InSAR imaging of Aleutian Volcanoes - Earth Online - ESAearth.esa.int/fringe2005/proceedings/presentations/278_lu.pdf · volcanoes deform. For example, the deformation may be confined

InSAR imaging of Aleutian Volcanoes

Zhong LuSAIC, USGS/EROS

Sioux Falls, [email protected]

http://edc.usgs.gov/Geo_Apps

Acknowledgements:

• Contributions by many colleagues.

• NASA and USGS for funding and support.

• Alaska Satellite Facility (ASF), European Space Agency (ESA) and Japanese Space Agency (JAXA) for providing us SAR imagery.

Outline• InSAR Studies of Alaska Volcanoes

– Akutan Volcano– Peulik Volcano– Aniakchak Volcano– Okmok Volcano– Westdahl Volcano– Shishaldin, Pavlof, Cleveland and Korovin

volcanoes

Deformation Interpretation

N57

°50′

N57

°40′

N57

°45′

w156°30′ w156°20′ w156°10′

5 km

2.8 5.68.4

11.2

14.0

16.8

0.0 cm

0 2.83 cm

• Current and Past Sensors• European ERS-1, 1991-2000, C-band, 35-day repeat cycle • European ERS-2, 1995-now, C-band, 35-day repeat cycle

(experiencing malfunctions since early 2001)• Japanese JERS-1, 1992-1998, L-band, 44-day repeat cycle• Canadian Radarsat-1, 1995-now, C-band, 24-day repeat cycle• European Envisat, 2002-now, C-band, 35-day repeat cycle

• Future Sensors• Japanese ALOS, 2006, L-band, 46-day repeat cycle• Canadian Radarsat-2, 2006(?), C-band• German TerraSAR-X, 2006(?), X-band• U.S. DOD Space-based Radar Constellations• U.S. ECHO+, forever?

• …

Wavelength (λ)• X-band: λ = ~3 cm • C-band: λ = ~5.7 cm• L-band: λ = ~24 cm

Synthetic Aperture Radar Satellites

• U.S. SIR-C Mission, April (10 days) and Oct (10 days), 1994X/C/L-band, Fully Polarized

Motivations (1)• In the Aleutian volcanic arc, nearly persistent cloud cover precludes acquisition of

useful optical satellite images for a particular time period. In the past 30 years, only a few cloud-free images have been acquired for most of the Aleutian volcanoes.

• Although the rate of eruptive activities in the Aleutians is very high, these volcanoes remain relatively poorly studied due to the remote locations, difficult logistics and high cost of field measurement.

Okmok Volcano94 Landsat-7 images: July 1999 – Sept 2005

• With its global coverage and all-weather imaging capability, InSAR provides a means to map surface deformation at a high degree of spatial density. At volcanoes, such deformation is often an important indication of future eruptive potential and/or cause of unrest, and may be the first sign of increased volcanic activity, preceding swarms of earthquakes or other precursors that signal impending intrusions or eruptions.

• Measuring how a volcano’s surface deforms before, during, and after eruptions, provides the essential information about magma dynamics and a basis for mitigating volcanic hazards.

• The spatial distribution of surface deformation data, derived from InSAR images, enables the construction of detailed mechanical models to enhance the study of magmatic and tectonic processes.

Motivations (2)

Makushin, AK

Akutan, AK

Westdahl, AK

Augustine, AK

Kiska, AK

Seguam, AK

Peulik, AK

0 28.3 cm

Okmok, AK

Aniakchak, AK Shishaldin, AK

Aseismic Inflation and Magma Supply Dynamics at Westdahl Volcano, Alaska,

Revealed by InSAR

Lu, Z., et al., GRL, 2000 Lu, Z., et al., JGR, 2003Lu, Z., et al., RSE, 2004

5 km

Westdahl, AK

Aseismic inflation at Westdahl volcano, imaged by InSAR

0 2.83 cm

5 km

Observed interferogram for 1993-1998 (Lu et al., 2000).

InSAR is a useful tool to prospect for magma bodies by detecting deformation that is virtually aseismic and therefore difficult to detect by conventional means. It may extend the warning time before eruptions to years or evendecades, filling a crucial gap between traditional volcano monitoring of shallow seismic precursors (hours to months of warning) and the long-term volcano hazards assessments based on eruptive history.

June 1992 - Sept. 1993

Sept. 1992 - Sept. 1993

Sept. 1993 - Aug. 1995 Oct. 1997 - Aug. 1999

Oct. 1995 – Oct. 1998

Oct. 1993 - Aug. 1998 Aug. 1999 - Aug. 2000

Oct. 1992 - June 1997

Sept. 1993 - Oct. 1998

post-eruption InSAR images (several examples)

21 Nov. 1991 – 30 Nov. 1991

co-eruption InSAR image

InSAR images can characterize transient deformation of Westdahl volcano before, during and after the 1991 eruption

7 Sept. 1991 – 28 Oct. 1991

pre-eruption InSAR image

Lu et al., GRL, 2000Lu et al., JGR, 2003

1 color cycle = 2.8 cm deformation

Def

orm

atio

n hi

stor

y of

Wes

tdah

l vol

cano

Magma plumbing system for Westdahl volcano, inferred from InSAR and modeling

~7 km

Sea level

Shallow Reservoir

~100 - 200 kmLu et al., JGR, 2003

Volcanic deformation associated with severe seismic swarms

– Akutan Volcano

• The second most active in the Aleutian arc; 27 separate eruptiveepisodes since 1790

• Latest seismic crisis: March 1996

• Lu et al., Journal of Geophysical Research, 2000.• Lu et al., Canadian J. Remote Sensing, 2005

Courtesy of AVO

5 kmN

OpenBight

Lava Point

0 2.83 cm

W166°00’

N54

°10’

W165°50’

N54

°05’

City of Akutan

CascadeBight

Deformation mapped by ERS (C-band, λ = 5.66 cm) InSAR

5 kmN

0 11.76 cm

Akutan

1996 Cracks

Deformation mapped by JERS (L-band, λ = 23.53 cm) InSAR

5 kmN

0 11.76 cm

Akutan

1996 Cracks

Deformation mapped by JERS (L-band, λ = 23.53 cm) InSAR

A-A’B-B’

A-A’B-B’

Fault Scarps

Dormant volcanoes are working!!!– Mount Peulik Volcano

• Mount Peulik: Last eruption occurred ~150 years ago

• Lu et al., Journal of Geophysical Research, 2002.

Ugashik

Peulik

5 km

Inflation of 24 cm (1996-1998) due to magmatic intrusion

M5.1

M5.2

Littoral cone

Ukinrek Marrs

Mt Peulik

?

Becharof Lake

Ugashik

? M4.8

Becharof LakeMount Peulik volcano

Ugashik caldera

Magmatic Inflation at a Dormant Stratovolcano: 1996-98 Activity at Mount Peulik Volcano, Alaska, Revealed by InSAR

InSAR study of the 1996-98 inflation episode at Mount Peulik 1) confirms that InSAR can detect magma accumulation beneath dormant volcanoes at least several months before other signs of unrest are apparent, and perhaps years before an eventual eruption; 2) represents a first step toward understanding the eruption cycle at Peulik and other stratovolcanoes with characteristically long repose periods, and could lead to improved eruption forecasting and hazards mitigation.

Anchorage

00

100 miles200 km

PACIFICOCEAN

ALASKA

BERINGSEA

N

Vent Mountain

1931 Crater SurpriseLake

10 km0

Aniakchak

Kwoun et al., 2005

Aniakchak Volcano

• A stratovolcano with a collapse caldera formed 3500 years ago• Last two significant eruptions: 1600, 1931

0 14 mm

range change

10 km0

Examples of Deformation Interferograms

Aug. 4, 1992 ~ Jul. 13, 1995 Oct. 7, 1995 ~ Aug. 27, 2000

B

A

10 km0

0 14 mm/yr

range change rate

Averaged Deformation Interferogram (1992 ~ 2002)

-14

-12

-10

-8

-6

-4

-2

0

2

0 1 2 4 5 6 7 8 10 11

distance (km)

LOS

def

orm

atio

n ra

te(m

m/y

ear)

0

200

400

600

800

1000

1200

Elev

atio

n (m

)

A B

Deformation profile

DEM profile

D

Mean Deformation Rate

Least Squares Inversion of Deformation

1992 1994 1996 1998 2000 2002 2004-140

-120

-100

-80

-60

-40

-20

0Deformation at the center of the Aniakchak Caldera

years

LOS

defo

rmat

ion

(mm

)

-5 0 5 10Range Change Rate (mm/yr)

Depth: 4.2 km Volume change: -0.0012 km3/yr

0 5 10 15 20 25 30Local Easting (km)

Loca

l Nor

thin

g (k

m)

0

5

10

1

5

20

25

3

0

Data Synthetic Residual

Modeling of Average Deformation with a Mogi Source

Observed Modeled Residual

Based on water contents and Fe-Ti oxide data, Bacon [2000] estimated that the pre-eruption magma storage depths for the last two significant eruptions at Aniakchak were 5 km for the 1931 event and 3 km for the 400 BP eruptions.

~3 km

~4.2 km

~5 km

Mogi source

400 BP eruption

1931 eruption

Our result from Mogi modeling suggests that the surface subsidence is related to the cooling/contraction of magma body.

Source Location

• Eruptions in this century: 1931, 1936, 1943, 1945, 1958, 1960,1981, 1983, 1986, 1988

• Latest eruption: February - April 1997

Magma Supply Dynamics and Post-emplacement Lava Flow Deformation – Okmok Volcano

Lu et al., EOS, 1998 Lu et al., JGR, 2000. Lu et al., IEEE TGRS, 2003. Lu et al., JGR, 2005

Facts about Okmok

• Shield volcano• Caldera formed 2050 years ago• ~10 minor explosive eruptions (ash) in 20th century• 3 large effusive eruptions (basaltic flows ) in 1945, 1958 and 1997• All eruptions from Cone A

Transient deformation of Okmok volcano, Alaska

ERS-1: 1992-1996ERS-2: 1995-2003Radarsat-1: 2000-2003JERS-1: 1992-1998

1992-1993 1993-1995 1995-1996

1995-1997 1992-1998

5 km

(a) (b)

(d)

(c)

(e)

0 28.3 cm

0 2.83 cm 0 2.83 cm 0 2.83 cm

0 11.8 cm

Transient Deformation of Okmok Volcano, AlaskaP

re-e

rupt

ion

Co-

erup

tion

A

BBA

Def

latio

n (c

m)

Distance (km)

1997-1998 1998-1999 1999-2000

2000-2001 2001-2002 2002-2003

5 km

0 2.83 cm

(a) (b) (c)

(d) (d) (f)

Post-eruptive inflation of Okmok volcano, Alaska

Magma supply rate for the shallow reservoir

• A magma reservoir residing at 3.2 km beneath the center of the caldera, is responsible for the observed deformation before, during and after the 1997 eruption.

• By the summer of 2004, 45~75% of the magma volume from the 1997 eruption had been replenished.

Pre-eruption Post-eruption

1997 eruption

Deformation fields of Okmok volcano

Interferograms after the 1997 eruption suggest at least four distinct deformation processes:

1. volcano-wide inflation due to replenishment of the shallow magma reservoir,

2. subsidence of the 1997 lava flows due to thermal contraction, 3. deformation of the 1958 lava flows due to loading by the 1997

flows, and4. continuing thermal contraction of 1958 lava flows buried beneath

1997 flows.

Lack of expected deformation associated with recent eruptions at several stratovolcanoes: Shishaldin, Pavlof, Cleveland and Korovin volcanoes

Lu et al., EOM, 2003Moran et al., JVGR, 2005

Shishaldin

Central and eastern Aleutians

Western Aleutians

Korovin

ClevelandPavlof

100 km

Shishaldin, Cleveland, Pavlof, and Korovin Volcanoes

• All of the four volcanoes are stratovolcanoes, have symmetric cones, and are frequently active.

• Co-eruption interferograms with 1- or 2-year separation for Shishaldin, Pavlof, Cleveland, and Korovin volcanoes do not show significant InSAR measurements of deformation associated with the recent eruptions. Three possible scenarios: 1) no significant pre-eruptive and co-eruptive deformation was associated with these eruptions; 2) pre-eruptive inflation was balanced by co-eruptive deflation and no net displacement could be observed; 3) the magma source is very shallow and magma strength is small so that deformation could only occur over the region of lost coherence.

• Interferograms with shorter time separations (a few days) and a few continuous measurements provided by GPS may help characterize how these volcanoes deform. For example, the deformation may be confined to the areas near the summits of the volcanoes, for which InSAR images, having annual repeat passes, are persistently incoherent due to snow and ice. Short repeat-pass (a few days) InSAR images and/or GPS data have the potential to capture this localized deformation if it exists.

Lack of expected deformation

The deformation patterns and the associated magma supply mechanisms over Aleutian volcanoes: • are diverse and vary between volcanoes, • help us understand how these volcanoes work,• have significant implications for future volcano monitoring effort.

Conclusions

Thank you!

• Interferometric synthetic aperture radar (InSAR) combines phase information from two or more radar images of the same area acquired from similar vantage points at different times to produce an interferogram.

• The interferogram, depicting range changes between the radar and the ground, can be further processed with a digital elevation model (DEM) to image ground deformation at a horizontal resolution of tens of meters over areas of ~100 km x 100 km with centimeter to sub-centimeter precision under favorable conditions.

Observed and modeled deformation images

Observed Modeled

Deformation sources:

• b1: a shallow expanding source representing intrusion of magma.• b2, b3, & b4: contracting sources that together account for observed subsidence of the eastern part of the island.

InSAR images in the field!

Westdahl is a young glacier-clad shield volcano located on the central Aleutian arc. The edifice is composed of a thick sequence of pre-glacial basalt lava flows. The volcano was frequently active during the latter half of the 20th century, with documented eruptions in 1964, 1978-79, and 1991-92. The background level of seismic activity since the last eruption in 1991-92 has been low (about five M < 3 earthquakes per year) and no unusual activity has been detected.

A group of satellite radar interferograms that span the time period from 1991 to 2000 shows that Westdahl volcano, Alaska, deflated during its 1991-92 eruption and is re-inflating at a rate that could produce another eruption within the next several years. The rates of inflation and deflation are approximated by exponential decay functions having time constants of about 6 years and a few days, respectively. This behavior is consistent with a deep, constant-pressure magma source connected to a shallow reservoir by a magma-filled conduit. An elastic deformation model indicates that the reservoir is located about 6 km below sea level and beneath Westdahl Peak. We propose that the magma flow rate through the conduit is governed by the pressure gradient between the deep source and the reservoir. The pressure gradient, and hence the flow rate, are greatest immediately after eruptions. Pressurization of the reservoir decreases both the pressure gradient and the flow rate, but eventually the reservoir ruptures and an eruption or intrusion ensues. The eruption rate is controlled partly by the pressure gradient between the reservoir and surface, and therefore it, too, decreases with time. When the supply of eruptible magma is exhausted, the eruption stops, the reservoir begins to repressurize at a high rate, and the cycle repeats. This model might also be appropriate for other frequently active volcanoes with stable magma sources and relatively simple magma storage systems.

Summary: Paul - FYI

In March 1996, an intense earthquake swarm beneath Akutan Island, Alaska, was accompanied by extensive ground cracking but no eruption of Akutan volcano. Radar interferograms produced from L-band JERS-1 and C-band ERS-1/2 images show uplift by as much as 60 cm on the western part of the island associated with the swarm. The JERS interferograms have greater coherence, especially in areas with loose surface material or thick vegetation. They also show subsidence of similar magnitude on the eastern part of the island, plus displacements along faults reactivated during the swarm. The axis of uplift and subsidence strikes about N70°W, which is roughly parallel to a zone of fresh cracks on the volcano’s northwest flank, to normal faults that cut the island, and to the inferred maximum compressive stress direction. A common feature of models that fit the deformation is the emplacement of a shallow dike along this trend beneath the volcano’s northwest flank. Both before and after the swarm, the northwest flank was uplifted 5-20 mm/year relative to the southwest flank, probably by magma intrusion. The zone of fresh cracks subsided about 20 mm during 1996-1997 and at lesser rates thereafter, possibly because of cooling and degassing of the intrusion.

Summary: Paul - FYI

Summary: Paul - FYI

Peulik, a stratovolcano, is located on the Alaska Peninsula. The volcano has been active only twice during historical time, in 1814 and 1852. A series of ERS interferograms that collectively span the time interval from July 1992 to August 2000 reveal that a presumed magma body located 6.6 km beneath the Peulik volcano inflated 0.051 km3 between October 1996 and September 1998. The average inflation rate of the magma body was about 0.003 km3/month from October 1996 to September 1997, peaked at 0.005 km3/month during June 26-October 9, 1997, and dropped to 0.001 km3/month from October 1997 to September 1998. Deformation before October 1996 or after September 1998 is not significant. An intense earthquake swarm, began on May 8, 1998 near Becharof Lake, ~30 km northwest of Peulik. More than 400 earthquakes with a cumulative moment of 7.15×1017 N⋅m were recorded in the area through October 19, 1998. Although the inflation ended at about the same time when the earthquake swarm occurred, static stress changes we calculate in the epicentral area due to inflation beneath Peulik appear too small to provide a causal link. The 1996-1998 inflation episode at Peulik confirms that satellite radar interferometry can be used to detect magma accumulation beneath dormant volcanoes at least several months before other signs of unrest are apparent. This application represents a first step toward understanding the eruption cycle at Peulik and other stratovolcanoes with characteristically long repose periods.

Summary: Paul - FYI

Nineteen ERS-1/-2 interferometric synthetic aperture radar images from 1992 through 2002 of historically active Aniakchak Volcano, Alaska, reveal that its 10-km-wide caldera subsides about 13 mm/yr. The pattern of subsidence does not reflect the distribution of pyroclastic deposits from the last eruption in 1931 and therefore is not related to compaction of fragmental debris. Least-squares inversion of the deformation maps indicates a relatively constant subsidence rate. Modeling the deformation with a Mogi point source, locates the source of subsidence at about 4 km below the central caldera floor, which is consistent with the inferred depth of magma storage prior to the 1931 eruption. Magmatic CO2 and He have been measured at a warm soda spring within the caldera, and several sub-boiling fumaroles persist elsewhere in the caldera. These observations suggest that recent subsidence can be explained by cooling or degassing of a shallow magma body (about 4 km deep), and/or reduction of the pore fluid pressure of a cooling hydrothermal system. Ongoing deformation of the volcano detected by InSAR, in combination with magmatic gas output from at least one warm spring, and infrequent, low-level bursts of seismicity below the caldera, indicates the volcanic system is still active and, as such, requires close attention for the timely detection of any unforeseen hazard.

Summary: Paul - FYI

Okmok volcano, located in the central Aleutian arc, Alaska, is a dominantly basaltic complex topped with a 10-km-wide caldera that formed circa 2.05 ka. Okmok erupted several times during the 20th century, most recently in 1997; eruptions in 1945, 1958, and 1997 produced lava flows within the caldera. We used 80 interferometric synthetic aperture radar (InSAR) images (interferograms) to study transient deformation of the volcano before, during, and after the 1997 eruption. Point-source models suggest that a magma reservoir at a depth of 3.2 km below sea level, located beneath the center of the caldera and about 5 km northeast of the 1997 vent, is responsible for observed volcano-wide deformation. The pre-eruption uplift rate decreased from about 10 cm/year during 1992-1993 to 2~3 cm/year during 1993-1995 and then to about –1~–2 cm/year during 1995-1996. The post-eruption inflation rate generally decreased with time during 1997-2001, but increased significantly during 2001-2003. By the summer of 2003, 30~60% of the magma volume lost from the reservoir in the 1997 eruption had been replenished. Interferograms for periods before the 1997 eruption indicate consistent subsidence of the surface of the 1958 lava flows, most likely due to thermal contraction. Interferograms for periods after the eruption suggest at least four distinct deformation processes: (1) volcano-wide inflation due to replenishment of the shallow magma reservoir, (2) subsidence of the 1997 lava flows, most likely due to thermal contraction, (3) deformation of the 1958 lava flows due to loading by the 1997 flows, and (4) continuing compaction of 1958 lava flows buried beneath 1997 flows. Our results provide insights into the post-emplacement behavior of lava flows and have cautionary implications for the interpretation of inflation patterns at active volcanoes.

Deformation of lava flows erupted before 1997

Deformation of lava flows after 1997 eruption

C-b

and

Imag

es

Deformation of 1997 lava flows from JERS-1 Imagery

L-ba

nd Im

ages

Surface displacement due to lava contraction and consolidation can be 2 mm/day or more four months after the emplacement