Granule ripples in the Kumtagh Desert, China: morphological and sedimentary characteristics, and development processes

1. Observation of composition and morphology

First, we selected typical granule ripples of different sizes and development stages at each observation site, and sampled the surface particles (to a depth of 1 cm) on the ridges of 85 granule ripples by scraping the surface with a steel ruler. The sample weight averaged 428 g (with values ranging between 249 and 697 g). We determined the grain-size distribution using a set of 15 standard sieves with a mesh size ranging from 0.0565 mm to 20 mm. We used the parameters proposed by Folk and Ward (1957) to calculate the grain size characteristics: the cumulative grain size of the 20 percentile (P20), and the average grain diameter, sorting coefficient, skewness, and kurtosis. Here, P20 represents the particle size corresponding to the cumulative distribution rate of 20% in the cumulative particle-size distribution diagram.

Second, we defined ripple morphological characteristics (the wavelength, ripple height, ripple index, and symmetry coefficient) for 142 granule ripples by measuring their wavelength, ripple height, and ridge position. The ripple index (Z) equaled the ratio of the wavelength to the ripple height. The symmetry coefficient equaled the ratio of the horizontal windward slope length to the horizontal leeward slope length.

Wavelength represents the horizontal distance between two crests, and the ripple height is the perpendicular height from the trough to the crest.

We also excavated vertical profiles of five typical granule ripples with different sizes, and measured the stratification thickness for one ripple with a wavelength of 5 m, for a vertical profile whose depth was 60 cm. Based on the different grain size characteristics, we identified five strata and therefore obtained five sediment samples with different depths, and analyzed the vertical distribution of particle sizes in the sediment.

We installed HOBO U30 field anemometers (Onset Computers, Bourne, MA, USA) at a height of 2 m and installed an eight-direction sand trap in the two areas in January 2012 to obtain the wind speed, wind direction, and the single-width sediment transport for a full year. The collection frequency for the wind speed and direction was 5 minutes, and the resolution of the wind speed and direction were 0.38 m/s and 1.4°, respectively. The collection frequency of the sediment transport rate was 1 month, and the resolution was 10 g. The width and height of each collecting port of the sand trap were 2 cm and 1.2 m, respectively.

2. Wind tunnel experiment to detect the gravel threshold velocity

We performed a wind tunnel simulation experiment to test the impact threshold velocity of the gravels. We used the high-speed wind tunnel at the High Speed Railway Construction Technology National Engineering Laboratory of Central South University in Changsha, Hunan Province in the validation test. We used surface sediments from granule ripples collected at our study area, and screened these sediments through 12 sieves with a mesh size ranging from 1.6 to 31.5 mm. We used the sieved material to create 10 different test surfaces with an average particle size ranging from 1.8 to 25.8 mm, which we subsequently bombarded with saltating particles from an upwind source. The upwind supply rate for the impacting particles was 1.2 kg/min, and their particle size ranged from 0.8 to 1.0 mm, with an average size of 0.9 mm, in all tests. The impacting particles were also obtained from the surface sediments of the granule ripples and were obtained by screening. The gravel bed surface was 80 cm long (parallel to the long axis of the wind tunnel), 40 cm wide, and 3 cm deep with its upper surface level with the wind tunnel floor. To increase the roughness of the surface, we glued a layer of coarse sand with a mean particle size of 0.9 mm that was 4.2 m long by 40 cm wide to the floor of the wind tunnel immediately upwind of the gravel bed.

We observed the particles with a high-speed camera in the wind tunnel to determine the type of motion (vibration, roll, obvious roll, saltation, obvious saltation). The wind velocities were measured using a pitot tube at 10 heights (6, 7.5, 13, 21, 41, 83, 123, 161, 220, and 242 mm above the gravel bed).

3. Measurement of the granule ripple migration rate

We chose a 25-m2 area at the yardang site, and monitored the migration rate of some smaller typical granule ripples from 11 October 2011 to 15 October 2012 using six marking pins, at an east–west spacing of 5 m between pins. The average wavelength of these granule ripples was 1.6 m and the average height was 0.09 m. The mean grain size in the ridge was 2.08 mm, and the maximum grain size was 6.5 mm. We also chose 11 larger granule ripples  (5 to 9 m in wavelength, 30 to 50 cm in height, granule size in the ridge of 6 to 10 mm) that were continuously distributed at the yardang site, and measured their migration rate over an 8-year period from 2011 to 2019 by using marking pins. Based on the migration rate, we applied the equation developed by Jerolmack et al. (2006) to calculate the sediment creep flux (qc) at the ridge of the granule ripples:

qc = (1-p) ρcH / 2

where p is the porosity of the ripple sediment (30%), ρ is the particle density (2650 kg m-3), c is the ripple migration rate (m/year), and H is the ripple height (m).