© Christopher Earls Brennen


"They cannot scare me with their empty spaces
Between stars - on stars where no human race is
I have it in me so much nearer home
To scare myself with my own desert places".

From "Desert Places" in "A Further Range" by Robert Frost (1936).


Out in the vastness of California's Mohave Desert, in a place where no-one lives and few ever venture, there is a 40mi stretch of wind-blown sand known as the Kelso Dunes. Superficially there is nothing unusual or surprising about these dunes unless you happen to be nearby during one of those special moments when the wind blows and these dunes resonate with the most amazing and other-worldly noise, a loud low-frequency rumble that, as yet, defies scientific explanation. In the geological literature they are known as "booming dunes"; the documentation suggests that the phenomenon is rare, that it occurs at only about 30 sites around the world. Wherever they are found, they are the subject of much primitive legend and myth, whether in North Africa, in Australia or in the southwestern United States. We know that the noise is emitted when the wind builds the steep slope on the leeward side beyond its maximum sustainable gradient. When the slope then fails, it generates an avalanche of sand that also triggers a deep resonance within the dune. The sound emerges as an eerie, single tone that can continue for many minutes and can be heard for miles around. Ever since I first heard of this bizarre natural phenomon I have wanted to witness it first hand.

So it was that on February 9, 2002, graduate student Steve Hostler and I left Pasadena about 8.00am and travelled Interstates 210, 10, and 40 into one of the most desolate places anywhere in the world. Interstate 40 crosses the Mohave Desert from the city of Barstow to the town of Needles on the Colorado River, a 144mi stretch of road with only a few small outposts of human presence along the way. About 51mi east of Barstow we passed the wayside stop called Ludlow, a small cluster of buildings with a gas station, the last supply point for many, many miles. Another 26mi east of Ludlow, we came to the well-marked exit for the Kel-Baker Road. Leaving the highway we turned to travel north on this good two-lane asphalt road, almost immediately entering the Mohave National Preserve. The Kel-Baker Road proceeds north over a low saddle in the Granite Mountains and then continues north in a gradual descent over the flat alluvial plain on the north side of the Granites. At the saddle a broad vista opens up to the north, the bright yellow dunes clearly visible in the distance. Scattered over a flat plain, they seem to stretch from the distant northwest horizon to a terminus at the bottom of the long and gradual upslope created by the Granite mountains.

It was cold (in the low 50s?) and windy here. The turn-off to the Kelso Dunes viewpoint was just 100yds past an easily recognized pumping station on the left side of the road. There we turned left onto a well-maintained dirt road that headed west, arriving after 3mi at the restrooms, sign boards and parking area that constituted the welcome at the Kelso Dunes trailhead.

Kelso Dunes from the trailhead Windblown sand and the steep lee slope

Though the direction often changes, the most common wind is from the west. Far away in that direction, it funnels down the wash of the Mohave River, through a narrow canyon and out onto a broad flood plain that stretches off to the southeast for almost 40mi. The sand carried along this route over many eons has piled up in a series of dunes, the last and largest of which lies in an EW direction just to the north of where we were now parked. We could see the dramatic ridge of dunes out across the sand, the highest peak bearing off to the northwest. Though it appeared quite close, the summit was over 2mi from the trailhead and 560ft higher in elevation.

From the trailhead a use-trail headed directly toward the peak, petering out after just 200yds where it crossed the distinct boundary of the sand dune material. We then ploughed across a series of low dunes, most with sparse grass-like vegetation. These gradually increased in height, elevation and sand mobility. With more mobile sand, the vegetation became thinner. Hiking through soft sand on steep slopes is not easy and so this was a more difficult hike than might have been imagined. Where the sand had been disturbed (for example along the trail) or recently deposited (on the steeper lee slopes of both the large and small dunes) it yielded easily underfoot. However on the tops of many of the foothill dunes where the surface had been scoured by the wind the surface was markedly different. Not only was it quite firm underfoot and therefore much easier to hike on but it also looked different, having larger, more visible black grains, probably of volcanic origin. This gave the surface a gray color, sometimes in bands running roughly perpendicular to the direction of sand transport. This coloration was in marked contrast to the light yellow color of the steep lee slopes. It seems that a process of sand sorting is occurring on these small and moderately-sized foothill dunes. The small grains (yellow colored) are being preferentially removed from the relatively flat tops and deposited on the steep lee slopes thus leaving a greater proportion of the larger black grains behind to produce the gray coloration.

The route steepened markedly as we approached the bottom of the slope leading up to the crestline. There appeared to be two common climbing routes. One was the direct route that necessitated a long and very steep final climb to the summit. We took the alternate route and headed for the saddle about 75yds east of the high point. By following the firmer, gray-tops of the foothill dunes we climbed to within about 20ft of the crest of the saddle. Over the last 20ft we had to crawl up the loose lee-side sand.

Close-up of the crest View along the crest

We paused at the saddle to look around. My impression was that the sharp crest of the dune consists of a steep pile of very fine sand sitting like a triangular cross-sectioned dike on top of a lower ridge of firmer, less-mobile sand. The wind (from the north on the day we were there) seems to carry only the very smallest grains to the crest. There the particles become airborne and are deposited by falling onto the very steep lee slope. The ridgetop geometry is extremely uniform and well-defined; it extends for miles in both the east and west directions. In cross-section the crest consists of a fairly steep windward slope, a small flatter area on the top, a sharp break and then the very steep leeward slope (30-45 degrees).

From the saddle, we trudged the additional 75yds to the highest peak, arriving there 1hr 20min after leaving the trailhead. The summit had the added feature that it was at the junction of the main crest and a secondary crest that branched steeply down the south facing slope in a southwesterly direction. This seemed caused by wind blowing both over and around the peak.

The wind at the crest was much stronger than we experienced even 100ft down the northside. Sand was blowing everywhere, getting into everything, coating our lips, eyes and nostrils. When I sat down to rest the density of wind-blown sand impacting my face was much greater than when I was standing. Thus I estimate that the thickness of the blowing sand layer on the windward side at the crest was about 3-4ft. This sand is then projected (roughly horizontally) over the top and cascades down onto the steep leeward slope.

The leeward slope at the ridgeline is so steep that by kicking on the crest one could generate a small landslide (perhaps 2-3ft in width) that would proceed slowly, almost viscously, downhill. The typical speed of these small landslides was no more than about 1ft/s. But they would continue (unexpectedly) for minutes, usually until they reached the bottom of the lee slope. These landslides also occur naturally without human intervention and leave vertical striations all the way across the lee slope. We also observed similar landslides on the lee slopes of the foothill dunes. However, during our entire climb we could hear no sound from any of these naturally occurring landslides and so began to speculate on why we could hear no "booming". Perhaps, we thought, the day was too cold for booming.

After a pause, we got to work and loaded a large bucket of sand into one of the backpacks for transport back to the laboratory. Then, having apparently failed in our effort to hear the booming, we decided to head straight down the lee slope on our way back to the trailhead. The only way to do this was to sit down and slide. Within a few feet of descent we had set into motion a much larger area of sand than occured with the small landslides. Then it happened. There emerged from under us the most awesome rumbling tone, increasing in volume as we continued to slide and persisting as long as the mass of sand was moving (and maybe even a little longer). It was an eerie and other-worldly experience. The whole dune seemed to shake as though, deep within, there was a large resonating cavity. This sound continued as we descended the 150ft steep lee slope but became more muted. About 50ft down from the summit we encountered a slope with some vegetation but still very steep. Even on this slope we could still detect the booming though it virtually disappeared near the bottom. It was a marvellous, exhilerating experience that sowed the seeds of a determination to find the origin of this mysterious phenomemon.

Excited and determined, we began our hike back to the trailhead, taking turns to carry the backpack with our precious 80lbs of "booming" sand. On the way back we tried several times to make the lee slopes of foothill dunes boom but detected only very slight squeaking noise, never the low frequency rumble that so awed us at the summit. We arrived back at the trailhead about 3hrs after setting out. Then, determined to make full use of our efforts we hiked back to the foothills to collect a second sample bucket of sand, this time from the area below the steep lee slope.

After a late lunch, we drove west along the increasingly rough dirt roads that parallel the south edge of the dunes. Our intent was to try to gather a third sample bucket of sand from a different geographic location. About 3mi west of the restrooms, the rough dirt road came within about 100yds of the edge of the dune sand. There I spotted a faint track down a shallow draw and headed down that track to try to drive closer to the dunes. We crossed a wash and found very soft sand at the far side. My vehicle would normally have no difficulty with this except that it was not in 4WD, despite the fact that I had put it in 4WD and the warning lights said it was in 4WD. The result was that when we tried to backtrack we became inextricable stuck in the sand in the middle of absolutely nowhere.

Steve and I tried for about 2hrs to either get it into 4WD or to get it out of the ruts it had dug. First we tried digging ramps in front of the buried rear wheels. The vehicle would just bury itself again. Then we tried lining the ramps with two old Mexican blankets I carry in the car. That only resulted in chewed-up blanket. After 2 hours we had managed to move the vehicle about 4 feet. It was clear that we would need outside help. Though I did have a cell phone I doubted that I could get a connection this far from any civilization and so I made plans to hike the 10mi or so to help, thinking that I would leave Steve with the vehicle since he was less accustomed to long distance hiking than I was. But before setting out I decided to try my cell phone anyway; I was quite surprised that someone quickly answered when I dialed 911. It was the Mohave National Reserve Ranger Station. They noted my verbal description of our position and said that they would summon the special off-road vehicle recovery service in Baker, about 50mi away at the northern end of the Kel-Baker road. We settled in to wait, for it would be more than an hour before this help could arrive. After a while we chose to climb to the top of a small hill where we would have a wide view of the desert off to the east and so be able to spot an approaching vehicle.

They took about 1.5hrs to arrive. Sitting there on our hilltop, scouring the horizon for any sign of the approaching rescuers, we were somewhat bemused by the approach of an unmarked helicopter and positively alarmed as it began to circle overhead. Clearly they were assessing the situation on the ground below. I had heard stories of illegal immigrants trying to cross through this unoccupied land and wondered if the helicopter might be trying to assess this possibility. Alternatively one sometimes hears stories of drug smuggling to remote desert airstrips. Leaving Steve on the hilltop, I walked back to my stranded vehicle in an effort to reaffirm our connection with it. Then the helicopter descended and landed about 100yds away. I moved slowly toward it but stopped as two aliens emerged from the small cockpit. Dressed from head to toe in gold lame suits, their heads encased in dark glass helmets, I truly began to wonder if I was hallucinating. But their soft California voices dispelled the craziness and I could now make out the California Highway Patrol labels on their flight suits. Apparently, the off-road vehicle recovery unit could not, at one point, get through to my cell phone and so they had asked the CHP helicopter to check our location and condition. Once I reassured them that we had enough water and blankets to last the night should that become necessary, the CHP aliens left us alone again in the middle of absolutely nowhere.

Steve and I again took up our vigil on the low hilltop. Finally, only about 15 minutes before sunset, we spotted the dust of the off-road recovery unit approaching from the east. Soon we were in cell phone contact and I was ready to guide them across the desert to our lonely resting place. First, however, the ungainly vehicle stopped about a mile away. It was then that we recognized that it consisted of a truck with an off-road vehicle mounted on the back. This rugged jeep was unloaded and then driven the final mile to where we were waiting, guided there by cell phone communication from my hilltop. The jeep had no trouble crossing the loose sand and winching my vehicle out to firmer ground. There were just minutes of twilight left.

If I had not had such a feeling of relief, it might have been anti-climatic. As it was I paid the hefty three-figure charge for the five minutes of work by the off-road recovery unit and then set out for home. In the heat of the moment, we failed to collect the third bucket that had been the cause of our grief. But we did remember to reload the first two buckets; Steve had hidden them behind a bush so that our elicit purpose would not be detected during the visit from the airborne cavalry.

Just another routine day at the lab.

Last updated 3/10/03.
Christopher E. Brennen