<P> Some of the largest antidunes on Earth are formed by turbidity currents . One observed sediment - wave field is located on the lower continental slope off Guyana, South America . This sediment - wave field covers an area of at least 29 000 km at a water depth of 4400--4825 meters . These antidunes have wavelengths of 110--2600 m and wave heights of 1--15 m . Turbidity currents responsible for wave generation are interpreted as originating from slope failures on the adjacent Venezuela, Guyana and Suriname continental margins . Simple numerical modelling has been enabled to determine turbidity current flow characteristics across the sediment waves to be estimated: internal Froude number = 0.7--1.1, flow thickness = 24--645 m, and flow velocity = 31--82 cm s . Generally, on lower gradients beyond minor breaks of slope, flow thickness increases and flow velocity decreases, leading to an increase in wavelength and a decrease in height . </P> <P> The behaviour of turbidity currents with buoyant fluid (such as currents with warm, fresh or brackish interstitial water entering the sea) has been investigated to find that the front speed decreases more rapidly than that of currents with the same density as the ambient fluid . These turbidity currents ultimately come to a halt as sedimentation results in a reversal of buoyancy, and the current lifts off, the point of lift - off remaining constant for a constant discharge . The lofted fluid carries fine sediment with it, forming a plume that rises to a level of neutral buoyancy (if in a stratified environment) or to the water surface, and spreads out . Sediment falling from the plume produces a widespread fall - out deposit, termed hemiturbidite . </P> <P> Prediction of erosion by turbidity currents, and of the distribution of turbidite deposits, such as their extent, thickness and grain size distribution, requires an understanding of the mechanisms of sediment transport and deposition, which in turn depends on the fluid dynamics of the currents . </P> <P> The extreme complexity of most turbidite systems and beds has promoted the development of quantitative models of turbidity current behaviour inferred solely from their deposits . Small - scale laboratory experiments therefore offer one of the best means of studying their dynamics . Mathematical models can also provide significant insights into current dynamics . In the long term numerical techniques are most likely the best hope of understanding and predicting three - dimensional turbidity current processes and deposits . In most cases there are more variables than governing equations and the models rely upon simplifying assumptions in order to achieve a result . The accuracy of the individual models thus depends upon the validity and choice of the assumptions made . Experimental results provide a means of constraining some of these variables as well as providing a test for such models . Physical data from field observations, or more practical from experiments, are still required in order to test the simplifying assumptions necessary in mathematical models . Most of what is known about large natural turbidity currents (i.e. those significant in terms of sediment transfer to the deep sea) is inferred from indirect sources, such as submarine cable breaks and heights of deposits above submarine valley floors . Although during the 2003 Tokachi - oki earthquake a large turbidity current was observed by the cabled observatory which provided direct observations, which is rarely achieved . </P>

Turbidity currents are thought to have aided in the formation of