This sampling method is designed to ensure that the collected sand is a random, representative sample of the stream (Folk 1974:15-16):
A channel made by recent stream flow is chosen for sampling. In smaller drainages, a single channel with low banks is usually present within the arroyo bottom. In larger washes, there may be multiple channels present. In each case, a sketch is drawn illustrating the relationship of the sampled channel to the wash.
A plastic tarp is laid out in the channel, and sand is collected from a transect perpendicular to the channel flow and placed on the tarp. If the sampled channel is small (<~4 m wide) a trench one shovelful deep and ~ 20 cm wide is dug across the transect. For larger channels, a shovelful is collected every 25 cm to 30 cm. Some of the largest channels (> 15 m wide) require sampling a shovelful of sand every 50 cm to 60 cm.
The sand is mixed thoroughly and uniformly by shovel.
The resultant sand pile is quartered and subquartered until a pile of approximately 2-3 kg remains.
The quartered section is screened through a 2 mm mesh screen into a collecting pan to remove all material coarser than sand size. The coarse material is examined, its grain sizes and lithology are recorded, and it is discarded. A Munsell color is recorded for the screened sand and it is placed in a soil sample bag. If it does not fit in the bag, it is quartered again and then collected. Sand in the bottom of the collecting pan is not simply discarded, since size sorting could occur during screening. In some cases, a close-up photograph is taken of the coarse and fine fractions to show relative colors, grain size, and lithology.
The geomorphology of the collection site is recorded, especially noting the degree of soil development and the cohesiveness of the drainage's banks. A photo is taken of the wash to further document its size and shape and the relationship of the sampled transect to the overall wash.
The sample location is plotted on a USGS 7.5' topographic map.
Sand Sample Preparation and Thin Sectioning
The collected sands are split and cleaned in the lab before being used for hand samples or thin sections. (The "hand sample" is loose sand viewed with the eye, hand lens, or binocular stereomicroscope. The "thin section" is sand mounted on a microscope slide for examination through a petrographic microscope.) To prepare the samples, each one is repeatedly halved using a "riffle" style sample splitter until it is small enough to fill a 30 dram vial, about 130 grams. The sample splitter is used to ensure that the small samples are random representatives of the larger samples without sorting or settling biases.
Once a 130 gram sample is obtained, each sample is washed in a 10 percent HCl solution for 15 to 30 seconds to remove caliche coatings from the grains. This is not long enough to remove limestone or caliche grains from the sample. The sample is then rinsed at least three times to remove all HCl, and wet screened in a .075 mm sieve to remove grains smaller than silt size. This sieve is slightly larger than the accepted sand-silt break at .0625 mm; however, most samples are so coarse that little information is lost.
Washed and screened samples were placed into jars and oven dried. Once dry, they are split again (using the sample splitter) down to a size appropriate for thin section preparation (approximately 1 tablespoon, or 25 g). The remainder of the washed sand is retained for use as a hand sample. The remaining unwashed sand is placed into storage in "Tyvek" bags.
To prepare thin sections, each sand sample is mixed with epoxy and allowed to set into a small block. (Ice cube trays are used for this purpose.) Once the blocks harden, they are treated as rocks and thin sectioned. All thin sections are etched with hydrofluoric acid and stained for potassium and calcium so that potassium feldspars and plagioclase feldspars can be readily identified. All thin sections have permanent cover slips placed on them.
