) from the road. Equations are proposed that allow annual sediment delivery rates to be estimated when net rainfall, road slope, road area, and truck traffic are known. Roads produce runoff rapidly and were found to deliver sediment for about the same duration as rainfall is falling, in this study varying between 5 and 10% of the time. The patterns of sediment delivery measured from the experimental roads (frequency, duration, and intensity) in this study are similar to levels that have been shown to alter the composition of in‐stream macroinvertebrate communities in small (e.g. <10 l s^?1), clean, mountain streams. However, in larger well‐mixed streams (e.g. >500 l s^?1), dilution is sufficient to prevent concentrations reaching critical levels that are likely to result in biological impacts. Copyright © 2006 John Wiley & Sons, Ltd.     

Integrated erosion control measures and environmental effects in rocky mountainous areas in northern China

Based on observations of runoff plots and field investigations of gully cross-sections, impacts of various soil and water conservation measures on runoff and sediment yield are analyzed for different rainfall conditions. The results show that antecedent rainfall and rainfall intensity are the main factors affecting the runoff and soil erosion processes. Rainfall events with antecedent rainfall can produce high runoff and sediment yield. Large differences in the characteristics of two rainfall events will result in greater variations of total runoff and sediment yield from the same runoff plot. Under the same soil control measure and rainfall condition, soil and water conservation measures can reduce the impacts of antecedent rainfall and rainfall intensity on runoff and soil erosion. Among various measures, level terrace seems to be the greatest for soil conservation purposes. Combining with engineering measures,Vegetation measures is also effective in controlling runoff and soil erosion. In the initial stage of vegetation enclosure measures, engineering measure is necessary to improve the environment for ecological recovery. Gully head protection can control gully erosion effectively, but the effectiveness of gully head protection would be reduced when rainfall intensity increases. Therefore, the design of a gully head protection structure must be based on local hydrological conditions.     

Runoff and suspended sediment yields from an unpaved road segment,St John,US Virgin Islands

Unpaved roads are believed to be the primary source of terrigenous sediments being delivered to marine ecosystems around the island of St John in the eastern Caribbean. The objectives of this study were to: (1) measure runoff and suspended sediment yields from a road segment; (2) develop and test two event‐based runoff and sediment prediction models; and (3) compare the predicted sediment yields against measured values from an empirical road erosion model and from a sediment trap. The runoff models use the Green–Ampt infiltration equation to predict excess precipitation and then use either an empirically derived unit hydrograph or a kinematic wave to generate runoff hydrographs. Precipitation, runoff, and suspended sediment data were collected from a 230 m long, mostly unpaved road segment over an 8‐month period. Only 3–5 mm of rainfall was sufficient to initiate runoff from the road surface. Both models simulated similar hydrographs. Model performance was poor for storms with less than 1 cm of rainfall, but improved for larger events. The largest source of error was the inability to predict initial infiltration rates. The two runoff models were coupled with empirical sediment rating curves, and the predicted sediment yields were approximately 0·11 kg per square meter of road surface per centimetre of precipitation. The sediment trap data indicated a road erosion rate of 0·27 kg m^?2 cm^?1. The difference in sediment production between these two methods can be attributed to the fact that the suspended sediment samples were predominantly sand and silt, whereas the sediment trap yielded mostly sand and gravel. The combination of these data sets yields a road surface erosion rate of 0·31 kg m^?2 cm^?1, or approximately 36 kg m^?2 year^?1. This is four orders of magnitude higher than the measured erosion rate from undisturbed hillslopes. The results confirm the importance of unpaved roads in altering runoff and erosion rates in a tropical setting, provide insights into the controlling processes, and provide guidance for predicting runoff and sediment yields at the road‐segment scale. Copyright © 2006 John Wiley & Sons, Ltd.     

Unpaved rural roads as source areas of sediment in a watershed of the Brazilian semi-arid region

Approximately 80% of the road network in Brazil is unpaved and shows evidences of a high erosion potential. In the semi-arid Caatinga Biome in the northeast of the country, a monitoring programme has been done for two years in order to analyze runoff and sediment production from unpaved rural road-ways and from embankments. Sediment production ranged from 0.30 to 0.92 Mg/ha yr, higher than in undisturbed areas, but generally lower than that reported for unpaved roads in other regions. However, this is a semi-arid area with low rainfall and runoff and, hence, with a limited hydrological connectivity and sediment production. Sediment production on an embankment with no vegetation was around ten times higher than on an embankment with vegetation. On the road surface, annual sediment production (normalized for gradient) in a section with traffic was three times higher than for a road surface without traffic. In addition, events that occurred after roadway maintenance activities generated peaks of sedi-ment concentration of over 5000 mg/L. These results suggest that sediment production from roads and embankments with bare surfaces is at least one order of magnitude higher than in undisturbed catch-ment areas. Maintenance activity and vehicle traffic contribute to an increase in sediment availability and impact on the sediment concentration, but less intensely on sediment loads, which depend on the runoff magnitude of the events occurring after roadway maintenance. It was also found that the natural vegetation of the semi-arid region potentially captures sediment on roadway embankments;thereby playing an important role in breaking connectivity between the sediment flow from unpaved roads and the natural drainage system of the catchment.