Geology of the Salt lake Area, Oahu, ! KOST A. PANKIWSKy]2

ABSTRACT: A series of pyroclastic eruptions and one lava flow, all of melilite­ nephelinite composition, were produced in the environs of present-day Salt Lake, Oahu, Hawaii, beginning about 500,000 years ago and terminating more than

100,000 years ago. The age of the flow has been dated by a K 40/ A40 method at 400,000 years to 470,000 years. The earliest eruptions took place at a time of higher sea levels than at present, and their products are water-laid. Later eruptions, in­ cluding the major ones from Salt Lake Crater and Makalapa Crater, took place at a time of lower sea levels than at present, and their products are air-laid. Sufficient time elapsed between individual eruptions to allow for erosion of portions of the earlier cones, to deposit silt and gravel, or to develop a soil prior to a subsequent eruption. The pyroclastics are medium- to fine-grained, lithic-vitric tuffs and con­ tain, in addition to juvenile material, a considerable quantity of particles of older rocks-volcanic and sedimentary, as well as ultramafic. The youngest deposits of tuff are unconformably truncated by reef and limestone which have been correlated with rocks dated by a Th23o/UZ38 method at 90,000 to 140,000 years.

THE PYROCLASTICS of the Salt Lake area of the topography in the present Salt Lake area con­ Oahu, Hawaii have a long history of geologic sisted of stream-cut valleys filled with silts and investigation. Hitchcock (1900) , Wentworth gravels, separated by ridges of erosional remnants (1926), and Stearns and Vaksvik (1935) com­ of Koolau lavas which were covered in places by prise the most notable stratigraphic studies. New red or brownish red soil. The initial deposits cuts were made in the early 1950s along the of Pleistocene pyroclastic material in the Salt widened Highway north of Salt Lake Lake area are water-laid and are intercalated and, in the late 1960s, on the inner slopes of with the silts and gravels noted above. The Salt Lake during development of a residential water-reworked cinder has been attributed to subdivision. This investigation was undertaken Aliamanu Crater (Stearns and Vaksvik, 1935, in order to examine new exposures as they were p. 109). More than one horizon of pyroclastics being developed and before they became covered is mapped, attesting to either a number of sep­ by man-made structures or were permanently arate eruptions or else removal and reworking removed in the process of earth-moving. The of ash from one major eruption. Land vegetation aim of the investigation was to establish a de­ came to cover the entire Pearl Harbor-Salt Lake tailed stratigraphy and geologic history of the area following a lowering of the level of the pyroclastic section in the Salt Lake area. sea. At this time eruptions from several craters in the Salt Lake area formed subaerial cones of Summary of Geologic History tephra. Those which I was able to recognize are The extinction of the Tertiary Koolau shield shown in Fig. 1, and are referred to in this paper volcano, which built the eastern part of the is­ as: Aliamanu School Cone, Moanalua Cone, land of Oahu, was followed by a period, prob­ Akulikuli Vent, and the Wiliki Cone. These ably in excess of 1 million years, of weathering cones were subject to erosion prior to being and erosion. About one-half million years ago, mantled by the major, most voluminous erup­ tions of air-laid tuff from Salt Lake Crater, 1 Hawaii Institute of Geophysics Contribution no. which, together with tuff erupted contempora­ 460. Manuscript received 13 October 1971. 2 University of Hawaii, Department of Geology and neously from Makalapa Crater, covers approxi­ Geophysics, , Hawaii 96822. mately 12 square miles at the present time. The 242 Geology of the Salt Lake Area, Oahu-PANKIWSKY] 243

~ RIM Of CRATER X. lOCATION MENTIONED IN TEXT ROADS 21"22'30"~------+-+=--J,.~:::::":::::::::"'~(

112 0 SCALE IN MilES 21"20 '~~;.LS--9':-0-0':-'--'--.:>...------1-57-•.J..--:'-30--:'--:'---....:..------.::c----..:-15-7-•..l54--:'-00--:,--:, ---I 56

FIG. 1 extent of surface exposure of air-laid tuff from of this tuff are capped by reef, testifying to a these eruptions is shown in Fig. 2. gradual rise in the level of the sea. Stearns On the seaward sides of the craters deposits (Stearns and Vaksvik, 1935, p. 127-128) related

IN 1923 IWENTWORTH,1926)

AT THE PRESENT TIME liN PART AfTER HAY,1966)

SCALE IN MilES 21·20'00"L__---l~...;,'- J...... J...... J 157"59'00" 157'56'30" 157'54'00"

FIG, 2 244 PACIFIC SCIENCE, Volume 26, April 1972 the high, low, and final high sea levels to the along the Oahu Railroad less than one-half mile Kaena, the Waipio, and the Waimanalo stands, west of Moanalua Station (near the present respectively. intersection of Puuloa Road and Kamehameha Highway?), which from bottom to top consisted Summary of Stratigraphy of: "the main coral reef; thin layer of tuff; coral The stratigraphic section in the Salt Lake area reef or limestone; decomposed rock sustaining a is divided into four groups for the purpose of soil covered by forest; eruption of tuff from ease of discussion and comparison: Aliapakai [Salt Lake Crater] ..." This ex­ Group 1. Tertiary rocks of the Koolau posure is no longer in existence. Wentworth Volcanic Series and their weathering products, (1926) specifically separates the typically water­ which are typically highly oxidized red-colored laid tuff below the air-laid tuff and discusses in soils showing little if any stratification. detail the geologic history of the area. The most Group 2. Water-laid, valley-filling se­ recent comprehensive description of the strati­ quences of well-stratified reddish, brownish, graphic sections in the area is by Stearns and and buff silts and clays; beds and lenses of Vaksvik (1935). I will call on their descrip­ gravel; reworked cinder of the Honolulu Vol­ tions from place to place in the discussion which canic Series; air-laid tuff from the several small follows. cones which were formed prior to the major Stearns (Stearns and Vaksvik, 1935, p. 105) eruption from the Salt Lake and Makalapa cra­ describes in considerable detail a cut on a pri­ ters; and various lagoonal and littoral sediments vate road to the Damon Estate leading up a including calcareous sandstone and siltstone, bluff on the southeast side of Manaiki Stream lime muds, coral, and shelly reef fragments, about 1,500 feet northwest of the main entrance some of which are intercalated with fine vol­ to (the present-day Pineapple canic ash. Street?). This location is shown as A on Fig. 1, Group 3. Typically air-laid tuff from the and the description of the section is shown at Salt Lake and Makalapa craters including the A in Fig. 3. Today, this exposure is heavily youngest pyroclastic deposit, a poorly consoli­ covered with vegetation along its slope and by dated dark grey ash. houses on top. The best present exposure in this Group 4. Calcareous deposits overlying vicinity is along the loop of road from Moanalua the Group 3 tephra. Highway to the seaward part of Fort Shafter, shown as locations Band C on Fig. 1, B being Description of Individual Cuts seaward of C. Eruptions from the crater, or more likely from These sections are shown in Fig. 3. They con­ several craters, in what subsequently developed sist entirely of the Group 2 of this report and into Salt Lake, completely blanketed the deposits can be divided into three parts. In decreasing of earlier eruptions. Some of the earlier deposits age these are: became uncovered by erosion, but most remained Part 1. Roughly bedded, rounded, water­ unexposed until excavated by man in the 20th laid dark grey cinder interlayered with a light century. It is to be expected that, in addition to brown silt and several boulder lenses. This the craters described in this paper, others may grades laterally toward the south into inter­ still be identified with further excavation. layered silt and gravels. The first geologic description of the Salt Lake Part 2. A 2- to 3-inch layer of tuffaceous craters is by J. D. Dana (1849, p. 245-248), who was the geologist of the 1838-1842 U.S. mud containing numerous plant fragments, well­ exploring expedition under the command of rounded cobbles and boulders of Koolau lithol­ Charles Wilkes, U.S. Navy. He describes the ogy in a tuffaceous matrix, and a well-bedded topography of the area and recognizes the three water-laid tuff. craters of Salt Lake, Aliamanu, and Makalapa. Part 3. Light brown silt with several Hitchcock (1900) describes the Salt Lake re­ lenses of gravel. In the southern part of cut B gion in more detail and notes particularly a cut also appears a bed of lithified calcareous sand- Geology of the Salt Lake Area, Oahu-PANKIWSKY] 245

stone containing fragments of reworked tuff lenses of coarse gravel, those in turn overlain by and cinder. a dark grey, but deeply red-weathering soil A section equivalent to the Part 1 above also which, near the top, contains well-preserved appears to be present near the top of the bluff plant remains. Above this is the base of the on the east side of Moanalua Gardens, the sec­ Salt Lake Tuff, characterized by a 2- to 6-inch­ tion A taken from Stearns. He described it thick layer of light grey poorly consolidated as a "stratified water-laid pumiceous fire-foun­ fine-grained tuff, rich in fossil plant remains. tain deposit." A similar section is also seen at Above this follows well-consolidated air-laid the present time just southeast of the crossing tuff. The roughly bedded tuff underlying the of Mahamoe Street over Manaiki Stream, loca­ soils and gravels is considered to be a remnant tion Z on Fig. 1, where it is seen to grade of a minor tuff cone, here referred to as the downward into a buff soil, and yet farther down Moanalua Cone which erupted on the floor of into spheroidally weathering aa flows of the the pre-Salt-Lake Moanalua Valley. It is this Koolau Series. cone, in my opinion, that first dammed Moa­ The imposing bluff at the corner of Moanalua nalua Stream and diverted its course to the Highway and Puuloa Road consists of a section southeast. of Group 2, very similar to that at Fort Shafter, Within the Salt Lake Crater area, on the but for the presence of the Salt Lake Tuff, northeast side of Ala Puumalu just southeast Group 3 of this report, capping the Group 2 of Ala Napunani (location Y on Fig. 1), one section. can at this time still find a similar section of As one traverses Moanalua Highway from the poorly consolidated cinder, covered by brown corner of Puuloa Road toward the northwest, to red soil, which, in turn, is covered by a very the following section can be observed in three poorly sorted tuff containng numerous xenoliths successive cuts on the southwestern side of the of Koolau lava. highway, located on Fig. 1 as D, E, and F, Farther westward along Moanalua Highway, respectively. the base of the Sale Lake Tuff dips below the At the base of cuts E and F is found a Group level of the highway until the highway reaches 1 section, the eroded remnants of Koolau lavas, the crest of Red Hill and then drops into Ha­ covered by a reddish or brownish soil layer. lawa Valley, shown as location G in Fig. 1. At Then in all three cuts are found sections com­ the crest of the hill a weathered aa flow is posed of sediments of Group 2. The silts are capped by only a few inches of red soil, and thickest in cut D, which is a section through a this is capped by air-laid tuff. Partway down fluvial flood plain. The gravels are coarsest in the slope to Halawa Valley the soil is thicker cut F, which lies near the center line of the and one can see the section illustrated at G ancient Moanalua River Valley. These gravels in Fig. 3. form lenses within interlayered water-laid cinder Lower in Halawa Valley, beds of coarse grav­ and brownish silt, equivalent to part 1 of the els are found intercalated with the soils below Fort Shafter cut (B and C in Fig. 3). Above the Salt Lake Tuff. Stearns (Stearns and Vaks­ this follow more silts and gravels, but also a vik, 1935, p. 110) reported a similar section at 2- to 3-foot layer of tuff containing plant re­ the former Aiea Station of the Oahu Railroad. mains. This is then capped by the Salt Lake This description is shown as H in Fig. 3. Tuff, the Group 3 section. The lower contact of I have not located this exposure; however, a the Salt Lake Tuff is at the present time partic­ recent cut along the H-l expressway southeast ularly well displayed on an excavated hill slop­ of Aiea, shown as location X on Fig. 1, shows ing south toward Salt Lake directly west of the a similar stratigraphic section, with the excep­ northwestern entrance to the Henry Damon Es­ tion that the subaerial tuff attributed to Salt tate, located at N on Fig. 1. At the bottom of Lake-Makalapa craters is not present, possibly this section, with its base not exposed, is a 5­ having been removed during excavation, and foot thickness of a very poorly sorted, very the horizontally bedded gravel and silt is seen roughly bedded tuff, overlain by brown silt with to grade downward into a brownish red soil, 50

THIN BEDDED LITHIC

AIR-LAID TUFF AIR-LAID TUFF AIR - LAID TUFF AIR -LAID TUFF L------....----- REDDISH SOIL

REDDI SH SOIL BOULDER LENS FINE BROWN SOIL WITH PEBBLES REDDISH SOIL 40 AIR-LAID TUFF

30 TAN SOIL DARK II GHT

f- UJ MA SSIVE AA N UJ ,j>. LL 0\ AA CLINKER E 20 F

10

RED SOIL

WEATHERED KOOLAU AA BASAL T o A FIG. 3. (lOCAllY OVER 100 fT. THICK) WEll-SORTEO~ 50 WEll- &EOOEO, fiNE TO ., COURSE - GRAINED TUff MUD LAYER WITH FOSSil PLANT REMAINS (UP TO 15 fT. EXPOSED) FINE-GRAINED, LIGHT GREY TUFF ~ ~:~"'''AIR -LAID TUFF WITH Well-SORTED, COURSE- WITH FOSSIL PLANT REMAINS XENOLITHS OF GRAINED TUFF, CEMENTED WITH ZEOLITES CI SOILS, SOME WITH GRAVEl > ULTRA MAFIC NODULE 5 WEll- &EOOEO AND CROSS- 40 l05J WATER - LAID TUFF OF CINDER, WITH OR ~lY OVER 100fT. THICK) WITHOUT SILT AND GRAVEL THINLY BEDDED. FINE- BEDDED, fiNE TO COURSE- GREY~<> GRAINED TUFF GRAINED TuFF 51 LT AND OR GRAVel WITHOUT TUFF • AND SPATTER WITH ~ POORLY-BEDDED, MEDIUM- XENOLITHS Of GRAINED TUFF ~ KOOlAU LAVA - MODERATELY WElL· BEDDED TUff WITH PE&&lES Of POORLY-SORTED TUFF WITH LIMESTONE & KOOLAU LAVA OCHRE CINDER ~ XENOLITHS OF LIMESTONE, 30 AND SPATTER KOOLAU LAVA, AND ULTRA- MEDIUM TO COURSE- MAFIC NODUL ES GRAINED TUFF, CEMENTED BY ZEOLITES K POORlY- SORTED TUff I­ CONTAINING 2"106" BEDS w REO STAINE 0, WEll- BEDDED WEll- &EOOEO, fiNE w AIR-LAID TUFF OF WATER-REWORKED TUFF u.. AIR-LAID TUFF GRAINED AIR-LAID TUFF FROM SALT lAkE­ MAKAlAPA CRATERS 20 RED· 501 L WITH BROWNISH SOil &ROWNISH REO BROWN CLAVEY SOil WEATHERED AA CLAYEY SOil

BROWNISH RED SOil 10 ------

DARK RED SOil

EXFOLIATING AA

0 H G FIG. 3 (continued) . 248 PACIFIC SCIENCE, Volume 26, April 1972 which grades farther downward into deeply silts; then about 12 feet of another tuff, locally weathered aa and pahoehoe flows of the Koolau underlain by about 2 feet of tuffaceous sand; and Series. then by an undetermined thickness of clays and A long series of terraces has been cut by the silts. On the basis of this description, correlation contractor for the Lakeside Subdivision to the of this section with any exposed in actual cuts south of Salt Lake, shown as location W on is extremely tenuous. The low tuffaceous sand Fig. 1. In the westernmost part of this cut is may perhaps be equivalent to the water-laid cin­ found, with its base not exposed, a thickness of der and the intermediate tuffs intercalated with up to 16 feet of a rather poorly sorted and gravel may be derived from the Moanalua or poorly bedded, palagonitized tuff rich in frag­ the Alaimanu School cones. ments of limestone and chunks of silt and mud. Southwest of the lake, in the back of the This section is unconformably overlain by the house at 3457 Ala Akulikuli just south of Ala Salt Lake Tuff, here rather fine-grained and Hinalo (shown as location T on Fig. 1) is ex­ moderately well bedded. The Salt Lake Tuff posed a small part of another local center of itself is truncated and mantled by a loosely con­ eruption, here referred to as the Akulikuli Vent, solidated dark grey ash. The limestone-bearing which predated the main eruption of Salt Lake tuff is also exposed along Salt Lake Boulevard Crater. This is composed of an irregularly north of the First Baptist Church. I consider it shaped mass in two dimensions, extending to 10 to be a remnant of a tuff cone, whose center was feet above ground level and 8 feet wide at the in the vicinity of the present Aliamanu School, base and with the third dimension being unclear, and, therefore, refer to it as the Aliamanu of medium grey, poorly bedded, and poorly School Cone. To the east, the contact between sorted lava balls, ranging in diameter from sev­ the Aliamanu School Tuff and the Salt Lake eral millimeters to about 20 cm. These balls are Tuff dips below the level of the ground. Farther cemented by zeolites and calcite, and resemble east, the base of the Salt Lake Tuff becomes ex­ sugar-dipped candy. The lava ball deposit is posed in contact with a reddish brown soil. truncated by a strongly palagonitized, moderately Here the bottom layer of the tuff is the light well-bedded tuff derived from Salt Lake Crater. grey, loosely consolidated, plant-remains-bearing To the south on Ala Akulikuli, this tuff trun­ rock type that was noted on the hill north of the cates another very similar tuff, possibly derived lake, at location N. Near the eastern end of this from an earlier eruption from Salt Lake Crater, cut, the Salt Lake Tuff is truncated and mantled but also possibly from another crater now com­ by the dark grey ash. In back of the parking lot pletely mantled by later eruptions. An even more of the apartment building at 3019 Ala Ilima impressive truncation between these two tuffs is (Polynesian Vista Apartments), location WI on in back of the house at 3411 Ala Hinalo, at the Fig. 1, the ash mantle has been scalloped out by corner of Ala Akulikuli, just 50 feet north of erosion and filled with a breccia consisting of the lava ball deposit. angular fragments of reef material as well as The large semicircular cut to the east of Salt pieces of tuff. These limestone blocks may be Lake, shown as location S on Fig. 1, now sched­ derived from the reef which caps the Salt Lake uled to be the site of a high school, shows sev­ Tuff south of here (see location V on Fig. 1). eral episodes in the eruptive history of Salt Lake Several borings were made by the Dames and Crater. Near the south part of the cut is a hair­ Moore Company in the lot prior to the construc­ line-sharp unconformity between two identical tion of the apartment building called the Ala tuffs, with the lower tuff dipping north, and the

Ilima at 2907 Ala Ilima (locaton W 2 on Fig. upper tuff dipping south. Continuing to the 1). Mr. Song of the Dames and Moore Com­ north, the lower tuff gradually assumes a dip pany noted (personal communication) that the of 60° to the northeast, suggesting postdeposi­ tuff at the surface is about 8 feet thick; under­ tional deformation. This tuff is then mantled by lain by 3 to 10 feet of clays and silts; then by the dark grey Salt Lake ash, which has been 3 feet of gravel, locally underlain by about 2 described previously. Fossil roots, some extend­ feet of tuff; then by 10 to 15 feet of clays and ing for over 6 feet along bedding planes, are Geology of the Salt Lake Area, Oahu-PANKIWSKYJ 249

abundant in the upper part of the ash in this north rim road and nos. 1, 2, and 36 near the location. Several good unconformable contacts administration house on the east side of the between palagonitized tuff below and dark grey crater. These locations are shown collectively as ash above are exposed in a series of cuts be­ l on Fig. 1. The older tuff is probably derived tween Ala Puumalu and Ala Napunani, north from eruptions from Aliamanu Crater, although of the high school site. possibly from Salt Lake Crater. The younger Aliamanu Crater is used at the present time one can be traced continuously to Salt Lake as an ammunition storage reserve by the U.S. Crater. Army. This crater was considered by Stearns Just east of the prominent bend in Salt Lake (Stearns and Vaksvik, 1935, p. 109) as the Boulevard, west of Salt Lake, an amphitheater­ source of the dark grey cinder underlying the like topographic feature suggests an old cone, Salt Lake Tuff, the best exposure of which is in breached on the southwest side, completely man­ the cut at the corner of Puuloa Road and tled by tuff from Salt Lake Crater. This feature, Moanalua Highway (location D in Fig. 1). shown as location R in Fig. 1, is here called the However, there is no proof that some of the Wiliki Cone, after the name of the street along cinder could not have been contributed by erup­ its side. Stearns (Stearns and Vaksvik, 1935, p. tions from other now-completely-mantled vents 109) noted that the topography southwest of in the Salt Lake area, or even from vents higher Aliamanu is suggestive of vents mantled by on the slopes of the Koolau ridges. Similar but Salt Lake Tuff. This area is covered by private considerably coarser cinder is exposed along the homes at this time, and thus such evidence is no north rim road within the Aliamanu Military longer present. Reservation between storage areas no. 3 and Makalapa Crater is topographically the low­ no. 4, shown as location II on Fig. 1. This par­ est, least prominent of the large craters of the ticular cut is directly in line with the crest of Salt Lake group. Along Kamehameha Highway, the Koolau spur which extends seaward from between Halawa Gate in the north to just south Red Hill. The thin water-worked tuff beds near of Makalapa Gate in the south, shown as loca­ the base of this cut closely resemble layers in­ tion K on Fig. 1, are exposed interlayered very tercalated with the red soil at Red Hill, as fine-grained and medium- to coarse-grained tuffs shown in G in Fig. 3. Two tuffs are mapped with xenoliths of Koolau lava and of coral reef above the cinder deposit in Aliamanu Crater. and calcareous sandstone. The coarser tuffs are Though basically very similar in composition undoubtedly derived from Makalapa Crater, and texture, the lower tuff can commonly be dis­ whereas the very fine-grained tuffs may be in tinguished from the upper one. Characteristics part or entirely from Salt Lake Crater. Tuffs of the older tuff are a lack of well-developed below these can be seen within the Pearl Harbor stratification and of sorting and an abundance of Naval Reservation, directly west of Kameha­ angular xenoliths of Koolau lithology. Charac­ meha Highway, shown as location J on Fig. 1. teristic of the upper tuff is a locally developed Cuts J and K are shown in Fig. 3. excellent stratification in the finer grained lay­ Tuffs from Salt Lake and from Makalapa ers, with the bedding planes clearly outlined by Craters are overlain by calcareous sandstone, stringers of coarser grained xenoliths of Koolau reef, and reef breccias in a number of exposures rock which are more resistant to erosion than the south and west of the craters. Good relation­ fine-grained tuff matrix. The imposing cliffs on ships can be seen at the following exposures: the north side of Aliamanu Crater, including 1. In the back of the parking lot of the Puu Leilono, the highest peak, with an elevation Polynesian Vista Apartments at 3019 Ala Ilima, of 485 feet above sea level, are composed of just south of Salt Lake, location WI on Fig. 1; the lower tuff. The entire south and west rims 2. Along the east shore of Waipio Penin­ of the crater are covered by the younger tuff. sula, location PIon Fig. 1; Prominent angular unconformities between the 3. Along the west shore of the Submarine two tuffs are well exposed in the mouths of Base part of the Pearl Harbor Naval Reserva­ storage areas nos. 10, 11, 12, and 13 along the tion, location P2 on Fig. 1; 250 PACIFIC SCIENCE, Volume 26, April 1972

4. Along the ditch parallel to Aolele with root casts of limonite-1/2 foot thick; 3. Street, immediately north of the Honolulu In­ overlain by a bed of fossil oyster shells-1.2 ternational Airport, location Q on Fig. 1. fe~t thick; 4. overlain .by a brown clayey silt Exposure 1 has already been described. Ex­ WithOut apparent beddmg-o.3 foot thick; 5. posures 2 and 3 face one another across the o.verlain by a compact limestone in a single mas­ sive bed containing fossil shells, pebbles and 1,500-foot-wide entrance to Pearl Harbor and numerous .wel1-rounde~ calcareous sand grains are quite similar. Exposure 2 and others on arranged 10 crude lamlOae-5.7 feet thick; 6. Waipio Peninsula figure prominently in discus­ overlain by a hard brownish black carbonaceous sions of the age of the Salt Lake Tuff. l~yer. containinl{ a few calcareous sand grains Stearns (Stearns and Vaksvik, 1935, p. 49­ [Italics by Pankiwskyj]-oA feet thick; 7. over­ 50) described a location near the south tip of lai~ by friable limestone consisting of sand Waipio Peninsula as follows: "... four feet of gralOs and a few fossil shells and containing laminated (air laid) Salt Lake tuff full of molds numerous root casts-2.5 feet thick; 8. overlain of stems and branches of small trees rests on 4 by friable brown clayey silt containing gravel to 12 inches of red soil overlying reef limestone and fossil shells at the bottom-more than 4 feet thick. ... about 100 feet to the south along the coast line the tuff passes beneath sea level, still resting on the soil and containing tree molds .... The The dark carbonaceous layer, number 6 above, is described as a shallow lagoonal de­ tuff at this point is overlain by six feet of emerged reef ... thin seams of soft white lime posit. Stearns also expresses the opinion that a fraction of an inch to an inch thick [ar; the limestone beneath the carbonaceous layer was not far separated in time from the lime­ present] in the tuff ... these seams [are of] secondary caliche ... ." stone above the layer. Rube, Williams, and Hill (1965, p. 493) state that this black layer, col­ Exposures in the ditches along Aolele Street, lected from a different outcrop than the one ~ear the Honolulu International Airport, loca­ described by Stearns, is similar to the Salt Lake tion Q on Fig. 1, contain an almost identical Tuff in X-ray diffraction pattern and in thin sec­ section. At the base is up to 2 feet of brown tion. They further continue that since this layer soil containing decomposing fragments of reef is "interbedded" with limestones which Stearns material. This is overlain by up to 5 feet of recognized as belonging to the Waimanalo fine-grained, pisolite-bearing, air-laid tuff, con­ t~ining upright tree trunks now replaced by stand, the Salt Lake Tuff "must be younger than lunestone. The upper 2 to 3 feet of the tuff is Waipio and at least as young as Waimanalo." interlayered with, and also crosscut by, stringers Hay and lijima (1966, p. 367) concur that this of very fine-grained, pinkish buff, faintly layer, which they refer to as a bentonitic clay­ banded limestone, which very prominently stone, is probably correlative with the Salt Lake crosscut the tree stems. This clearly demon­ Tuff, and that it is "interbedded" with limestone strates the limestone to be secondary, formed and oyster-shell deposits. However, they draw after the deposition of the tuff. The tuff and no conclusion as to the age of the tuff from this tree trunks are unconformably overlain by reef evidence. Moreover, they claim that the Salt and reef breccia. The bands of limestone, which Lake Tuff was not only deposited, but also correspond to the caliche described by Stearns, palagonitized and cemented, with zeolites be­ also are found cutting the reef above the tuff. fore the sea rose to the Waimanalo stand on the Another cut in a bluff on the west side of basis of "coraline limestone with rounded and Waikele Peninsula, one-half mile south of the angular fragments of zeolitic palagonite tuff Waipahu Railroad Depot, now just south of the lying at elevations of 15 to 20 feet near the Waipahu dump, is described by Stearns (Stearns foot of a wave-cut cliH of similar zeolitic and Vaksvik, 1935, p. 53) as follows: palagonite tuff [in the vicinity of location V on Fig. 1 of this paper]. 1. Brown clayey marine silt of uniform tex­ I visited the Waipahu dump site, possibly the ture and without apparent bedding-about 15 same cut examined by Rube et al. (1965), and feet thick; 2. overlain by a brown soil filled found that the one-half-foot dark layer sane!. Geology of the Salt Lake Area, Oahu-PANKIWSKYJ 251 wiched between the two limestone beds contains ation. However, in all grains alteration can be numerous calcareous sand grains as well as noted by the replacements of mineral grains and rounded fragments up to 1/8 inch in size, and the filling of vesicles. In addition, two contrast­ also that it is rich in fragments of plant casts ing types of juvenile particles were observed, in random orientations. There is a pronounced each of which I consider to have solidified at lamination in the rock due to alternating bands different depths. The shallower formed particles of dark "tuffaceous" material with bands rich are composed of glass of any of the colors de­ in calcareous sand. It would thus appear that scribed above, containing various amounts of the layer in question is an interbedded mixture olivine and magnetite. Less abundant, if at all of detrital calcareous sand, a silt which is tuffa­ present, are crystals of melilite, needlelike clino­ ceous, at least in part, and fossilized plant frag­ pyroxene, and nepheline. The deeper formed ments which were possibly washed in rather particles contain fragments of a colorless to very than growing in situ. The tuffaceous material pale green (in section) clinopyroxene (2Vz = may be original infall from an eruption in the 60° ± 1°; Z 1\ c= 44°) showing a reaction Salt Lake area. On the other hand, it may be zone and rimmed by a strongly zoned, honey­ reworked tuff, just as the calcareous sand in colored (in section) clinopyroxene (2Vz = 65° which it is interbedded is reworked reef. In this ± 2°; Z 1\ c = 54°) with dispersion r> v. case it would imply that this tuffaceous layer Other megacrysts present in the deeper formed was reworked in Waimanalo time (provided particles are 'embayed blocky grains of olivine that the underlying and overlying limestones are and smaller, more elongate grains of olivine of Waimanalo age) and not that the Salt Lake which are not embayed. The groundmass is com­ craters were still erupting in Waimanalo time. posed of fine-grained, needlelike clinopyroxene (Z 1\ C = 54°), altered melilite, and devitrified Mineralogy and Petrology glass (?) filled with dust-size opaque material. The tuffs of the Salt Lake area are composed The deeper formed particles were recognized to of particles from several origins: juvenile pyro­ be such because their mineralogy is identical clastic particles belonging to the Salt Lake Phase with that of the dark rims which are found of the Honolulu Volcanic Series; fragments of around most of the ultramafic xenoliths present a melilite-nephelinite lava erupted from one or in the tuff, and also because they have been more of the Salt Lake craters; xenoliths of found forming cores rimmed by material of the medium- to coarse-grained ultramafic material; shallower formed type. Fragments of xenoliths and xenoclasts derived from underlying coral of Koolau Series lithology, coral reef, various reef, sediments, sedimentary rocks, and the Koo­ sedimentary rocks, and of the melilite-nephe­ lau Volcanic Series, as well as fragments of tuff linite lava do not have such deeper formed rims. from earlier eruptions of various craters in the Brief descriptions of the mineralogy and Salt Lake area. petrology of the deposits in the Salt Lake area The juvenile pyroclastic particles are of varied were made by Dana (1849), Hitchcock (1900), mineralogy, primarily as a function of their de­ Stearns and Vaksvik (1935), and Winchell gree of crystallinity. The amount of glass ranges (1947). A more complete description was pre­ from 95 percent to less than 25 percent. The sented by Hay and lijima (1968). I concur with glass ranges from perfectly fresh sideromelane, the description of the Salt Lake Tuff as a lithic­ through palagonite, to a cryptocrystalline devit­ vitric tuff of melilite-nephelinite composition rite. In standard thickness thin section the color containing considerable amounts of fragments of of the glass ranges from chestnut reddish brown tholeiitic material (from the underlying Koolau or pale greyish green to practically opaque black. volcanics), as well as other introduced particles. Alteration from the reddish brown to the pale Stearns (1940, p. 55) mentioned the dis­ greyish green has been observed in a number of covery in a well sunk in Aliamanu Crater of a thin sections, and it is the green that replaces "thirty-one-foot thick dark grey dense nonpor­ the reddish brown. The opaque glass does not phypitic basalt" below Salt Lake Tuff and above lllow one to observe color changes due to alter- alluvium resting on Koolau basalt. He feels that 252 PACIFIC SCIENCE, Volume 26, April 1972

this lava erupted "at the close of the explosive TABLE 1 phase in Aliamanu Crater." The petrographic COMPOSITION OF "ALIAMANU BASALT" description, by G. A. Macdonald, in this report BY PERCENT states that CHEMICAL ANALYSIS

... it is a melilite-nepheline basalt containing Si02 37.10 colorless microphenocrysts of olivine reaching Ti02 2.93 0.3 mm across. The groundmass consists of a A120 3 11.75 network of subhedral grains of altered melilite, Fe20 3 6.39 lath-shaped in cross section, and between them Cr20 3 0.04 euhedral to subhedral grains of nepheline, an­ FeO 9.14 MgO 10.56 hedral grains of augite and olivine, equate grains CaO 11.48 of magnetite, and greenish-brown interstitial MnO 0.23 glass. Highly acicular crystals of apatite are Na20 4.92 abundant.... K 20 1.61 H 2O+ 1.22 H O- I was not able to locate samples from this 2 0.91 P 0 1.23 well, nor thin sections of these samples. How­ 2 5 CO2 0.04 ever, numerous angular blocks within the Salt Cl 0.06 Lake Tuff, especially concentrated in the pe­ F 0.09 riphery of the lake, correspond exactly to this Total 99.70 type of rock. A chemical analysis of one of these NORM (CIPW) blocks was made for the writer at the labora­ an 5.84 tories of the U.S. Geological Survey at Menlo ne 22.15 Park. This analysis, together with a CIPW norm Ie 7.41 and a thin section mode, are presented in { wo 12.41 Table 1. di en 8.60 fs 2.77 12.46 01 { fo Ages of the Rocks in the Salt Lake Area fa 4.59 Eruptions during the Pleistocene epoch in the a ~50 Salt Lake area span times of sea level consider­ mt 9.28 il ,3.65 ably higher than at present, followed by other ap 2.69 levels considerably lower than the present, and MODE (FROM THIN SECTION) finally one level about 25 feet higher than the Nepheline 70 present. These have been correlated by Stearns Melilite 15 (in Stearns and Vaksvik, 1935, p. 127-128) Clinopyroxene 6 respectively with the Kaena, Waipio, and Wai­ Glass 4 manalo stands of the sea about the island of Magnetite 3 Oahu. On a worldwide scale, these three are Olivine 1.5 Apatite 0.5 correlated in time respectively with the Yar­ mouth Interglacial, the Illinoian Glacial, and the Sangamon Interglacial (Stearns, 1966, p. in three separate runs (Gramlich, Lewis, and 23). The dates of these are still a matter of con­ Naughton, 1971). If this rock can be correlated tention. One time scale, by Ericson, Ewing, and with the 31-foot-thick flow at the base of the Wollin (1964, p. 731) gives the end of the Salt Lake Tuff below the present-day bottom of Yarmouth at 420,000 years B.P., the end of the Aliamanu Crater (Stearns, 1940, p. 55), this Illinoian at about 340,000 B.P., and the end of dates the period of eruption in Aliamanu Crater, the Sangamon at about 120,000 years B.P. from which the water-laid cinder below the Blocks of fresh melilite nephelinite collected Salt Lake Tuff may be derived. The water re­ in the Salt Lake Tuff yielded a K40/ A40 age of working of the cinder attests to a higher level 430 -I- 25, 446 -I- 2, 418 -I- 18 thousand years of the sea than at present, and the 400,000 to Geology of the Salt Lake Area, Oahu-PANKIWSKY] 253

467,000 year age places this comfortably during 1840, 1841, 1842. Under the command of the period of the Yarmouth Interglacial, as esti­ Charles Wilkes, U.S.N. Vol. 10. G. P. Put­ mated by Ericson et al. (1964). nam, New York. Coral reef terraces rising to the Waimanalo ERICSON, D. B., M. EWING, and G. WOLLIN. level have been dated by Th23o/U238 methods. 1964. The Pleistocene epoch in deep-sea sedi­ Four samples from Oahu from near Nanakuli, ments. Science 146:723-732. Kailua, Kahuku, and Haleiwa gave ages of re­ GRAMLICH, JOHN W., VIRGINIA A. LEWIS, and spectively 120 -1- 30, 110 -1- 20, 110 -1- 20, and JOHN J. NAUGHTON. 1971. Potassium-argon 140 -1- 30 thousand years (Veeh, 1966, p. dating of Holocene basalts of the Honolulu 3383). The calcareous sandstone capping Salt Lake Tuff is correlated with the Waimanalo volcanic series. Bull. Geol. Soc. Amer. 82 (5) : stand of the sea. Thus the date for the youngest 1399-1404. eruptions from Salt Lake Crater predates 90,000 HAY, R. L., and A. II]IMA. 1968. Nature and to 170,000 years B.P. origin of palagonite tuffs of the Honolulu Group on Oahu, Hawaii. In R. R. Coats, R. L. ACKNOWLEDGMENTS Hay, and C. A. Anderson, Studies in volca­ I thank Dr. G. A. Macdonald for spending nology. Mem. Geol. Soc. Amer. 116:331-376. a day in the field with me and for reading the HITCHCOCK, C. H. 1900. Geology of Oahu. manuscript and making suggestions. Any errors Bull. Geol. Soc. Amer. 11:15-60. in the interpretation of field data, however, are RUHE, R. V., J. M. WILLIAMS, and E. L. HILL. my own. I appreciate the discussions and the 1965. Shoreline and submarine shelves, Oahu, time spent in the field with Dr. E. D. Jackson, Hawaii. J. Geol. 73:485-497. who also provided me with the chemical analysis STEARNS, H. T. 1940. Supplement to the geol­ reported in Figure 2. Mr. D. Song provided me ogy and ground-water resources of the island with interesting and useful spoken communica­ of Oahu, Hawaii. Bull. Div. Hydrog. Terr. tions. Permission to enter the Pearl Harbor Haw. 5. Naval Reservation, the Aliamanu Military Res­ ---. 1966. Geology of the state of Hawaii. ervation, and the Hickam Air Force Base was Pacific Books, Palo Alto, California. 266 p. granted by the United States Navy, the United STEARNS, H. T., and K. N. VAKSVIK. 1935. States Army, and the United States Air Force, respectively. Permission to work on the grounds Geology and ground-water resources of the of the Dan10n Estate was granted by Mr. H. E. island of Oahu, Hawaii. Bull. Div. Hydrog. Damon; and the area about Salt Lake by Mr. Terr. Haw. 1. 230 G. S. Hong, attorney for the International De­ VEEH, H. H. 1966. Th /U238 and U234/U238 velopment Company. The illustrations in this ages of Pleistocene high sea stands. J. Geo­ paper were drafted by Mrs. Jan N. Hunt. phys. Res. 71: 3379-3386. WENTWORTH, C. K. 1926. Pyroclastic geology LITERATURE CITED of Oahu. Bull. Bishop Mus., Honolulu 30. DANA, J. D. 1849. Geology. liz U.S. Exploring WINCHELL, H. 1947. Honolulu Series, Oahu, Expedition. During the years 1838, 1839, Hawaii. Bull. Geol. Soc. Amer. 58:1-48.