A multiyear dust devil vortex survey using an automated search of pressure time series

Abstract

Dust devils occur in arid climates on the Earth and ubiquitously on Mars, where they likely dominate the supply of atmospheric dust and influence climate. Martian dust devils have been studied with a combination of orbiting and landed spacecraft, while most studies of terrestrial dust devils have involved manned monitoring of field sites, which can be costly both in time and personnel. As an alternative approach, we describe a multiyear in situ survey of terrestrial dust devils using pressure loggers deployed at El Dorado Playa in Nevada, USA, a site known for dust devil activity. Analogous to previous surveys for Martian dust devils, we conduct a posthoc analysis of the barometric data to search for putative dust devil pressure dips using a new automated detection algorithm. We investigate the completeness and false positive rates of our new algorithm and conduct several statistically robust analyses of the resulting population of dips. We also investigate possible seasonal, annual, and spatial variability of the putative dust devil dips, possible correlations with precipitation, and the influence of sample size on the derived population statistics. Our results suggest that large numbers of dips (>1000) collected over multiple seasons are probably required for accurate assessment of the underlying dust devil population. Correlating long-term barometric time series with other data streams (e.g., solar flux measurements from photovoltaic cells) can uniquely elucidate the natures and origins of dust devils, and accurately assessing their influence requires consideration of the full distribution of dust devil properties, rather than average values.

Dust Devil Populations and Statistics

“…If it must be done, then ideally an information-theoretic criterion might be used in bin-size selection (e.g. Jackson and Lorenz, 2015), or at least binning be done in some manner well-suited to strongly skewed data e.g. logarithmically (bins of size ratio equal to the square root of two, as used often in the impact cratering community, is one approach).…”

“…A preliminary inspection of data from a trial field campaign at El Dorado Playa in Nevada (Lorenz, 2013a) suggested detectable encounter rates (>0.3mbar) on average of 1.3-2.6 per day. More systematic evaluations recording more than a hundred candidate events (Lorenz and Lanagan, 2014) and indeed some thousands of events (Lorenz and Jackson, 2015) have been conducted. This is an author-produced, pre-peer-reviewed version of this article.…”

“…With 252 million pressure readings, Jackson and Lorenz (2015) recovered more than 1,000 putative dust devil pressure drops during a two-year terrestrial field campaign and were able to compare different parametric fits to the histogram of pressure drops. Figure 5 shows data from that study which illustrates the pressure drop histogram, with bin sizes determined using Knuth's Rule, a binning scheme that uses a Bayesian framework to determining bin size.…”

“…The aggregated counts of Jackson and Lorenz (2015) shown as a differential population function with bins sized to maximize information content. A power law fit is shown with the solid line, and an exponential with a dashed line.…”

“…Both candidate functions deviate from observations below 0.2 mb, either due to an intrinsic minimum intensity of the population, or low detection efficiency for events near the noise level. Jackson and Lorenz (2015) also leveraged their large number of recovered pressure drops to investigate how the number of detections influenced the resulting fits and determine how robustly a given power-law index could be recovered if fewer detections were made. Randomly selecting sub-samples of the full, recovered population and fitting a power-law to the each resulting histogram, the study found that smaller populations of drops could give wildly different power-law fits and, to retrieve the final best-fit to the full sample, at least 75% of the full sample would have had to be recovered.…”