 About one third of our planet's land mass is arid. These ecosystems are as complex as they are varied and are home to an astonishing number of plants, animals, and human cultures. Yet within most of these highly varied desert ecosystems, it is the rivers that are the vital lifelines in these parched environments. They provide water to support communities, wildlife populations, agriculture, transportation, and industry. The river valleys in these desert environments also provide a vast and highly interesting geological history of the region if you have the tools to read the geologic records. This is a look at the geological history that scientists in the desert project unlocked in one small but highly significant desert river valley, the Rio Grande Valley in southern New Mexico. What makes this story especially interesting is that it was one of the first desert river valleys to be decoded, and that the knowledge and techniques developed here are now used by scientists all over the world as a basis for understanding many other desert river valleys. The Rio Grande stretches 1500 miles through much of the Chihuahuan Desert in North America. It also marks half of the boundary between Mexico and the United States. Today, here on the banks of the Rio Grande, I'm standing on the river's modern flood plain. But over there on the West Mesa, you can still see the flood plain of the ancestral Rio Grande. It's called the La Mesa surface, and it's the relics of the basin floor that once extended across the entire area now occupied by the Rio Grande Valley. Four huge desert basins contain sediment laid down by the ancestral Rio Grande. The hornada, messia, tulerosa, and hueco. By tracing the Rio Grande's river pebbles, scientists discovered that at one time the river passed through that gap over there into the tulerosa and hueco basin. Later, when the land terrain changed probably due to faulting and uplifting in the past, the ancestral river shifted back to the west side of the mountains and flowed into Lake Cabeza de Baca, located in what is now northern Mexico. Over millions of years, sediments raised the floor of the lake and the base level of the Rio Grande in this region. Later, the lake and river sediment's upslope were cut or incised when the Rio Grande Valley was formed. Walking along the edges of this river valley, most people just see beautiful rock formations and interesting terrain. But from 1957 to 1972, podologists and geologists working with the USDA's Desert Soil Geomorphology Project understood that the land terraces, the alluvial fans, and the relic basin floor along the valley had a rich history to tell. Scientists first began their study. Very little was known about the desert soils of the world, although they cover much of the Earth's land area. This is a cross-section of landforms along the Rio Grande, studied by scientists in the Desert Project, as it became known. These fans and terraces slope into the valley from mountains that provide sediments. Portions of these fans and terraces were preserved as the valley entrenched, and today form a stepped sequence of landforms that are related to age. Scientists working for the Desert Project learned they could determine the relative ages of these landforms by their position in the step sequence and by the amount of calcium carbonate built up in their soils. As scientists began investigating these soils, they discovered that many of the older soils were cemented with large amounts of calcium carbonate. Yet the actual rocks in the sediments contained very little calcium. Thus, in the early days of the Desert Project, the question was, where did all this calcium carbonate come from? For ten years during the windy season, scientists set out dust traps in these mountains, alluvial fans, and basin floors, and the captured dust was analyzed. The findings showed that the dust blowing into the entire region was calcareous, and that even the rainwater falling through this dust contained dissolved calcium. Yet the surrounding soils and rocks were not calcareous. Therefore, the scientists knew that the carbonate deposited in these soils had come from the atmosphere, blown in from elsewhere in the region. What scientists also discovered was that the older the soil, the more calcium it contained. For example, in Terrace No. 5, the youngest in the valley, calcium deposits show up as white filaments, coating sand grains and pebbles in the soil. This is stage one of carbonate accumulation. Given time as carbonate accumulation increases, it becomes dense enough to plug the soil, and eventually a thin laminar zone forms. Here, water percolating through the soil is blocked and begins to move laterally. This is stage four of carbonate accumulation, and we find it in Terrace No. 3 of the step sequence of landforms. Lee Gile, one of the original researchers on the desert project, shows us an example of this phenomenon. This soil is in Terrace No. 3 and is estimated to be 50,000 to 100,000 years old. The soil has stage three carbonate and a thin, plugged zone. Here's the plugged zone. This is the laminar zone. When the soil solution can no longer move vertically, then it starts to move laterally, forming these thin laminar horizons. In the step sequence of surfaces, soil age and carbonate accumulation increase with increasing elevation of the step. This is shown by the ridge in the background, which is the number one landform and is the oldest soil in the step sequence. Research in the Hornada Basin in southern New Mexico near Rincon has helped provide some of the clues as to ages of the sediments and time of the Rio Grande Valley down-cutting in that area. These sediments beneath the La Mesa surface behind me were deposited by the ancestral Rio Grande between about four million years ago and the beginning of the down-cutting event. We date these rocks using various radiometric techniques. They're too old to date with carbon-14, so instead we use a variation of the potassium argon method called argon-argon dating. And we date volcanic material like this piece of pumice that's inner bedded with the river sediments. Pumice deposits in southern New Mexico were derived from the Hamas volcanic field, 250 miles to the north. Volcanic eruptions choked the river with pumice which was then transported very rapidly downstream in a matter of days or weeks. These pumice deposits are important for determining the age of the river sediments of the ancestral Rio Grande. We have identified five distinct pumice beds ranging in age from 3.1 million years to 1.3 million years ago. This particular one right here is dated at 1.6 million years ago. With that information, scientists can use a second test, the magnetic polarity of the river sediment, to help date these sediments even further. Here's how it works. Look at this compass. Today, the needle end of the compass points north because the Earth's magnetic field is oriented north. Iron minerals within sediments will naturally align themselves toward magnetic north, just like this compass needle. However, about 780,000 years ago, the Earth's magnetic field changed. If you had looked at this compass 780,000 years ago, the magnetic north-seeking end of the needle would have pointed south. Iron minerals within sediments deposited during that time would have aligned themselves toward the geographic south pole. These reversals in polarity provide important clues for scientists. Throughout geologic time, the Earth's magnetic field has reversed several times. There were periods when the magnetic pole was in the north, which we call normal polarity, and periods when it was in the south, called reverse polarity. This pumice layer is in a region of reverse polarity. So are the upper few meters of sediment beneath the La Mesa surface. So these sediments must be between 780,000 and 900,000 years old. In general, magnetic polarity can help us date these sediments. But in the case of this La Mesa surface, more paleomagnetic analysis is needed. This is because in the messiah basin, the upper few meters of the La Mesa sediments have a normal polarity, indicating that they were deposited sometime after 780,000 years ago. In any case, we can conclude from paleomagnetism that the La Mesa sediments are younger than 900,000 years old. But what happened after 900,000 years? It would stop sedimentation and cause the river's base level to quit rising. The answer lies with the topography around Lake Cavesa de Vaca. The lake was surrounded by high terrain, which formed a basin to hold the water in. Over thousands of years, the river filled the lake until a time when water reached a low point in the rim and spilled over. The rushing water eroded rocks and soil, cutting a gash in the landscape, which allowed even more water to escape. And what is now northern Mexico was instantly pirated of its freshwater lake. This happened in the area of what is now El Paso, Texas. Here the pirated water poured into the Waco basin. The draining of the lake impacted the river upstream as well. Water flowed more rapidly and the river eroded or downcut through its own sediments. The river channel became deeper and narrower, forming something like a gully but on a grand scale. But downcutting was not continuous. Instead, the Rio Grande downcut in an episodic manner stopped and started, leaving stepped terraces, such as the ones behind him. But when did this spillover and downcutting begin? The sediments themselves provide some of the food. Remember that the polarity reversal data and the argon argon dating tells us that the downcutting began sometime after 900,000 years ago. After the formation of the La Mesa surface, we see the development of the first inset terrace, caused by the draining of the lake. Fortunately for us, this terrace contains a volcanic ash called the bishop ash. This ash is from a volcanic eruption near Bishop California, which originally spread across a wide portion of the western United States. Radiometric dating tells us that the ancestral Rio Grande had already begun to downcut and that this first inset terrace had formed at 740,000 years ago, the age of this ash. Thus the age of the downcutting of the ancestral Rio Grande was between 900,000 years ago and 740,000 years ago. Next, we look at the other four terraces in the Rio Grande system. As Leland Gile and others observed, the lower terraces contained progressively less calcium carbonate in their soil, indicating that each was younger than the one above it. In general though, when we think of sedimentary beds, we think in terms of superposition, where the oldest sediments are deposited on the bottom and then younger sediments are deposited on top. However, with these inset terraces formed by downcutting and backfilling, the younger terraces are at the lower level and the older terraces are at a higher elevation. Why would this be so? Consider a bathtub. As you know, if you don't drain your bathwater right away, a soap ring forms around the edges of the tub. When you drain out part of the water, the level drops but the ring stays behind. Put the plug back in the tub and wait a while longer, another ring will form. Drain it again, plug it again, and the same thing happens. By the end of the day, you've got a nice series of circles around your tub with the oldest at the top and the youngest at the bottom. In a similar manner, the five inset terraces of the Rio Grande indicate downcutting at least five times since the draining of Lake Cabeza de Baca. However, unlike our bathtub analogy, the Rio Grande was eroding and downcutting, lowering its base level. Then the river began depositing sediments, partially backfilling and slightly raising the base level. The river downcut, backfill partway, downcut again, backfill partway, and so on. The result of this deposition of sediment is that the elevation of these five terraces can be as much as twenty meters higher than it would have been through downcutting alone. What forces caused the Rio Grande to downcut in backfill producing these terraces? Apparently, the explanation lies with the ice ages when much of the northern United States was covered with glaciers. The Rio Grande begins in the southern Rocky Mountains, which were also covered with glaciers during the last ice age. During glacial periods, when glaciers in the Rocky Mountains were advancing, the Rio Grande carried less sediment downstream and had a greater discharge. As a result, the river had a greater capacity to downcut. In contrast, during the interglacial periods, the river discharge was less. More sediment entered the river because there was less vegetation to prevent erosion. The river became choked with muddy sediments which were carried downstream. These sediments backfilled the channel, causing a slight rise in base level. With each glacial period, the process repeated itself. So although far from the glaciated parts of North America, these river terraces in the warm, arid southwest also record glacial advances and retreats. Scientists are learning that the principles clarified here are somewhat typical of actions that took place in many other desert river valleys around the world. Like many desert river valleys, the Rio Grande and its surrounding landscapes evolved and left evidence of periodic climate change. By using the clues provided by the soil carbonate, ash layers, fans, terraces and river sediments, scientists in the desert project were able to look back in time and discover how it happened. They found some basic principles that are still fundamental to similar scientific work today. First, the relative ages of landforms can be observed by determining their position in the step sequence and by measuring the amount of soil carbonate in the soil. Next, the approximate ages of the soils and landforms can be determined by using radiometric dating and paleomagnetic data. In the case of the Rio Grande, argon dating and paleomagnetism indicate that river sediments were filling Lake Cavesa de Huaca until sometime after 900,000 years ago. Then, the lake suddenly broke free of its mountainous boundaries. As the lake drained, the Rio Grande began its cycle of downcutting and backfilling. The presence of the Bishop ash layer within the oldest inset terrace indicates that this date was before 740,000 years ago, so the entrenchment must have occurred sometime between 900,000 and 740,000 years ago. The cycles of downcutting and backfilling were driven by glacial activity. So, that's the story locked inside the rocks and terraces of the Rio Grande Valley near Las Cruces, New Mexico. Is the story unique to just this valley? Hardly. In fact, similar geological actions took place in many desert river valleys around the world. That's why the principles the desert project scientists learned here are so important. Today, international scientists who have never even been to this valley refer back to this landmark research and use it as a guide to find clues to the history of their desert river valleys. The stories hidden within these soils and sediments may never be completely understood. More research is always needed. But just like layers of sediment, knowledge builds on knowledge. One mystery solved gives clues to understanding several more. And who can guess what stories these rich valleys still have to tell?