PHOTOGRAPHIC INTERPRETATION OF THE STRATIGRAPHY OF BLOCK ISLAND, RHODE ISLAND

Application of standard geologic photointerpretive techniques to a series of offshore photographs of the Block Island cliffs has shown signifi~ant correspondence with previous stratigraphic field studies. Constant scale while mapping and a vertical exaggeration for the final profile were achieved with the aid of a Zoom Transfer Scope. Photogeologic unit boundaries were defined by: the nature of bedding visible in.the photographs, the extent of different erosional and drainage patterns on the cliff face, changes in texture of the cliff face, tonal variations, and variations in clast size. Five photogeologic units have been defined on the northern cliffs of the island by using the above criteria .. Their boundaries correspond closely with those of units defined in the field by previous workers and observed in field work for this study. These include a basal outcropping Cretaceous unit, the Raritan Formation; and Pleistocene units equivalent to two members of the Montauk Drift of Sirkin (1976) and the New Shoreham Outwash and New Shoreham Till of Sirkin (1976). Nine photogeologic units, representing six depositional stages, were identified on the southern cliffs of the island. These six stages are equivalent to the three members of the Montauk Drift of Sirkin (1976), his New Shoreham Outwash and Till, and channel gravel deposits laid down during the: f·:nal stages of glacial retreat (Sirkin, 1976). •. Hhere Sirkin provided unit thicknesses or outcrop locations, there is also agreement with photointerpretive results.

The New England Islands, which include Long Island, Block Island, No Mans Land, Marthas Vineyard, and Nantucket, are cored by a cuesta of Cretaceous sediments (Johnson, 1925;Fennemnn, 1938;McMaster and others, 1968). This cuesta is cut by several preglacial river systems running southward to ancient shoreline positions on the continental shelf (McMaster and Ashraf, 1973). This is overlain in places by Tertiary deposits (Woodworth and Wigglesworth, 1934) and capped by Pleistocene glacial sediments.
the Pleistocene stratigraphy and palynology.
STRATIGRAPHIC r1APPING BY PHOTOGRAPHY Although remote sensing techniques have been used for geologic mapping from aerial photographs (Ray, 1960;van Bandat, 1962;American Society of Photogrammetry, 1966, 1968, 1975Avery, 1977}, 1 ittle attention has been paid to the value of ground photographs.
As photographtng the entire geologic section does not require detailed study at the time, the worker can cover lc.rge areas of the cliff in a short time. Later, by mosaicing the photographs, he can see larger areas than might be visible in the field, and with minimal distortion from offset in the section. In addition, he can review two or more sections which may be geologically separated by using the photographs, which a field worker would not have available.
For later, detailed studies, the geologist need only return to the photographs to remap the section in the required detail. In addition, he can easily map a section without the distortion induced by viewing an outcrop from one limited vantage point. In coastal studies, cliffed areas which are relatively inaccessible can be photographed from a distance offshore or from a plane.
The field time required to collect data for initial mapping is less for a photographic survey than for an equivalent on-site study.
The Block rsland photographs were taken in one day from offshore, while even cursory field studies of the cliffs, which must be done on foot due to a lack of roads at the base, require two days field time (Hallick, 1896;Bierschenk, unpub.).

AVAILABLE GROUND DATA
Ground data is any information collected on the ground, or derived from that data, to aid in the interpretation of remotely sensed data (American Society of Photogrammetry, 1975 worth and Wigglesworth, 1934;Hansen and Schiner, 1964;Sirkin, 1976).
These data could be plotted on scale sections of the cliffs and used as standards against which to measure the detail recorded in the photo-graphic study. There also exist a number of brief studies, descri"bing, but not tllustrattng, specific sections, and numerous articles and books on the regional geology which were used to provide supplementary data to aid in the interpretation.

BEDROCK GEOLOGY
The crystalline bedrock, as exposed on the Rhode Island mainland, is primartly Paleozoic and Mesozoic granite or gneiss (Quinn, 1971). The metasediments of the Narragansett Basin may pass just to the east of Block Island (Woodworth and Wigglesworth, 1934;Tuttle, Allen, and Hahn, 1961). The bedrock surface on the mainland has been eroded to form an anc,~ent peneplain (Davis, 1895) which dips to the south at five to ten meters per kilometer (Woodworth and Wigglesworth, 1934). Seismic work on the island showed the crystalline bedrock to be at a depth of about 330 meters at the north end of the island (Tuttle, Allen, and Hahn, 1961). Tuttle, Allen, and Hahn (1961) (Fuller, 1906(Fuller, , 1914Hoodworth and Wigglesworth, 1934;Fetter, 1976). The unconsolidated sediments on Block Island are between 150 and 200 meters thick (Tuttle, Allen, and Hahn, 1961), the upper portions of which are Known from well logs (Hansen and Schi'ner, 1964) and outcrops (Livermore, 1877; Woodworth and Wigglesworth, 1934;Christopher, 1967;S.irkin, 1976}~ (Woodworth and Wigglesworth, 1934). Several boulders found near Clay Head contained fossils characteristic of the Calvert Formation of Maryland and Virginia, and regionally unique to this occurrence (Shimer, 1916;Woodworth and Wigglesworth, 1934). No associated strata are found nearby, so the boulders presumably weathered out of the till (which im-plies an as-yet unknown source to the north) or served as ballast for some boat (Shimer, 1916).

PLEISTOCENE GEOLOGY
By far the greatest amount of the exposed sediments on the island are Pleistocene. Upham (1879) Woodworth and Wigglesworth (1934) and Fuller (1914), but did not attempt to relate directly his work to theirs.
Sirkin (1972, 1976}  and this complexity has resulted in various contradictory interpretations of the stratigraphic sequence. Some of these will be discussed later. Shaler (i888, 1894, 1897Shaler and others, 1896) reported highamplitude, sharply compressed folds of considerable size in the Tertiary and Cretaceous beds of Marthas Vineyard. He also cited communications from Woodworth (Shaler, 1897) indicating similar features existing on Block-Island. Shaler considered these folds to be the result of regional orogenic activity rather than of glacial deformation for two reasons.
One was that he saw a well-developed preglacial topography superimposed on his folded sediments. Secondly, he observed that the apparent direction of applied stress on the folded sediments was at 90° to the direction of apparent motion of the ice sheet.
Upham considered some of the deformation in the lower beds on Marthas Vineyard to be the result of ice thrusting of the frozen sediment. He ascribes much of the volume of Shaler's "preglacial erosional remnants" to glacfofluvial deposition in contact with the ice sheet.
Shaler considered the southern margin of the ice sheet to have been very thin and, in fact, flof!ting en the sea surface between supporting points on his erosional remnants; thus it would not be capable of deforming sediments. Upham, tn contrast, consfdered the sea had retreated out to the conttnenta1 slore, and that the ice sheet was much thicker and more widely spread than did Shaler, and therefore more likely to cause deformatfon.
Woodworth (1897) also considered the folding and thrusting on Block IsTand and Marthas Vineyard to be ice-pushed_ features, but he felt that positive evidence either way was lacktng, as the deformation, even i·f by gl acia 1 tee movement, had occurred before the deposit ton of the glactal sediments.
Kaye (1964a} stressed the thrust-faulting and folding found on Marthas Vfneyard and cons.idered the deformation to be of glacial origin.
He considered the thrusting may have resulted in displacements of up to several miles, and stated that this complicated interpretation of the secttons as it was difficult to determi'ne whether differences in adjacent deposits were from faulting or some other factor. The photographs were scanned twice, with details being transferred to tracing paper at a constant horizontal scale of 1:1,000

METHODS OF
and 2x vertical exaggeration. The first mapping was for the purpose of noting lineations and the lithologic variations, while the second was to determine apparent unit boundaries.
Constant scale plotting was achieved by vary1'ng the photo magnification so the cliff height, as determined from the U. s. Geologic Survey topographic map of the island (U. S. Geologic Survey, 1970} was to scale.
At least three control points per strip were used to provide the maximum posstble accuracy.
Print enhancement by unsharp masking, a well-known remote-sensing technique (American Society of Photogrammetry, 1975; r 2 s, undated; Mirkin and others, 1972) was employed in apparently featureless or confused areas to help disttnguish lineations. Duplicate negatives of increased contrast range, made for use in the unsharp masking process, were also printed directly to provide improved tonal discrimination between units.
These two techniques are applicable to different parts of the problem as unsharp masking is considered to provide increased line enhancement, while the increased tonal range achieved by printi'ng the duplicate negatives without masks simplifies the discrimination of areas of different tonal values. As the preparations for the two processes are the same, both techniques were used where any enhancement was needed, the informatton from the resultant prints being transferred directly to the paper section with the Zoom Transfer Scope.

PHOTOGEOLOGIC MAPPING INTERPRETATION
Lithologic di-fferences were determined by tonal variations, textural vartations on the photographs, differences tn erosional features and drainage patterns on the cliff face, the nature of bedding, and to a lesser extent, by the topography as determined from air photos and the U.S. Geologic Survey topographic map of the island.
Tonal variattons within a single photograph, but not across photo boundaries, were considered. This was done to minimize errors due to  (Ray, 1960;van Bandat, 1962; American Society of Photogrammetry,. 1966Photogrammetry,. ,· 1975Avery, 1977), changes in drainage patterns were considered significant indicators of lithologic changes.
Hoodoos developed in massive units composed of dominantly fine sedtments, while fine· parallel drainage was most apparent in coarse sediments or areas of recent slumping, presumably poorly compacted and relatively permeable, thus retarding growth of the drainage channels.
Both of these features can be considered variants on the vee-shaped gullies, as adapted to their special lithologic conditions. The time required to develop these features is probably also variable, with hoodoos and large vee-shaped gullies requiring the longest time to develop, and parallel drainage representing a short-term feature in areas of recent Bedding and structures were defined on the basts of tfie number of beds per five meter vertical sectton of cliff. Ftne-bedded sections had from three to twenty beds per five meter vertical section (twenty beds per ftve meter section represented the limit of clear resolution}, In pl aces,. the texture appeared to be that which would be expected from sharply contrasting bands at a spacing just beyond the limit of resolvable detail. "Fine bedded" was chosen as it has no ffold sedimentological connotations, and therefore would be 1 ess 1 H.ely to be misconstrued than more traditional terms such as "thin-bedded." Between one bed per ten • meter section and three beds per five meter section, units were considered to be thi'ck-bedded, while beyond that point they were considered massive. Massive units in this case could include units wi-yt bed thicknesses of thtrty cm or less (medium-bedded or below, Blatt, Middleton, and Murray, 1972), which were too fine to be distinguished.
If the units, or portions thereof, appeared resistant, this was also noted, as a possible indication of thrust planes or coarse layers cemented by iron oxide deposits from groundwater action.
Topographic expression, studied with the aid of stereo paired air photo coverage of the island, and with the aid of a conventional 7½ min topographic map (U. s. Geologic Survey, 1970) was plotted for those areas adjacent to the cliffs to aid in the delineation of units present.
The possible influence of geology on topography was suggested by various authors (Woodworth and Wigglesworth, 1934;Merrill, 1896).

FIELD STUDIES
To asstst tn both the photogeologic mapping and interpretation, geologic field studies were conducted throughout southern New England and Long Island. In these areas, the type sections of several of the major units in the area were studied, and tnterpretations were discussed with recent v10rkers. These field studies included one on Long Island which included di•scussion by Dr. L. A. Strkin, whose Block Islarrd paper were then compared with the interpretations of the three most informative work avai1a.b1e on the island's geology (Woodworth and Wigglesworth, 1934;Hansen and Schiner, 1g64;and Si'rkin, 1976)  The third glacial unit found is the Jameco Outwash, which lies unconformably on the Nebraskan sediments. Its basal member is a boulder bed locally cemented by iron oxides and outcropping on Clay Head. This   (Woodworth and Wigglesworth, 1934, p. 39, 52) describes it as a late Sankaty (Yarmouth) sand laid do11m by a retreating sea, and twice (Woodworth and Wigglesworth, 1934, p. 52, 220) identifies it as a glacial gravel and sand.
The next unit, the Manhasset Formation, ts the fourth glacial unit.  Hansen and Schiner (1964)  .. : SIRKIN, 1976 Sfrktn (_19761 describes, only two glacial fonnations, the Montauk Formation and the New Shoreham Drtft, and isolated occurrences of a presumed interstadial unit, the whole column being capped by scattered occurrences of channel gravels, peat bogs, and lake deposits ( fig. 6, table 21.
The first fonnation, the Montauk, is of early \Hsconsinan age, and extends below sea level wherever it occurs on the island, alt~ough The second glacial unit is the rlew Shoreham Drift. This forr,ation Photogeologic unit boundaries were determined in the cliff photographs by tonal and textural variations, differences in erosional and drainage patterns, and the character of the bedding. Unit boundaries were not plotted for cliffs less than ten meters htgh as they did not provide sufficient continuity. Cliffs of that height should provide no difficulty to the field worker in any event.
Tonal, erosional, and drainage differences were used to determine Textural variattons were of primary importance in determinfog grain stze. Texture is an indication of detail below the resolution of the film. Units of smooth or velvety texture were consiclered fine-grained (less than 2 mm diameter). Units with coarse texture were considered to contain sediments of betvreen 4 and 128 mm diameter, while particles of greater than 128 mm diameter were considered to occur in areas of extremely coarse texture. Particles in this last class were normal_ly visible as distinct objects, although at the limits of resolution of the film used.
Bed thicknesses, as discussed previously, were fine (three to twenty beds per ftve meter interva,11, coarse Cone to six beds per ten meter secttonl, and mas.sive (less than one bed per ten meter section}.  whi'le to the west it averages two meters thi'ck. Displacement along the fault is two to three meters. This unit was only photographi'cally studied. Unit J Cftg, 9b I, foynd at th.e base of Barl ows Point, ts up to twenty-two meters thick, and consists of fi'ne parallel beds and massive b-eds. I't has a re.ugh surface and texture, and grain size probably ranges from fi'ne to coarse, with clast stzes up to twenty or twenty-five cm.

Descrtpttons apply only to the locattons referred to. Changes in
Only photoi'nterprctation \'/as used to describe this untt. Only photographs were used in its stud~,.
Unit N ( fig. 11), also from Vaills Beach, has massive bedding, a rough, resistant surface, and it appears to contain material wi'th a size range from ftne materials up to clasts of twenty cm dtameter. It was studied from photographs only. A occurs as an asymmetrical hump wtth the gentler slop~ to the north. It ts ove.rlatn by unit B·, whi'ch is strongly contorted at the southern end of the exposure and dips• steeply to the south there. Uhi't C, which ts flat .. lying above these two untts, appears to fJe extremely crumpled just south of them and at the same level. This suggests that units A and B may be the ttp of a thrust plate, emplaced by tee-shove of the frozen . ground, as has been observed in Long Island (Sirktn and Mills, 1975} and Marthas Vineyard (Kaye, 1964a. Relocating Hansen and Schiner's (1964) contacts to positions closer to those reported by Woodworth and Wigglesworth (1934) and Sirkin (1976) yields results which agree with those of the other workers, and which are further confirmed by this study.
North of Balls North Point, however, Hansen and Schiner show a till unit which, starting from the top of the cliff, dips to the south at 9° until ft jotns up with the unit A equivalent at Balls North Point. This unit is not visible in the photographs, which plainly show untt C, the only untt in the area, either fl at-lying or dippfog to th·e north at a very low angle, wtth bedding planes vtsi·ble throughout the section.
Hansen and Schiner's (1964r i'nland sectfon in this area (fi'g. 5, section E-P) shows the untts t-n that area a 11 dipptng to the north. As the control for lithologic changes from the well logs (measured from the surfacel has no perspective problems and provtdes better opportunities for exami'ntng th.e lithologies of the· unt-ts involved, and as all other data suppot'ts the probability of north-dipptng units, tt would seem that the upper portion of Hansen and Schiner's (1934} sections, at least, should be viewed with caution. The till Hansen and S'chiner (_1964) show at the base of the cliff tn the northern Clay Head area corresponds to a possible outcrop of unit A under the thrust plane in the same araa; _,,unfortunately, fan deposits at the base of the cliff prevent tracing this contact on the photographs.
The depostts on the northern cliffs mapped in this study and by prevtous workers show a close correspondence. The boundartes of the five uni-ts were photogeologically mapped first and then were observed to be.  (Strki'n, 1976) and the Nantucket Ground Morai'ne (Woodworth and Wigglesworth, 1934)   The results of this study support Strkin's (1976)  Si-rktn has channel gravels mapped.
Predictions on the northern and southern cliffs were borne out by field checks.

APPENDIX 1 -RECOMMENDATIONS
Th i's study was-carried out using existi'ng bl acf<: and white photo-. graphs. In future studies, if funds and ttme are aval'lable, preliMinary testtng should be carrted out to determtne the most suitable film and filter combtnatfons for resolutfon and photogeologic unit discrimtnatfon for a parttcular type of depostt.
To make full use of stratigraphic photointerpretation, photo coverage should also be obtained i'n color, whether by use of a color·.sHd"e or negative film or by use of a tri'color separation process. Color slide and negattve films are generally of poorer resolution and coarser grain than black and white films, so use of the tricolor method ts preferable i'f opttrnal resolution ts required. It has the disadvantage of requiring either more cameras or ~ore complex equtpment and some means of viewing or printfog the combined separation negati~es, but penntts, by addition of a fourth camera, procurement of "false color" infrared data and finer disttnctton between units by selective filtration. There are presently several multi-band or multi-camera systems available for this type of work. Offsetting the advantages of the tricolor method to some extent is the convenience of the color transparency or negative system with only one camera and film to be handled, a significant help both in operation of the system and in viewing the final product.
Color negatives are somewhat less convenient than transparencies as prints must be made before the sectton can be viewed with any facility.
In additi"on, paper prints have i'nherently 1 ess detai'l than images viewed by, transmitted ltgnt, Th.tsts trequs~ tn vi~wing an tmage· by reflected 1 tgf1t the· Hght ray must pas-5' through· the fmage-bear.i-r.s errul s-ion twtce before reaching the-eye, whereas an image vi"ewed by, transmttted ltght onl,y requtres the li'ght ray to pass through the emulston once. Thus for the same image dens.ity, more detai"l wtll be visible i"n a transparency than tn a print, or, gtven the-same_ range of visible detatl, a transparency wi-11 have more contrast. Transparencies may be vtewed directly on a Zoom Transfer Scope or other camera lucfda, or prints may be made and assembled into mosaic form if this is desired. Unmounted prints or transparencies are desirabTe for areas of great reli'ef, as they can be viewed under stereoscopes, thus pennittfog more accurate models to be made.
In securing the initi'a 1 photographs, a means of keeping the. range To ensure continuity of coverage_ b'lcr cameras-should be available,

S9
or interchqn~eable m~gazine.s tf the camera useq has thjs feature, Th.is .. wN1 permt,t drangtng ft,lm at the·end of tn:e-·ro11 w-i'thout 1eavi'ng any gaps tn the· coverage. rn 3S-mm cameras-, where the-fi'1m must EYe rewound after each roll, or tn aeri"al cameras, which must b"e unloaded and reloaded tn total darkness, thts ts a disttnct prob·1em. • rn ana lyztng the photographs, stereo i·mages-may prove helpful, enahling resolutton_ of detai-1s which may otherwtse be lost in rubble, vegetatton, or other masking matertal. Wtth practice, the observer can also esttmate slopes of resistant faces or loose sediments, th.e former . gtving the idea of the relattve· reststances of the untts to erosion and the latter indtcating the grain size of the sediments present.