a rough, simple formula to estimate unknow % of particle could be the follow: (text taken from the SoilPar 2 software help). It is no more than a quite stupid interpolation, the true carachteristics of the soil are not considered.

The interpolation criteria is based on a assumption of log-normal distribution of the particle size. The cumulated distribution is computed after a logarithmic interpolation between the two values closest to the requested values is performed. The equation used is the following:

mass = (y2 - y1) / (Log(x2) - Log(x1)) * Log(x) - (y2 - y1) / (Log(x2) - Log(x1)) * Log(x1) + y1

mass = % of mass from 0 to the diameter in which interpolation is required

x = diameter to interpolate

x1= % of mass in the range from 0 to the available diameter lower than x and closest to x available

x2= % of mass in the range from 0 to the available diameter greater than x and closest to x available

y1,y2 = particle size diameters corresponding to x1 e x2

I repeat again: it is better to do again the analysis, but, if is not possible, the previous formula is a possibility.

Hi, I think it would be possible only if you have the data on the cumulative soil texture curve available (from 0 to 2000 um). Then you would be able to estimate any fraction. Up till now I do not know about any formula, using which you would be able to recalculate the fraction size distribution to other fraction size distribution. So I am also curious if somebody will give you some suggestion.

Good luck!

Elena

I agree with Elena. You can predict a distribution of soil particles if you have a group of observed distributions to base your prediction on. If you are investigating a well-studied soil type, this information probably already exists. If not, I suggest filtering the 2-20 um fraction out of ~10% of your samples and using this information to form your prediction on the remaining unfiltered samples.

It coulb be estimated from the distribution curve of the size particles if you have used a laser system or one like that, one of the equipment most commonly used for this determination is the Malvern. If you have determined the fractions showing by the pipette methods then it will be difficult to estimate the missed fraction, or you decide to do a further determination to get the fraction missed. Anyway the pipette method is less accurate, and in any case all depends on the purpose for which the data will be used.

Hi, I don't understand um (2-20 um in the question)? please explain more about the question that help me for answer the question?

um=micrometer (1m = 1 000 000 um)

I have results of a soil analysis with the percentage of clay (0-2um), silt (2-50um) and sand (>50um) made by pipette method. From that data, is it possible to make a rough estimation of the fraction of silt in the range 2-20um?

Option 1 is redo the PSD analysis !

option 2 (a very rough method) is to do a log interpolation (assuming a log normal distribution). The SoilPar2 software do this when asked to convert between textural classification systems, but, I repeat, is a rough approximation. Other kind of interpolation and methods based on neural network and similarity criteria was developped by A. Nemes.

@Elena, I think there is no formula for particle size distribution, but you have to prepare a grain size distribution curve -- on vertical axis keep % finer then and horizontal axis size of particle range of your interest. ... in which material passed from any sieve size weigh it note down, then passed from next smaller size note and so on...that weight in % versus size graph constructed. Coincide range of different sizes of textures to horizontal axis .... now take median point of particular texture, draw vertical line to cross curve from this point draw horizontal line which crosses vertical axis it shows % of that texture, same way find % of other textures presents in your sample.

It is not possible to delineate soil mineral fractions as the soil is a heterogenous system which varies significantly in particle size distribution and other parameters. It is a different matter for pure systems which are more or less static to a greater extent.

If your purpose to estimate to estimate the 2-20um fraction for your publication, I do agree with Elena suggestion, Its called pedotransfer approach. The data on the cumulative soil texture curve available (from 0 to 2000 um), should be representative to cover your study areas. As far as I know, currently there is no pedotransfer equation to answers your question. If this approach to complicated then is better to direct measurement using the standard pipette method, that take only one or three day(s).

Yes it is possible to separate 2-20um fraction with help of particle size analysis by sieve method.

From your question, I understand that you do not have a proper sieve to separate 21-50 um particles. that's why you want to know whether there is any method to bypass this. The answer is "No". Soil component varies from layer to layer and you cannot just estimate it. The best you can do is to spread the soil by shaking it cautiously and then look into the finer fraction and estimate a percentage.

I agree with the opinion of redoing the analysis by pipette there is no other option

There is no other option than to reanalyse the partcile size distribution. For estimate, you may use hydrometer method for getting the size data when teh particles begin to settle down.

you can find all indications in the PhD thesis titled "Efficient Methods for Predicting Soil Hydraulic Properties", by Budiman Minasny, you can download it from the following website http://ses.library.usyd.edu.au/handle/2123/853

a rough, simple formula to estimate unknow % of particle could be the follow: (text taken from the SoilPar 2 software help). It is no more than a quite stupid interpolation, the true carachteristics of the soil are not considered.

The interpolation criteria is based on a assumption of log-normal distribution of the particle size. The cumulated distribution is computed after a logarithmic interpolation between the two values closest to the requested values is performed. The equation used is the following:

mass = (y2 - y1) / (Log(x2) - Log(x1)) * Log(x) - (y2 - y1) / (Log(x2) - Log(x1)) * Log(x1) + y1

mass = % of mass from 0 to the diameter in which interpolation is required

x = diameter to interpolate

x1= % of mass in the range from 0 to the available diameter lower than x and closest to x available

x2= % of mass in the range from 0 to the available diameter greater than x and closest to x available

y1,y2 = particle size diameters corresponding to x1 e x2

I repeat again: it is better to do again the analysis, but, if is not possible, the previous formula is a possibility.

Dear Gabriel, statistical methods for predicting particle size distributions are what they are, an educated guess. If you are serious about your research you should measure them directly.

Here's an approach I've used in the past to estimate mass fractions which weren't measured:

Skaggs, T.H., Arya, L.M., Shouse, P.J., and Mohanty, B.P.; 2001. Estimating Particle-Size Distribution from Limited Soil Texture Data. Soil Sci. Soc. Am. J. 65: 1038-1044

According to the paper, you need to know three mass fractions and your silt content should be below 70%.

It is not possible to do that by sieving as before indicated!!!

But it is possible to do that by sedimentation...; the only problem is the calculation of time, because it is different if you have either round particles or planar ones. But consulting references as our neighborg Pfr. NUNES indicated, you can obtain a good approximation, at least for publication. I have some papers about, in relation physical fractionation of SOM, I do not remeber now where!!!

My suggestion is to redo the analysis and calculate time of partials fall according to Stokes law equation

Exccellent and ethical suggestion from Eman Aljanabi and I thank him for it. I suggest the initiator of the topic follow the advice without further waste of time and energy.

My be you can shaking your soil sample with sodium hexamytafosfat and then precipitated by the method of texture..

This topic highlights a possible source of important and widespread mistakes. We all know of the International Society of Soil Science size boundaries (2-20-2000 um for clay-silt-sand) and the United States Department of Agriculture size boundaries (2-50-2000 um for clay-silt-sand). But it is hard to find a texture triangle for ISSS sizes; we almost invariably see the USDA texture triangle. I believe many persons have ISSS size data and mistakenly use the USDA texture triangle or mistakenly infer ISSS size fractions from the USDA texture triangle. I suspect the questioner has realised this. I have an ISSS texture triangle which I take every opportunity to share and I can scan it to whoever requests. Alternatively, I believe you can find it in one of T.J. Marshall's older manuals/ books.

Thank you very much Mr. Peter Vine for the crystal clarity with which you exposed flaws in soil particle size distribution analysis. The discussion over the topic for me appears as unnecessary and it is better to adopt standard and approved soil test for particle size and distribution analysis.

Probably yuo finde moire informtion in: León-Arteta R. 2001. Comparación de métodos de análisis mecánico de suelos. Terra. Vol. 19 No 3 pp. 219-225

And in Agronomy 9. 1965 in Day R.P. work with pippete and Hydromether