Phreatic Eruption Clouds: the Activity of La Soufri6re de Guadeloupe, F.W.I., August - October, 1976
G. I-IEIKEN B. CROWE T. MCGETCHIN F. WEST J. EICHELBERGER Los Alamos Scientific Laboratory Los Alamos, NM 87545 D. BARTRAM R. PETERSON tC WOHLETZ
ABSTRACT with an approximate velocity of 10 to 25 m]sec. Upon reaching the sea the clouds continued to From August to October, 1976, La Soufri6re move forward, but at a decreased velocity, and de Guadeloupe was observed, and recorded spread laterally, having left behind the restric- with an automated sequence camera and nu- tions of valley walls. A thin gray veneer of merons handheld cameras. During the period moist tephra, ranging from several cm thick of observation, the nature of volcanic activity near the dome to less than 1 mm thick several ranged from mild steam emission to moder- km downwind~ was deposited along a narrow ately energetic phreatic eruptions. corridor southwest of the summit. Tephra from Background fumarolic activity (steam emis- the phreafic eruptions consisted mostly of hydro- sion) was characterized by the emission of gen- thermally altered lithic, mineral, and glass erally tephra-free steam clouds 50 to 150 m fragments derived from dome lavas; no fresh above the suramit. The clouds rose buoyantly (juvenile) pyroclasts were present in the above the vent and were blown downwind at tephra. prevailing wind velocities. Phreatic eruptions Absence of juvenile tephra at La Soufri6re were well-documented on September 22, Octo- supports the view that activity was due to ber 2, and October 4. In the latter two erup- groundwater circulating in a vapor-dominated tions, small bursts of tephra-laden steam erupted geothermal system, probably driven by a shal- at intervals of 30 to 45 min. and rose from 350 low heat source. At La Soufri6re, most vapor- to 500 m above the summit. In the largest dominated systems are located in elevated observed eruption, that of October 2, the steam areas of groundwater recharge where ground- and tephra cloud rose to a maximum height of water movement is downward and outward. 600 to 650 m in 20 rain. A white vapor cloud The sporadic phreatic eruptions may be relat- and a medium gray, tephra-laden cloud were ed to the rate of recharge of meteoric waters erupted simultaneously from the summit vent within the dome, the decrease in pore pressure and both were surrounded by a vapor collar; during fortnightly tidal minimums or both. the clouds were thoroughly mixed within 1 km Whatever the triggering mechanism, vapor- downwind of the summit. The concurrent dominated fluids eroded vent walls during phreatic growth of clouds from separate vents (summit eruptions and carded out fine-grained, hydro- and flank) implies a common source. Simulta- thermally altered, pre-existing dome material neous eruption of tephra-free and tephra-laden as tephra. clouds from the same vent is puzzling and implies: (i) lateral changes in the degree of al- teration of dome rocks along the elongate INTRODUCTION vent, hence erodability of the dome lavas, or (ii) differences in the gas velocities. These (
Bull. Volcanol., Vol. 43-2, 1980. 384 HEIKEN - CROWE - MCGETCHIN - WEST - EICHELBERGER . BARTRAM . PETERSON - WOI-ILETZ
I ATLANTIC or Basse-Terre, part is composed of NNE- OCEAN trending chain of Post-Miocene volcanoes. - 16= 30' N The (
I6"N T.,,. 0.,°.,. RECENT ACTIVITY
0 20= km Historic eruptions, as s-mmarized in I ROBSON and TOMBLIN (1966), have been 61 '= 30' W of smsil magnitude consisting of phreatic explosions and f11mArollc activity,and the FIG. 1 - Location map, Island of Guadeloupe. generation of small lahars. The last Pe- dies (Fig. 1). It is composed of an older 16ean activity occurred between 1277 and volcanic massif capped by a steep-sided 1577 A.D. (C14 date of 550 _ 150 years - dome that is 700 m in diameter and rises BRUET, 1953). Most of the activity has approximately 470 m above the massif been concentrated along the NW-SE surface (ROBSON and TOBLIN, 1966; LA- trending fracture system on the SE side SERRE, 1961). La Soufxi~re, the highest of the dome. Some of the past activityhas point on the island (1,467 m), towers been observed near the summit at above the populated regions, the largest Gouffres Dupuis and Tarissan (ROBSON being the towns of St. Claude, Basse- and TOMBLL~, 1966). Terre and Capesterre. During summer During October, 1956, steam and gray and aut~]mr~ 1976, increased seismic activ- tephra were erupted from a new fissure on ity and phreatic explosions at La Soufri~re the southeast slope of the dome. Large provoked the evacuation of most of the in- blocks were erupted onto the surface near habitants of the region and marked the the vents and a thin layer of fine-grained beginning of multidisciplinary studies of tephra was deposited west of the fissure the volcano by scientists from France, along a 12 km long, 2 km wide area (BAR- U.SA. and Trinidad. The objectives of the RAB#. and JOLIVET, 1958). About 105 m ~ of small group from the Los Alamos Scientif- tephra were erupted between October 19 ic Laboratory were: (1) Document and and October 24. The 1956 activitywas, in study eruption clouds, (2) Sample and most ways, similar to the activity describ- characterize tephra, (3) Measure the tilt of ed in this paper. the volcano, both as a predictive tool and to analyze ground deformation associated TECHNIQUES USED IN THE STUDY with the phreatic activity. This paper pre- sents the results of the eruption cloud The phreatic activity of La Soufri~re study; our results from the tephra and tilt was recorded with a stationary automated studies will be presented elsewhere. sequence camera and numerous hand- held cameras. (These are described in the Appendix). Although photographic studies GEOLOGIC SETTING were hampered by bad weather, a number of freatic eruptions as well periods of The Island of Guadeloupe is one of a typical fumarolic ac~vity were well docu- line of primarily volcanic islands that form mented. Formulae used in analyzing vent the Caribbean Arc. The island is divided velocities are detailed in the Appendix. into two major parts (Fig. 1). The north- Tephra from each of these eruptions, and eastern part is vnRirdy an uplifted lime- may more not photographed, were collect- stone platform, whereas the southwestern, ed for petrologic analysis. PHRF_~TIC ERUPTION CLOUDS: THE ACTMTY OF LA SO~ DE GUADELOUPE, ETC. 385
Table 1 - Activity Observed during the or, coinciding with increased tephra emission. Eruption Cloud Studies Tephra emission from phreatic explosions for 20 to 30 m~n_, depositing 4 mm thick layer at Aug. 30:1040 AST. Phreatic explosion carried summit car park. steam and tephra 2 km downwind from the Oct. 11: All day; quiet htmarolic activity. summit. Blocks fell on dome and near-summit car park. No photographic data; eruption clouds rnlued with atmospheric clouds. Explo- sions followed a period of increased seismicity. BACKGROUND Small lahars generated. FUMAROLIC ACTIVITY Sept. 14:1922 AST. Phreatic explosion and rock avalanche. Vegetation was flattened Background fumarolic activity occurred and]or defoliated over an 800 m long, wedge- nearly continuosly between phreatic ex- shaped area. Wind directed air fall to sea plosions during the period of observation. near Vieux Habitants. Not photographed or Several days of this fumarolic activity observed. were photographed with the Automax Sept. 22:0619-0627 AST. Phreatic explosion camera. A typical synthesis of these ob- with ash cloud rising 3500 m above vent. Docu- servations is presented here to character- mented from 0620 to 0630. Tephra-fall to west; ize the activity. 3 mm of tephra at geophysical observatory (Parnasse). Explosive phase ceased by 0645. Pale gray or white steam clouds rose buoyantly from vents located at the sum- Sept. 23: 1200. Mild fumarolic activity observ- mit and along the southeastern fissure to ed. a height to 20 to 100 m. The long axes of Sept. 26: 0602-0700. Mild fumarolic activity visible clouds extended downwind for dis- with very small phreatic explosion observed at tances of 250 m to several km (Fig. 2). 0635. Generally the clouds were wind-sheared, Sept. 28.'-1200. Mild activity, with small flowing to the west and down the valley of phreatic explosions. Tephra fall within 1 kin, the Rivi~re Noire. The smell of sulfur dispersed steam and tephra cloud moving compounds was detectable for 3 or 4 km down the Rivi~re Noire. downwind, although no clouds were visible Sept, 29: 1730. Quiet fumarolic activity photo- at that distance. The clouds were either graphed. continuous or broken into sections 40 to Oct. 2: 1350-1750. Fumarolic activity. Major 85 m long. Although very little tephra was phreafic explosion starting at 1638. The tephra- entrained during the background activity, laden cloud rose - 630 m above the summit. small visible tephra-falls were occasionally The clouds flowed down the valley of R. Noire carried for 100 to 350 m downwind. and across the sea. Activity had subsided by After small phreatic explosions, the 1750 when observation was terminated by cloud changed from white to a medium darkness. gray (presumably a reflection of the Oct. 3: 1100-1300. Fumarolic activity produced tephra content) and rose to 2 or 3 times a very dilute cloud that flowed west, along can- the height to the steam clouds of the yons radial to the summit. background fumarolic activity. In windless Oct. 4:1100 AST. Marked increase in fumaro- conditions, especially during the evel~ngs lic activity. 1150 AST Phreatic eruption, with just after sundown, the steam plumes rose cloud rising to -160 m above the summit~ as thin vertical columns from 500 to about flowed west. Tephra fall over summit and along 1000 m above the summit. the valley of R. Noire. 1245 AST; Phreatic ex- plosions. Lahars developed, flowing down dome flanks. 1618-1720: Numerous small phreatic explosions, with clouds Hsing to 250 to 300 m OBSERVED PHREATIC ERUPTIONS above the s,,mrnit. Oct. 8: 1200. Mild fumarolic activiLy. The most completely documented erup- tions were on October 2 and October 4. Oct. 9: 0800. Mild fumarolic activity observed. The eruption of October 2 was recorded Oct. 10. 1111 AST; 1 m, 2 sec.harmonic trem- by the Automax camera and by observers 386 m~ - CROWE - MCUGETCHIN - WEST . EICHELBERGER - BARTRAM - PETERSON - WOHLETZ
FIo. 2 - Mild phreatic activity, September 28, from the location of the Automax camera on the Plateau du PAlmlstes, 4 km SE ofLa Soufri~re, The main (summit) vent, that erupted steam and a small amount of tephra, is to the left. The pale gray streak left (southwest) of the summit consists of tephra-fall and laharic breccia. The plume to the right is rising from a fissure vent on the south- east flank of the dome. located at Vieux Habitants and Marigot. ing westward, downwind, for d/stances of On October 4 the activity was photograph- 40 m to 1 km Tmmediately after an explo- ed by the AutomaY camera and observers sion, the clouds were dark gray and located at Basse-Terre and near the heavily laden with tephra. Tephra falls summit. were visible for several hundred meters downwind of the vent Oct. 2, 1976: At 1638 hrs., a large phreatic explosion lasting 7 minutes took place. Clouds rose Throughout most of the day, there was to 623 m above the s.mmlt vent and 452 fumarolic activity, with several small phre- m above the southeast fissure. Based on atic explosions. The activity was at two analyses of the Autem.x camera photo- vents; the summit (Gouffre Tarlssan) and graphy, this event was over by 1657 hrs., on the southeast flank along a new frac- with fumarolic activity continuing until ture system (near the Cratere du Sud of dark (1750 hrs.). FEUmT.ARD et al.,) (SEE Fig. 2). Observers at Vieux Habitants saw the The maximum height of the eruption beginning of the phreatic eruption at 1638 clouds above the vents (A/4) is described hrs. and followed the development of the in Fig. 3, along with the estimated escape eruption cloud during its westward move- velocities of gases at the sl]rnrnit vent ment to the sea. The eruption cloud (w0). from the summit vent, the only vent vis- During background fumarolic activity, ible as the flank vent was concealed be- cloud heights at both vents ranged from hind the dome, had two distinct parts; to 60 to 100 m. During phreatic explosions, the southeast it was medium gray and to clouds at both vents increased in size con- the northwest it was white. Both were currently (Fig. 3), reaching elevations of mixed and moved downwind. 200 to 500 m. The clouds were sheared off When the cloud reached a height of by winds passing over the S.rnmlt, mov- about 600 n~ the buoyant rise was over- PHREATIC ERUPTION CLOUDS: THE ACTIVITY OF LA SOUFRI~A~E DE GUADELOIrPE, ETC. 387
TIME (OCT. 2,1976)"-'-~ 1400 1500 1600 6T31700~m 600 i I I I I I I I i i i I I I I i I t I I 1 1 I t AH(m) WEST (CENTRN.) VENT 30O I %......
l 600 I J I I i I i t I i I I I I i t I I i T I i i ¢
AH(m) EAST(UPPER FISSURE) 300 VENT
IOO]llllT~lllli[illflflllll
ESTIMATED Wo WEST (CENTRAL) VENT (m/s) 60 .t j t..%~oy ('"
FIG. 3 - Graph of A H (plume rise above the vent) and W~ (vertical exit velocity at the vent mouth) versus local time for the eruption of October 2.
come by wind and the cloud moved to- had spread to the north across the sea by ward the west. It approximately followed at least 7.5 kin, cutting off the view of the the valleys of Rivi~re Noire and Rivi~re sun from Marigot Village. As the cloud des P~res, moving as a tephra-laden den- was moving over the sea, it was still pre- sity current. As the cloud moved forward, ceeded by a thin vapor veil several hun- confined by ridges parallel to the valley, dred meters thick and by a low, thin, light there was some upward motion (Fig. 4). gray cloud moving as a density current im- Viewed from the northwest side, the lowest mediately above the water surface (Fig. km of the density current was curved up- 5). The full extent of the cloud to the west ward and outward with a striated appear- and south was not determined. The cloud ance. Above this level, for a height of 0.4 left thin (nun-scale) layers of fine-grained to 0.8 km, it flared out into a massive, tur- (clay and silt-size) tephra along the river bulent form (Fig. 4). As the cloud moved valleys down which it flowed. down the valley, a thin, veil-like wreath of vapor was pushed ahead of it. The cloud Oct. 4: moved along the valley to the sea with an average velocity of ~ 25 m/sec, consider- Phreatic explosions were observed from ably faster than the wind speed of about several locations throughout the day. The 10 m]sec. When the cloud reached the sea earliest activity was observed at close and was no longer confined as a density range, from the summit car park at Sa- current by valley walls, it spread laterally vane a Mulets (1050 to 1150 hrs). The ac- as well as forward. By 1740 hrs, about an tivity was solely at the summit vent. As hour after the eruption began, the cloud was observed on Oct. 2, there were two 388 PHREATIC ERUPTION CLOUDS: THE ACTMTY OF LA SOUFRI~RE DE GUADELOUPE, ETC.
Soufri~re
~" SHORELINE 0 2 4km I , , , I ot 7m 30s
rm) PHREATIC ERDFI~ON CLOUDS: THE ACTIVITY OF LA S0~ DE GUADELOUPE, ETC. 389 cloud types at the vent; a dark gray tur- OCT. 4 SUMMIT bulent cloud in the center, flanked on the ends of the elongate fissure-like vent by pure white steam clouds. Surrounding both clouds was a thin veil of vapor (Fig. ~ APOR COLLAR 6). The clouds rose only 60 to 70 m above the summit before being swept downwind to the west. All three clouds were mixed within 1 km into a single medium gray cloud that moved down-slope into the val- ley of the Rivi~re Noire. The top of this cloud dropped to a level 100 to 200 m be- low the summit 4 km downwind (to the WSW). At this distance it began to rise again, to an elevation of about 2.5 km .1058 (about 1 km above the summit), aS a thin, light gray, vapor cloud (Fig.7). This may have been related to the pattern or ~
lAIR FALL
f FIG. 6 - Sketch of the summit during the phreatic eruptions of October 4; based on photographs taken from Savanes ~ Mulet (summit parking I ..,..~_.~ .---. :-9 --~ ,,~ ,~ fi/g._~,.~ .-.-- lot). A turbulent dark gray, tephra-laden erup- tion cloud was flanked by a white to light gray steam cloud and a diffuse < MAIN STREET, AT 1750 tephra than the one described on Oct. 4. MARIGOT Activity similar to that described above continued through the day, with ash-fall visible below much of the length of each FIG. 5 - Sketch of the terminus of the steam and tephra cloud that flowed over the sea, as a cloud. Phreatic activity later in the day density current at 1730 hrs. The main body of (1600 to 1750 hrs) is summarized in Fig. the medium-gray cloud was preceded by a low, 8. The cloud heights (AH) varied from 20 more rapidly moving cloud (b) and a thin vapor to 250 m until the wind speed dropped veil (a). and the air was calm~ Mild activity produ- *-- FIG. 4 - (a) Schematic drawing of the density current resulting from the phreatic eruption of October 2: (top) section along the long axis, down stream valleys from the summit of La Souf- ri~re to the sea and, (bottom) section perpendicular to the first sketch, across the axis of the stream valley. These drawings are based upon series of Photographs taken from Vieux Habitants at 7 minutes, 30 seconds after the eruption began. (b) Map of the approximate area covered by the dense clouds of steam and tephra erupted on October 2. The area covered by the clouds (stippled) moved down the valley of the R. Noire and R. des P~res in 6 minutes (1638 h to 1644 h local time) and then began to spread laterally and for- ward across the sea. By 1740 (1 hr, 2 rain after the beginning of the eruption) the lateral growth of the cloud had blocked the view of the setting sun from Marigot. 390 HEIKF.~ - CROWE - MCGETCHIN - WEST - EICHELBERGER - BAETRAM . PETERSON - W0HLETZ FIG. 7 - Photograph of the eruption of October 4, f~om Basse-Terre (9 kln southwest of the dome). The tephra-laden cloud drops to a level 100 to 200 m below the summit, flowingdown the valley of the R. Noire. At a distance of 4 km from the summit, it rises again as a more diffuse cloud, to an altitude of about 2.5 km. TIME (OCT. 4, 1976) "~i' 1600 1630 1700 1750 500 I I I I I I I I I I I [ 1 I I I I I .,~ I • '°°.t,1 a,.,¢,~, o~ , ,T,T l ,T , , , T,T ,T T, ,TT, , T , T,T, , STEAMING I ERUPTION8EG,NS ~NEW.URSTS=t~ ~WIND SPEED DROPS "l I I I I I i I I I i i i i i I I I i ] I 6.2 Wo (m/sec)5. I 0 ' CROSS-WIND ~9M/s ~L= CALM FIG. 8 - Eruption sequence, evening of October 4, from 1600 to 1750. A H and Wo are plotted versus time. The horizontal wind velocity dropped at about 1720, allowing steam clouds to rise to 300 m above the summit. PHREATIC ERUPTION CLOUDS: THE ACTIVITY OF LA S0~ DE GUADELOUPE, ETC. 391 ced clouds rising to nearly 300 m during laden with tephra. The southeast vent windless conditions. Individual phreatic erupted steam only. explosions during this period were marked Based on maps provided by the obser- by a darkening of the cloud to a medium vatory and helicopter photographs of the or dark gray. Individual bursts were sepa- summit, the area of the main vent (G. Ta- rated by 2 to 5 minute periods, with some rissan) is about 600 m 2. Velocities based explosions (bursts) lasting up to 11 min- on the plume rise formula (Appendix) are utes. Near the end of this activity, the listed in Table 2. steam clouds turned white and contained very little tephra. TABLE 2 Velocity (m/sec) Period o[ Activity (rnin) Other Activity." 60 9 Other phreatic explosions are less well 40 4 documented. The explosion of 22 Sept. 20 6 may have reached an elevation of 3.5 km above the vent, but this was above the If we assume that steam and tephra field of view of the AutomR~ camera. How- were being released from a vent with an area of 600 m 2, at velocities listed in ever, according to the camera data obtain- ed, this eruption, which consisted of 2 Table 2, then a volume of 29.5 × 106 m a major explosions, 5 minutes apart, lasted was erupted from 1638 hrs to 1657 hrs (19 for about 27 minutes. minutes), the period of explosive activity. Although not observed by the authors, a The corridor along which the eruption group of observers were caught at the cloud flowed and along which tephra was summit during the phreatic explosions of deposited covers an area of 35 kin 2, exclu- August 30, 1040 AST. The tephra and sive of any tephra-faU over the sea. Based block-fall was heavy in the sumrrdt region. on only a few thickness measurements It was not possible, due to clouds, to ob- (4mm to lcm closest to the dome to serve the eruption cloud from Basse- 0.1 mm at a distance of about 8 Km) the Terre, The blast heard in Basse-Terre total volume of tephra deposits from this phreatic eruption is about 53 × I03 m 3. was similar in magnitude to a sonic boom, followed by a low roar for about 3 min- The eruption cloud must have entrained at least 2.6 kg]m 3 (assuming a bulk p of utes. Observers at the summit did not 1.5 gm]cm3 for the tephra) of erupted comment on any sound other than a low hissing noise. The tephra falling on them steam. This is clearly feasible for a fluidiz- was cool to touch, very sticky and fme- ed system of clay - silt - f'me sand size grained and several small lahars were ob- tephra. According to SPARKS (1976), en- served by those at the summit. traiument of material of that particle size in a fluidized system (steam passing through altered, fine-grained rock and clay) would require velocities as low as 10 cm]- Deposition of Fine-grained Tephra sec (0.1 m/sec). The observed velocities by Phreatic Explosions (20 to 60 m/sec.) are more than adequate to entrain large volumes of fme-grained Using data from the sequence camera tephra within the vent and carry it up in and individual observations of the erup- the eruption cloud. tion at 1638 hrs, Oct. 2, it is possible ap- As was described earlier, the tephra- proximately to define the mechanism of laden steam cloud flowed down valleys as the event. This event was selected a density current. The observed average because it seems to be fairly typical and velocity of 26 m]sec for the Oct. 2nd cloud ex- is the best-documented. Although both ceeds considerably the average wind velocity vents (summit and southeast fissure) of 10 m/sea It therefore must have flowed as were active, only the summit vent was a gravity-driven density current. 392 ~ - CROWE - MCGETCH1N - WEST - EICHELBERGER - BARTRAM - FETERSON . WOHLETZ TEPHRA FROM THE PHRF~TIC identical in composition and texture to the ERUPTIONS dome lavas. Twenty tephra samples, from 12 phrea- tic eruptions (12 August to 1 November, 1976) were studied. A manuscript is being DISCUSSION prepared on that subject by the Labora- toire de G6ochlmie et Cosmochimie, Ins- The absence of unaltered juvenile glass titute de Physique du Globe and Los Ala- in the tephra deposits from the 1976 ac- mos Scientific Laboratory; some of its tivity of La Soufri~re is highly significant. main conclusions are summarized here. This indicates that new magma was not Tephra from the phreatic eruptions is the direct driving mechanism for the erup- fine-grained with a mean grain size of tive activity. Rather the activity was most 4.7 ¢ (38 tim) sorting ~ 2.03 ¢, very probably the result of increased thermal poorly sorted)'!), forming very thin, gray pulses affecting groundwater circulation sticky deposits containing accretionary la- in a vapor-dominated systen~ The ther- pilli. mal pulses may have been initiated by a Most of the > 38 IJm fraction consists of major change or disruption of the existing hydrothermally altered andesite or dacite hydrothernml system by the introduction, fragments. The degree of alteration rang- at depth, of a magma body. Presence of a es from a trace of hematite on grain sur- magma body is suggested by the greatly faces to clay-pyrite mixtures, exhibiting increased shallow seismic activity (Bulle- only relict textures to identify them as tins of lTnstitut Physique du Globe). fragments of lava. Plagioclase grains, rang- Most vapor-domlnated geothermal sys- ing in composition from Ans0 to Anss, are tems are located in elevated areas of present in the tephra as broken subhedral groundwater recharge where groundwater crystals exhibiting progressive and oscilla- movement is downward and outward tory zoning. Pyroxenes are mostly pale (HEALY, 1976). The La Soufri~re dome is yellow to pale pink orthopyroxene (Wo3 the highest point on the island, and is the En63 Fss4 to W04 Ens2 Fs44). Smaller focus of the island's heaviest rainfall (LAS- amounts of clinopyroxene (Wo41 En41 Fsls SERRE, 1961). Within such systems, boil- to Wo4s En37 Fs20) are also presenk ing or warm springs may be located at the Amorphous masses of hematite and well- base of the elevated region and the steam developed pyrite grains are found in all of will pass upward where it may be explosively the samples. Glass particles (rhyolitic to released (HEALY, 1976). Again, this is andesitic) are present in small amounts. the case at La Soufri~re (summit at In all cases, the glass exhibits some de- 1466 m), where there are numerous hot gree of alteration and apparently was de- springs around the base, at elevations of rived from the glassy groundmass of the 950 to 1200 m (from 1 : 20,000 scale top- dome lavas. There was no evidence in ographic maps of Guadeloupe). the tephra of fresh (juvenile) magma. What~ then, controls the sporadic phrea- The La Soufri~re dome is composed of tic eruptions? The more explosive activity silicic andesite, with phenocrysts of pla- may be related to the rate of recharge of gioelase, augite and pigeonite making up meteoric water within the dome, the de- about 50% of the rocl~ Tephra from the crease in pore pressure during fortnightly phreatic eruptions were derived from tidal mlnlrrlum,q, or botl~ RINEHART (1976) hydrothermally-altered and acid-leached found that slight changes in state of strain rocks of the dome. The lithic, glass and modify pore pressure within permeable crystal components of this tephra are rock units and thus control the flow rates. Tidal changes do alter the flow characteris- (I) Sample from the phreatic eruption of tics of some geothermal systems, espe- October 2, collected at Matouba, 3 km SW of cially geysers. Fortnightly components of the summit. The statistical parameters of grain 11 to 16 days (RINEHART, 1976; MAUK size are those of FOLK (1968). and JOHNSON, 1973) have been observed PHREATIC ERUPTION CLOUDS: THE ACTIVITY OF LA SOUFRI~RE DE GUADELOUPE, ETC. 393 to affect geysers and erupting volcanoes, APPENDIX and may be a controlling factors for some of the explosive activity at La Soufri~re. Methods As was noted earlier, one of the periods of accelerated phreatic activity followed To asses the mechanics of the eruption extremely heavy rains by 2-112 hours. clouds, a 35-mm automatic camera (Auto- max) with a 50-mm lens was emplaced on Whatever the cause of the phreatic the Plateau du Palmiste. The camera was explosions, we suggest that vapor-dominat- located 4 km from the summit of La Souf- ed fluids eroded vent fillings and walls, ri~re and oriented normal to most of the and erupted fine-grained tephra and eruption clouds. One color photograph was steam. Observed velocities of erupted taken every minute. Although there were steam are more than adequate to entrain some problems with persistent cloud cov- fine-grained elastic material within the er and batteries, a number of small erup- vents. The simultaneous eruption of tions and several days of normal fumarolic tephra-laden and tephra-poor steam activity were photographed. Measure- clouds from the same vent has several ments of clouds are outlined in Fig. 9. implications. It may be that there were Other photographs were taken of the differences in the gas velocities along the activity from BassTerre, Matouba, St. length of the vent, and thus differences in Claude, the summit car park, Vieux Habi- the ability to erode vent walls, or there tants, and Marigot with hand-held 35-mm cameras having 50- to 300-mm lenses. may be lateral changes in the degree of Weather data necessary to analyze the vent-wall alteration and the ease with eruption cloud photographs were obtained which the alteration products were eroded from the National Climatological Cen- and carried out. ter, Ashville, N.C., U.S.A. The nearest sta- After rising buoyantly, the <~cool>~ tions where were data available for this tephra-laden clouds were pushed down- time period were Antigua, B.W.I. (Sonde wind and moved as density currents along data) and Basse-Terre, St. Kit~, B.W.I. stream valleys radiating from La Sou- (groundbased data). It was used, along fri~re. Movement of the clouds as density with subjective descriptions of the weather currents is supported by the fact that ob- by observers at the volcano. served clouds were moving faster than the To estimate the vertical exit velocities predominant winds and they were <~chan- of steam from vents on Soufri~re, we used neled>~ by topography. For example, the methods developed by BRIOGS (1969) and vapor-tephra cloud from the eruption of October 2 was confined to a valley until it reached the sea and began to spread lat- I__ ~ -1-[ erally. I: o, : Thin, corridor-like tephra deposits of fine-grained, hydrothermally altered dome rock, radiating from dome summits, and associated with laharic breccias may be used to identify past periods of intense hydrothermal activity at a volcano. Such deposits may be useful in studying past JH = HEIGHT of JET trends or eruptive cycles. Deposits similar .~ = CLOUD LENGTH to those deposited by the phreatic activ- of = LIMIT of VISIBLE TEPHA FALL ity at La Soufri~re have been described at Mt. Pel~e, Martinique, interbedded with FIc~. 9 - Schematic drawing, showing the mea- silicic pumice deposits (ROOBOL and surements made of each photograph from the SMITH, 1976). Automax (sequence) camera. 394 HEIXEN - CROWE - MCGETCHIN - WEST - EICHELBERGER - BARTRAM - PETERSON - WOHLETZ summarized by Settle (1976) for assessing oO = Potential temperature. the rise of industrial plumes. From the cam- oz ffi Elevation eras, the height of the cloud is determined, the average cross-wind speed is esti- The vertical exit velocities were approxi- mated from Sonde data, and the vent mated in the following way: width estlmsted from maps and photo- graphs. The formula from BRIGGS (1969), for F = us (2) stable conditions, windy, is: F AH ~ 4.0 (i) Wo = (3) where \ TsI AH = Plume rise height aove vent (m). D ~ Vent diameter [in this case, the vent width was 10 m; the vents ACKNOWI.EDGEMENTS were fissures, not circular cra- ters]. We wish to thank W. Lockyer, C. Lewis, W0 ~ Vertical exit velocity at the vent H. Fishbine, C. Lyon, C. Nelson, R. Riecker, mouth (m]s). for their help in this project. Barbara Hahn typed the manuscript. This work A T ~ Difference in absolute tempera- was funded by NASA Office of Planetary ture between ambient air and Geology, by the Energy Research and De- effluent gas (79°K). velopment Agency, Division of Physical T8 ~ Absolute temperature of efflu- Research and by the Office of Disaster ent stack gas (assume 373°K). Assistance, U.S. State Department. T, ~Absolute of ambient atmos- phere (292°K). REFERENCES F ~ Buoyancy flux ~ g (-) ~ Wo BARRABE, L. and JOLlVET, J. 1958, Les recentes (+)2, (m4/sec3) manifestations d'activitd de la Guadeloupe (Petites Antilles). Bull. Volcanol., 19, p. 143- 157. S =Atmospheric stability parame- BRIGGS, G.A, 1969, Plume Rise, U.S. Atomic g c~® Energy Comm. Critical Review Set. USAEC ter -- , T, oz Report TID-25075, 81 pp. BRUET, E. 1953, L'dge Absolu de la derni$re oO 9.8 °K grande ~ruption p$lSenns de la Soufri$re de --ffilr+ la Guadeloupe. Bull. Volcanol., 13, p. 105- oz 1000 m 107. FOLK, R. L. 1968, Petrology of Sedimentary Rocks, Hemphilrs, Austin, Texas, 170 pp. F = Environmental lapse rate HEALY, J., 1976, Geothermal Fields in Zones of Recent Volcanism. Proc. U.N. Second Conf. on Geothermal Energy, pp. 415-422. LASSElmE, G. 1961. La Guadeloupe. Un. Fr. d'Impression, Paris, 1135 pp. 1~ •Average cross-wind speed (9 MAUK, F.J. and JOHNSON, IVL 1973, On the m/s) Triggering of Volcanic Eruptions by Earth Tides. Jour. Geophys. Res., 78, p. 3356-3362. g = Gravitational acceleration. ~T, J.S. 1976, Alteration of Flow Char- PHRF,ATIC ERUPTION CLOUDS: THE ACTIVITY OF LA SOUFRIERE DE GUADELOUPE, ETC. 395 acteristics within Geothermal Areas by Clouds: relationship between Cloud Height Tidal Forces. Proc. U.N. Second Conf. on and Eruption Intensity. U.S. Air Force Geothermal Energy, pp. 549-551. Geophys. Lab. Environ. Res. Papers, no. 566 ROBSON, G.R. and TOMBLIN, J.F: 1966, Cata- (AFGL-TR-76-0127), 37 pp. logue of the Active Volcanoes and Solfatara SPARKS, R.S.J, 1976, Gram Size Variation in Fields of the West Indies. Cat. of the Active Ignimbrites and Implications for the Trans- Volcanoes of the World, part. XX. Intern. As- port of Pyroclastic Flows. Sedimentology, soc. Volc., Roma 56 pp. 23, p. 147-188, ROOBOL, M.J. and SMITH, A.L., 1976, Mount Pelde, Martinique: A pattern of Alternating Eruptive Styles, Geology, 4, p. 521-524. Ms. made available to the IA VCEI Publication SETTLE, M., 1976, Rise of Volcanic Eruption Office for the press July 1979.