Temperature affects fluid viscosity, but typically you have to change viscosity by an order of magnitude to see significant effects. From zero to fifty Celsius only gives you a change in viscosity by a factor of 3x, while this sounds large it will likely not be relevant in hydraulic applications for common water temperatures found in the earths environment.

Thanks for the answer dear Dr. Fritz, you mean that increasing temperature doesn't lead to a significant sediment movement threshold. what about decreasing temperature, does it lead to a change? since in environment we mostly have cold weather rather than warm.

I disagree with Hermann Fritz! The kinematic viscosity (which is here the relevant parameter) is very sensitive to changes in water temperature. If the temperature range is small, then the effect is not large enough to bother about. However, seasonal mid- to high-latitude water temperatures (e.g. in the Wadden Sea) fluctuate by >20°C and then the effect is very large, changing settling velocities of sediment particles up to more than 30% in dependence of particle size (the larger the grain size, the lower the effect). Expressed in environmental terms, the shore-normal location of sediment particles in the Wadden Sea which have the same settling velocity in winter (high kinematic viscosity) and summer (low kinematic viscosity) lie as much as 3 km apart!

You can download the following paper from my publications in ResearchGate:

Krögel, F., Flemming, B.W. (1998) Evidence for temperature-adjusted sediment distributions in the backbarrier tidal flats of the East Frisian Wadden Sea (southern North Sea). SEPM Spec. Publ. 61, 31-41.

For the question to be answered correctly we would need to know what the median grain diameters of the sediments in question are. If we are talking about sediments that are 63 micron or greater there is very little difference. However if you are talking fine silts and clays the difference becomes much larger with decreasing temperature and grain size, and not in a linear fashion. There are other factors to consider as well such as water movement; convection currents, flow, wind action, wave action (which are not the same).

For equations and demonstrative graphs, please see:

http://kleene.er.usgs.gov/sdct/images/e/ed/Bull_1181_a.pdf

Specifically pages 24-26

In my opinion the question is badly formulated: Shields’ diagram, as it is, does not change at all because it was derived for incoherent grains on which temperature has no effect at all. Obviously, if you change T, the kinematic viscosity changes and the grain Reynolds number changes (let us call it X, according to Yalin), so that the threshold that you read on the Shiedls diagram changes (see Pilotti M., Menduni G., Beginning of sediment transport of incoherent grains in shallow shear flows, Journal of Hydraulic Research, IAHR, 39, 115-124, 2001). Burghard is right on this point. Now you may wonder when this effect is relevant. Here comes the answer by Daniel, who rightly observes that it depends on the diameter of the grains. Actually, this is a detectable effect only when the threshold itself is a function of X, that is when X < 100. Moreover, when sediments are very small (clay) they are almost inevitably cohesive and then T has an additional effect on erosion that, however, is not represented on Shields diagram.

I agree with professor Pilotti, temperature change affects the particle Reynolds number in Shields diagram due to change in viscosity of water. However the more important issue is related to the kind of sediment which is cohesive or non-cohesive. Shields diagram developed for non-cohesive sediment. For cohesive sediment effect of temperature is more tangible.

Good discussion, guys! The temperature effect, however, is not only relevant at particle sizes below 63 microns, as suggested by Daniel, but is still tangible up to at least 1 mm, from there rapidly decreasing with increasing size. A look at the Rouse diagram (Rouse 1937) clearly shows this, and also a diagram showing the winter vs summer temperature effect in the Wadden Sea. I attach a modified version of the diagram to this message (redrawn by myself for teaching purposes). It will clarify much of the controversy. Indeed, we should really get away from geometric particle sizes and either use particle settling velocity per se (or a dimensionless form of it) or transformed settling diameters (with quartz spheres as standard). I have been using the latter ever since I constructed a settling tube for my PhD in the 1970s.

In the discussion we should also distinguish between erosion and deposition, which have different threshold conditions. For example, the erosion threshold for cohesive sediments depends strongly on the water content, whereas suspended particles settle out once the settling velocity exceeds the vertical velocity component of a current, irrespective of whether we are dealing with single mineral grains or aggregates, although concentration will at some level begin to affect the settling behaviour (hindered settling).

Sorry, the Rouse diagram was removed on the first occasion. It seems the system can only handle one file at a time. So, here is the Rouse diagram.

Burg

Another point to keep in mind are the log-scales along the two axes of the Rouse diagram!

Burg

Thanks for the discussion all dears. Why we need to consider sediment settling velocity as long as we focus on bed particle movements (threshold)?

I found out from the answers that temperature does not affect considerably on non-cohesive soils and the Shield's diagram is for the non-cohesives so the diagram won't change.but temperature may affect on both water and soil particle which this multi-effect may change what it is as shield's diagram.

It is a common misconception to believe that bedload transport of particles has nothing to do with the settling velocity of particles. As we are operating within a hydraulic system, a hydraulically normalized particle size will be superior to a geometric one! This is achieved by using equivalent settling dimaters. The relevance becomes immediately clear when dealing with oddly shaped particles, e.g. bioclastic sediment, in which case sieve diameters (i.e. geometric particle size which, incidentally, is also used in the dimensionless grains Reynolds number) overate particle sizes by as much as 60% when compared to equivalent settling diameters. The attached diagram, where the sieve grain size distribution of a bioclastic sediment sample is compared to the equivalent settling dimater distribution of the same sample, proves the point. You can read more about this whole issue in the article: Flemming, B.W. (2007). The influence of grain-size analysis methods and sediment mixing on curve shapes and textural parameters: implications for sediment trend analysis. Sedimentary Geology 202: 425-435 (cf. my publication list in ResearchGate).

Was the original question to do with the behavior of sediment in nature or in the laboratory? Other effects on the erosion threshold of deposits in the field may be mitigated through biological action, such as bioturbation rates and rates of mat formation by diatoms and cyanobacteria. An interesting question from this perspective since exposed intertidal sediments can vary in temperature by tens of degrees as they warm up after inundation ...a few thesis worth of work here probably!

Dr. Paterson, question was actually focused on temperature and sediment movement threshold but yes if we look at the problem more precisely it has worth of more studies.

Dr Flemming, Is it right to say "the more viscosity, the more shear stress production, so when temperature is higher water viscosity decreases furthermore it leads to less erosion and the threshold gets more in quantity and vice versa."?

I would formulate this as follows:

"The higher the kinematic viscosity (i.e. the ratio between density and the molecular viscosity of the water), the greater the shear stress at a constant shear velocity. Thus, because the kinematic viscosity decreases with increasing water temperature, the higher will be the critical shear stress required to erode a non-cohesive sediment."

I also wish to support the comment of David Paterson that biological activity, especially microbial and microphytobentic activity, can strongly influence the erosion behaviour of a sediment surface. Because such activity is stronger in summer (at higher temperatures), this further adds to the greater stability of a sediment surface in the summer season. As in the case of non-cohesive sediments, this bilogical effect also decreases as water temperature decreases towards the cold winter months. One must keep in mind, however, that the degree of the two influences depends very much on where you are on earth (e.g. tropics or high latitudes).

Dr Flemming, so data in higher temperatures go upper than Shields diagram I mean above the diagram and other diagrams may need to be provided in every temperature. And Shield's diagram would be useless and in any temperature we need to check for the movement or settlement of sediments by our own computations. Is that right?

Some clarifications:

The Shields diagram is a plot of the grain Reynolds number (u*D/v, where u* is the shear velocity, D the mean or median particle size, and v the kinematic viscosity) against the Shields parameter (which, incidentally, also includes u*). Because both u* and v change with temperature, the grain Reynolds number takes care of temperature changes. A change in temperature will thus result in a different grain Rynolds number for the same grain size. On the Shields diagram this results in a sideways shift along the curve representing the Shields criterion until the appropriate grain Reynolds number is reached.

However, what does the Shields criterion for the initiantion of sediment transport actually mean or, expressed differently, what is the state of the bed at the Shields criterion? Most people don't know this because this was not clear for a long time. Shields, unfortunately, did not provide a description of the bed state to which his criterion appied. It was only after van Rijn repeated the expiremt in the early 1990s using a modern flume setup that this question was resolved (Van Rijn, L.C., 1993. Principles of Sediment Transport in Rivers, Estuaries and Coastal Seas. Aqua

Publications, Amsterdam, The Netherlands.). As it turned out, the Shields criterion applies to the state where the entire bed is in motion. Van Rijn also describes six steps leading up to full bed motion, beginning with individual particles moving in some places through many particles moving in many places. The criterion for the initiation of sediment transport is thus a matter of definition. With respect to Shields, much happens at the bed long before the criterion is reached and, from a sedimentological point of view, this can be relevant provided a subcritical state persists for an extended period of time because a formerly undisturbed bed will begin to change its appearance long before the Shield criterion is reached. I attach a Shields diagram (modified after van Rijn 1993) with lines representing shear velocity (valid for 10°C) and grain size superimposed.

Hamed, what exactly is the nature of the sediment that you are concerned with? Location, grain size, hydrologic setting, etc. We are exploring a world of possibilities that may not even apply to your situation.

Dear Daniel, I wanted to know if we can find a general physics rule about it.

generally if temperature goes higher or lower how does it affect sediment movement?

actually how does it affect on critical shear stress and stability of the river bed.

as Dr Flemming mentioned: "The kinematic viscosity decreases with increasing water temperature, the higher will be the critical shear stress required to erode a non-cohesive sediment".

so it looks simple but it has said hear that it doesn't effect cohesive sediments and what was the scientific reason I still don't know.

Maybe the rule is not general, is it?

Real world studies of this are very difficult to conduct on this topic because the temperature in an uncontrolled natural setting is constantly changing, as are the velocities, and flow vectors. But I did find an in situ flume test study of stream bank erosion that looks at several aspects including temperature, suspended sediment, conductivity, and cross-sectional flow velocities.

It may be of interest to you.

http://ascelibrary.org/doi/full/10.1061/%28ASCE%290733-9437%282007%29133%3A3%28256%29

The general rule is that as viscosity increases, erosive power increases (if flow velocity remains constant at the bed interface), but numerous modifiers come in to play.

To answer the question, a simple calculation can be done using two formulas. One is the function of viscosity varying with temperature, as given in

http://ascelibrary.org/doi/abs/10.1061/(ASCE)HY.1943-7900.0000322

The other is the Shields number varying with the shear Reynolds number, as given in

http://ascelibrary.org/doi/abs/10.1061/(ASCE)HY.1943-7900.0000375

Article Evaluation of Flow Resistance in Smooth Rectangular Open Cha...