Initial processing and determining Mud, Sand and Gravel proportions

Sediment samples are first weighed, and then wet sieved over 63 um and 2000 um sieves to determine the percent (%) content of Gravel (> 2 mm diameter), Sand (2-0.063 mm) and Mud (< 0.063 mm). The results are presented as Mud Sand and Gravel percent (%).

Gravel and Sand fractions from the wet sieving are put into an oven at 40deg C overnight or longer to dry. The wet sediment samples, representing a discrete proportion (%) of the bulk sample, from the Mud fraction (< 63 um diameter) are then measured in one of two Malvern Mastersizer particle size analysers for measurement of the fine fraction. Each sample is analysed three times, and the average from these are used as the result for the mud fraction. Because the Mastersizer particle size data are created in 105 individual size bins, these data are re-assigned to the appropriate size classes used in the Wentworth scale, and the results are presented for the 0-250 um size range. This is because the laser measurement system produces data that is based on spherical particles, yet natural minerals and biogenic particles can have a very non-spherical form. Repeat analyses of multiple samples of < 63 um sieved sediment at the Pal Sed lab generally results in grain size data for sizes of up to 250 um diameter particles being produced. The laser measurements are undertaken on the volume of particles. However, because the Mud sample was originally weighed, these data represent the x % by mass of the original bulk sample, determined in the initial stage by wet sieving. Therefore these data can be summed to the outputs of Sand and Gravel % by weight to give baseline grainsize data for the Wentworth scale size bins.

Once the wet sand and gravel fractions are dry, these are then dry-sieved over sieves that match as close as possible the Wentworth scale. Sieve sizes are 63, 125, 250, 500, 1000, 2000 um for the sand fraction, and 2000, 4000, 8000 and 16000 um size for the gravel fraction. A single analysis is undertaken for the sand, and for the gravel fractions. The results are the proportions by mass as a percent of the sand and of the gravel fractions. Once the measurements are completed, and Quality Assurance duplicate analyses are undertaken on each of the Mud, Sand and Gravel fractions for approximately 1 in 10 of the samples submitted in each job, the samples are archived dry, and sent to the Repository.

Quality Assurance

Approximately 1 in 10 samples in a batch of samples submitted are analysed in duplicate (i.e. replicates). For the Mud fraction, this means that the duplicate sample will also undergo three individual analyses, and the result for the duplicate is the mean of the three analyses. For the coarse fractions, a single sieve analyses is undertaken for the duplicate, the same as for the original. The full grain-size distribution is produced, and sediment distribution plots are created to compare duplicates (Quality Assurance or QA samples) with the original samples.

Outputs of the new grain size method

The determination of Mud, Sand and Gravel proportions is the first step in creating a grain size distribution.

The proportions of grain size distributions in individual bins originating from Mud, Sand and Gravel proportions are summed to 100 %. For the Mud measurements made using the Mastersizer, this means that the determined size values in the range of 0-250 um is added to the sand fraction because they each are a % proportion of the total. Once the grain size distribution is determined, the Microsoft Excel-based software Gradistat (Blott and Pye, 2001) is used to calculate the Geometric graphical measures of Mean (MG, Standard deviation (= Sorting, σG), Kurtosis (KG) and Skewness (SkG) originally described by Folk and Ward (1957).

Additionally, to provide summary point information for each sediment sample, the Folk textural classification for coarse sediment (Folk, 1980) is determined for each sample, derived from the percentage proportions of Mud, Sand, and Gravel. Finally, the Shepard textural classification for fine sediments (Clay, Silt and Sand) is derived for each sample in MARS that has the full grain size distribution. At present, both the Folk and (soon) Shepard textural classes are derived when the final grain size distributions are entered into MARS. This is because these are products derived from data, and are not raw data. Thus the final sign-off of sediment data for release will be after data is entered into MARS.