University of Washington Publications in Geology The Geomorphology and Volcanic Sequence of Steens Mountain in Southeastern Oregon

The great eastern scarp of High Steens presents the most extensive exposures In south central Oregon. Locally outcrops are almost continuous for a height of over 5,000 feet, while across Alvord Desert, on the lower opposing wall of the graben the higher members of the series survive. Except for minor tuffaceous deposits and a few intrusions, the mountain is composed of flows of varying magnitude. The volcanic activity responsible for this lava was predominantly parallel to the potential line of weakness later followed by the great fault. Many of the vents are exposed in the deep valleys that cut the scarp.

Most of the members of this sequence are formed by the thick accumulation of the near vent phase. Owing to the high viscosity at the time of extrusion, some of the flows are very local in their distribution. On the other hand, the flows of more fluid types are widely distributed and, if of a characteristic variety, have considerable stratigraphic value. The tuffaceous sediments of this region show such marked lithologic similarity that their correlation must be based largely on a rather meager fossil content.

In piecing together the regional geology of this broad volcanic field, it is necessary to correlate the series exposed in isolated scarps, which are usually 20 or 30 miles apart. Aside from the topography, only the ubiquitous presence of a thick series of basaltic lavas, which were extruded in a highly fluid condition, suggests a definite correlation. Otherwise the rapid changes in the stratigraphic sequence even on a single scarp clearly indicate the impossibility of basing regional correlation on the distribution of viscous lava or on minor variations in the volcanic sequence.

SUMMARY

Near the base of the Steens Mountain scarp, between Cottonwood Creek and Little Alvord Creek, there are scattered exposures of well-stratified tuffs, which are at least 800 feet in thickness. A small flora indicates these beds to be of Middle Miocene age, thus approximately corresponding to the Mascall Formation, which overlies the great series of Columbia River Basalt in the John Day region. For the present, these tuffs are called the Alvord Creek Beds in reference to the location of their major outcrops. Unfortunately, owing to the difficulty in correlating isolated exposures, the complete structure was not conclusively interpreted.

To the north of Alvord Creek, however, the subsequent flows are concordant with these beds, while to the south, the overlying series is nonconformable. Here, the sediments have been intruded by a thick basaltic sill and subsequently uparched by an acidic intrusion, which probably was in the form of a laccolith. The uppermost beds of this uparched structure have been truncated by an elongate acidic vent which parallels the scarp for about a quarter of a mile. The absence of deformation indicates its intrusion to have been subsequent to that of the laccolith. Extrusive material from this vent, however, is not exposed and cannot have been very great.

The southern extent of this laccolith could not be accurately determined, but it may well be responsible for the southern dip of the tuffaceous beds exposed on the northern wall of the valley of Little Alvord Creek. The dip slope of these beds formed the retaining wall for a massive flow of rhyolite about 400 feet in thickness. In this same valley, the vent for this flow is exposed cutting the inclined well-stratified tuffs. Poor exposures render it difficult to trace this rhyolite flow to the south so as to permit an accurate correlation with a complex series of acidic lavas exposed in the adjacent valley.

This valley, which is known as Pike Creek, shows an alternation of stratified tuffs and acidic flows. At the base are exposed about 200 feet of tuffs beneath an equal thickness of platy rhyolite, which locally appears to overlie its vent. On this flow was deposited more tuffaceous material to a depth of almost 300 feet. A well-defined vent crosscuts these beds and wells into another thick flow of platy rhyolite, which is at the same stratigraphic level as the previously mentioned rhyolite at Little Alvord Creek. The two flows, however, appear to be distinct. This lava was followed by another bed of stratified tuffs, about 40 feet in thickness.

Above this are two thick dacitic flows of very similar composition. The upper one is the larger and shows distinct vent characteristics over an extensive region between Pike Creek and Indian Creek. Here the exposures have a thickness of over 500 feet in spite of the erosion of the surface features. Northward the upper dacite thins gradually to almost half that depth, and at its northern extremity appears to abut against the previously mentioned elongate vent exposed above the laccolithic structure.

To the north, in the valleys of both Cottonwood Creek and Willow Creek, other masses of acidic lava outcrop at the base of the scarp. The former locality shows only a fault contact with the later series. Although due to their isolation, it is impossible to determine the relationship of these lavas to the acidic series exposed to the south, they at least have a similar stratigraphic position, in as much as they were both followed by the eruption of andesite.

Between Alvord Creek and Cottonwood Creek, a flow of basic andesite is interbedded in the Alvord Creek Beds at about 100 feet below their top. These same beds are also overlain by another thick andesitic flow of similar composition. This upper flow of basic andesite was capped first by a well-stratified bed of coarse tuffs of intermediate composition, and then by a huge flow of slightly more acidic andesite, which locally shows a visible content of hornblende. Immediately to the north of Alvord Creek, this flow has a thickness of approximately 900 feet, while farther to the north in the valley of Cottonwood Creek it has thinned to about half of that thickness. In the valley of Mann Creek, a similar lava forms the lowermost exposures and shows a thickness of 300 to 400 feet. Presumably it is part of the same flow, although it was not observed in the intervening valleys. Possibly it was derived from a series of contemporaneous vents, but the only discordant relation observed is exposed on the southern side of the valley of Alvord Creek. The presence of another locus of extrusion, however, is also distinct on the northern wall of the valley of Little Alvord Creek. Here the andesite welled above its vent to a sufficient depth to permit a thin extension to cap the upper dacite of the Pike Creek series.

Prior to any obvious erosional interval, the irregular surface of this great flow was capped by the extrusion of thin aphanitic flows of similar chemical composition. This upper series, which shows a maximum thickness of over 1,500 feet, attained a relatively level upper surface, although it locally decreases rapidly towards the west. The flows are remarkable in the fact that they consist predominantly of vesicular breccias, which are locally cut by auto-injections from their own thin platy flows. The basal breccia of each flow, as a rule, merges with the surface of the previous one, so that their contacts are usually imperceptible unless defined by an accumulation of stratified tuffs. This series has been greatly complicated by a discontinuous series of vents, which are scattered through a zone roughly parallel to the scarp. Between Mosquito Creek and the valley of Little Mann Creek, two of these vents gave rise to explosive activity, which formed cinder cones about 400 feet in height near the top of the series. These cones were subsequently almost completely submerged by additional flows, which are similar to the lower ones except for being slightly more vitreous.

Before any marked erosion had taken place the andesite was covered to a depth of at least 3,000 feet by a series of basaltic flows. These flows have marked physical and petrographic characteristics which are largely dependent on their extreme fluidity at the time of extrusion. The identical series can be traced southward to Pueblo Mountain and westward to Warner Valley and Abert Rim.

R. J. Russell¹ refers to the capping basalt in the Warner Mountain region of California as the Warner Basalt and correlates it both with the uppermost basalt exposed to the north near Plush, and also with the great thickness of lava at both Bluejoint and Abert Rims. He did not realize, however, that in the Warner Lake region the capping basalt, which is merely several hundred feet in thickness, rests disconformably on stratified tuffs overlying a far greater series, which is apparently formed by the western continuation of the Steens Mountain flows. In consequence, the writer will follow Russell's nomenclature in referring to the capping flows as the Warner Basalt, while using the name Steens Mountain Basalt for the lower series which forms the great thickness both in Warner Valley and at Abert Rim.

---

¹R. J. Russell, op. cit., pp. 416, 427, 439.

---

On the lower scarps in the region, which have been subjected to less vigorous erosion, quite a complex later chapter of the volcanic history is still apparent. The series exposed is extremely variable in its sequence. In general it consists of miscellaneous alternation of acidic, intermediate and basic flows with light colored tuffs which may or may not show stratification. The uppermost member both to the east of Alvord Desert and at the northern end of Warner Valley is slightly nonconformable. In case of the former, however, the series is capped by basalt, while in Warner Valley an acidic flow locally closes the volcanic history. Otherwise the upper sequence appears approximately conformable with irregularities due merely to the viscosity of the lava. This varied series has a thickness of over 500 feet.