Abstract Ground failures, ranging from long tension cracks or fissures to surface faults, are caused by man-induced water-level declines in more than 14 areas in the contiguous United States. These failures are associated with land subsidence caused by compaction of underlying unconsolidated sediment. Fissures, which range in length from dekameters to kilometers, typically open only a few centimeters by displacement but are eroded by surface runoff into gullies 1 to 2 m wide and 2 to 3 m deep. Surface faults commonly attain scarp heights of 0.5 m and lengths of 1 km; the highest and longest scarps are 1 m and 16.7 km, respectively. Scarps grow by aseismic creep at rates approximately ranging from 4 to 60 mm/yr; modern fault movement is high angle and normal. Fault movement commonly correlates with seasonal water-level fluctuations, and examples of seasonal water-level recoveries halting fault movement have been reported. The greatest economic impact from ground failure is in the Houston-Galveston, Texas, metropolitan region where more than 86 surface faults have caused millions of dollars of damage and losses of property value. Most ground failures probably are caused by localized differential compaction, although this mechanism has not been demonstrated everywhere. Earth fissures formed by this mechanism are caused by stretching related to bending of the overburden that overlies the differentially compacting zone. Surface faults form when differential compaction is discrete across preexisting faults. Fissures that form complex polygonal patterns probably are caused by tension induced by capillary stresses in the zone above a declining water table. Ground failures can be predicted either by determining potential areas of differential compaction or by monitoring surface deformation in areas of ongoing water-level decline. Potential ground-failure sites can be resolved by either technique to within a few dekameters.

Great quantities of dust are transported to central Arizona by strong summer winds blowing mainly from southern and eastern Arizona and northern Mexico. Dust collected periodically from April 22, 1972, to July 21, 1973, from the cedar shingle roof of a house in Tempe, Arizona, in the lower Sonoran Desert of the southwest United States, indicates a rate of deposition of 54 g · m 2 /yr, a rate equivalent to 5,400 t · ha/10,000 yr. Most of the dust is deposited in association with summertime dust storms (haboobs), which on the average occur 3.5 times per year in the area. Approximately 75% of the dust particles have diameters between 0.05 and 0.005 mm. Chemical study and examination of heavy minerals indicate that the dust is mostly derived from igneous and metamorphic rocks. Three percent of the dust is CaCO 3 , 5% is iron oxides, and 0.7% is MnO. It can be calculated that 162 t of CaCO 3 is deposited per hectare in 10,000 years. The abundance of CaCO 3 in the dust strongly supports the suggestion that the source of CaCO 3 in caliche is from windblown dust. It has recently been demonstrated that the iron-manganese coating that forms desert varnish on rocks and stones in arid regions is at least 70% clay minerals, and that the source of these fine-grain particles is desert dust. Our studies indicate 270 t/ha of iron oxides and 3.8 t/ha of MnO are deposited every 10,000 years in the area. There is sufficient iron-manganese deposited annually in southern Arizona by windblown dust to form the desert varnish at the rate presently hypothesized. The process of blowing dust is a geologic hazard and affects man mainly in the form of air pollution. Manifestations of this air pollution are loss of visibility to motorists during dust storms, thereby causing accidents, many of them fatal, and a sometimes fatal illness known as “valley fever.”