Hi

here a milestone about all statistical parameters relating to grain size of sediments

good job!

regards

Giovanni

Dear Giovanni

Thanks very much

regards

daudi

Hi Daudi,

you're welcome!

I studied the sediments into the Posidonia oceanica meadows in the Mediterranean Sea; you can look also at my works for sedimentological aspects

I take this opportunity to ask you two questions:

1) what seagrass species lives on sampled sediments?

2) have you find monospecific or polyspecific meadows?

regards

Giovanni

Thanks Giovanni

They are generally monospecific, dominated by Thalassia hemprichii 

regards

Hi Giovanni

Unfortunately I am having trouble downloading Folk e Ward, 1957.pdf, can you please send it as an attacehment to [email protected]

regards

Daudi

sent!

Good reply by giovanni. If still in trouble you have all the gelogists at IMS.

I suggest you to use Tal for Win that is very simple software bringing out the texture of your sediment.

All the best

Take care, most of the software find percentile size (to enter F&W1957) trough interpolation in linear way. F&W are "graphical" parameters and interpolation must be done on a log-probabilistic scale (Excel han not this). I posed this problem to Blott & Pye. They agree and told me they were changing the program, but I don't know if they did it. Erros in Mz is approximateli half of the sieves interval, which for fine sand is a relevant error.

Thanks to Enzo and others. The most important point is not the wonderful graphs you get, but how you interpret it and come to the right conclusion, good comparison of sampling sites' etc. Good luck!

Daudi,

if few samples you can send me fraction weight and I send you back the data you need and some comments to the differences. Unfortunately our program doesn't work on new computers and we nevver hasd time to update it.

So you've sieved the sediments and now have the retained masses for each size class. Good. Be sure to sum the masses for each size class and compare to the initial mass; it should be within 2%. Now calculate the % retained in each size class as mass retained / sum(mass retained)*100. Now, for each size class, you calculate the percent passing as 100 - sum(% retained of the prior size classes); i.e., the % passing should cumulatively decrease from the largest to smallest sieve diameter.

This "percent finer" is plotted on the y-axis in linear scale, and the size class is plotted on the x-axis in log scale. The median diameter (D50) is the diameter at 50 percent finer. The D16 is the diameter at 16 percent finer, etc. You can read these values directly from the plot, or do a linear interpolation (the bulk of the distribution tends to plot linearly, say, between 20 and 80 percentiles). As mentioned before, a number of statistical measures can be derived graphically. There are other methods (e.g., the method of moments), but the graphical measures are pretty standard and give similar values if the particles follow a log-normal probability distribution.  

I've attached Folk's "Review of Grain Size Parameters", but keep in mind that the parameters use grain sizes in phi (phi = -log2(diameter)). 

Thanks for your contributions. I am a micro-paleontologist, however i found your responses quite useful.

I cannot see any purpose in calculating an average graisize. It would be more useful and easier to construct histograms for each site so that you can compare the grain size distributions.

Sometime you need numbers! In any case distributions can be compared through statistical parameters.

I don't compare average grain size as there no point to do that. You can put the grain size in percentage and compare between sites. Or from the data you collected convert into sediment category like muddy sand, fine sand etc. There are several protocols you can refer to like Udden 1914, Friendman and Sanders 1978, GRADISTAT program and many more. Good luck.