IAGI Maluku 2021

The
Banda Sea:
Recent and rapid lithospheric
extension in eastern Indonesia

Jon Pownall
University of Helsinki

with thanks to… Jeff Malaihollo and the IAGI committee, Robert Hall and SEARG, Phil Cummins, Richard
Armstrong, Ian Watkinson, Gordon Lister, Marnie Forster, Yasinto Priastomo, Ramadhan Adhitama, Adianto
Trihatmojo, Ryan Pranantyo, Jono Griffin, and the ARC…
 Indonesian
tectonics are
complex!
SE Asia – Australia collision

SE ASIA PACIFIC
(reference) 110 mm a–1

AUSTRALIA
75 mm a–1
Nugroho et al. (2009)
How did things begin?
Craddock (1970)

Started to break
up in Early Jurassic
(c. 184 Ma)

India and Australia

headed north, to
later collide with
Eurasia

GONDWANA
Australia–SE Asia Ma
collision Hall (2012)

The Himalaya & Tibet ROLLBACK

Indonesia
What is “Slab Rollback”?

Sinn and Schlemmer (2014), after Ring et al. (2010)

Slab Rollback

Crameri et al. (2012)

How does slab rollback make
curved mountain belts?

Moresi et al. (2014)

Australia–SE Asia Ma
collision Hall (2012)

The Himalaya & Tibet ROLLBACK

Indonesia
Rollback  EXTENSION Pownall et al. (2014)
After Spakman & Hall (2010)

Hall (2012)
Extended oceanic crust + continental slivers

Seram a
“mirror image” of
Timor?

The Banda Arc a

“typical arc-continent
collision zone”?
A cross-section through the Banda Arc…

Pownall et al. (in prep.)

Outline of rest of talk
1. (Ultra-)high-T metamorphism and
melting in the northern Banda Arc
2. The Weber Deep and the Banda
Detachment (and historical tsunami)
3. When? New 40Ar/39Ar results for Buru
and the Banda Sea region
2.

3.
1.
What is UHT metamorphism?
T > 900˚C; P of 7–13 kbar (Harley, 2008; Kelsey, 2008)

Requires a geothermal gradient of > 750˚C GPa–1 (approximately equivalent to > 20˚C km –1 or the extent of
the sillimanite stability field above 900˚C).

Kelsey, 2008
How are UHT conditions reached?
HOT BACK ARCS?
Collins, 2002

RIDGE
SUBDUCTION?
How are UHT conditions reached? Sizova et al., 2014

Early Earth: mantle 250˚C hotter  Long-lived ultrahot orogens

e.g.,
Napier Complex
(Archean;
> 1050˚C; spr–qtz;
30–100 Myr duration)
2x present-day crustal
heat production
Harley, 2016

‘Modern’ granulites
(c. 16 Ma)
Seram, E Indo
UHT metamorphism through time Kelsey, 2008

‘Modern’
granulites
(c. 16 Ma)
Seram,
E Indo
Kobipoto Mountains, central Seram Pownall et al., 2018
The Kobipoto Complex Pownall et al., 2017

Granulite-
facies
migmatites

& Lherzolites

HEAT SOURCE
for UHT met.
UHT granulites: petrographic evidence Pownall, 2015

spinel +
corundum +
sapphirine

spinel + quartz

Melt inclusions in grt

Spl: 1.5 wt% ZnO

Phase equilibria modelling Pownall, 2015

THERMOCALC P–T
pseudosection for
melanosome

H2O and XFe3+

determined from
iterative T–MH2O
and T–MO
pseudosections
___________________

925˚C
9 kbar
Zircon U–Pb dating Pownall et al., 2014

TWO ages
• c. 200 Ma ‘inner’ rims
• 16 Ma ‘outer’ rims
Further evidence for the UHT
event occurring at 16 Ma:

• Zr-in-rutile T > 900oC

• Very similar HREE profiles

between 16 Ma zircon
and garnet
Zircon U–Pb age summary Pownall et al., 2017
Evidence for rapid exhumation
Duration of 12 Myr duration: long-lived history
UHT met.
Probably IDENTICAL-WITHIN-ERROR:
< 2–3 Myr. 16 Ma Rapid cooling histories?
Did not reset U–Pb zircon 5.8 Ma 3.6 Ma
the Lu–Hf U–Pb monazite U–Pb zircon U–Pb zircon
system in 40Ar/39Ar biotite 40Ar/39Ar biotite 40Ar/39Ar biotite

garnet
Evidence for rapid exhumation

‘Normal’“P&T–t”
cooling of metamorphic complex at depth
plot
Extremely rapid cooling and decompression of thin tectonic slices
Only individual exhumation histories are recorded by geochronometers.
6400˚C MYR–1 COOLING?
Mapping has shown high-T slices to be ~ 1km thick
Rapid exhumation, and juxtaposition against colder upper crustal rocks

How quickly does a 1 km-thick,

1200˚C dyke cool?

yrs Over 40,000 yrs:

Average cooling rate: 20,000 ˚C Myr–1

after Jaeger (1968)

Banda slab rollback  EXTENSION
Simply not enough time for a long-lived
(> 30 myr) Napier-style UHT event.
Outline of rest of talk
1. (Ultra-)high-T metamorphism and
melting in the northern Banda Arc
2. The Weber Deep and the Banda
Detachment (and historical tsunami)
3. When? New 40Ar/39Ar results for Buru
and the Banda Sea region
2.

3.
1.
EXTENSION  Exhumation of hot rocks
E Seram

7 km-deep basin 
N S
Fadol Island

W 7 km-deep basin  E
Weber Deep
• 7.2 km deep
• Earth’s deepest forearc
basin, and deepest
point of the oceans
NOT in a trench
• Parallel grooves and
ridges present on
landforms flooring
basin
• The lineations are
parallel to the Kawa
Shear Zone on Seram
and parallel to inferred
direction of rollback
Pownall et al., 2016
and extension.
Weber Deep

• Interpreted as
grooves on vast
low-angle fault
plane (or subsidiary
fault planes to larger
fault structure)…

• The Banda
Detachment
Pownall et al., 2016
Similar to grooved detachment faults exposing core complexes on land

Pompangeo
Metamorphic
Complex
(Spencer, 2010)

Pownall et al., 2016

 Kawa Shear Zone (KSZ) is the northern bounding
Banda Detachment fault of the Banda Detachment,
accommodating strike-slip motion

Pownall et al., 2016

Banda Detachment
• Here’s where the
cross-section is from…
The Banda Detachment Pownall et al., 2016, GEOLOGY

Rollback  Extension  Mantle exhumation  (U)HT metamorphism + melting

 Rapid exhumation by detachment faulting
Palu, Sulawesi
Sep 2018

Tsunami picture
The 1852 Banda Sea earthquake + tsunami
• Tsunami struck Banda Neira island 20 MINUTES after earthquake
• First tsunami wave: POSITIVE polarity / Maximum amplitude: 8 m
• Reported damage indicative of MMI ≥ 8 (Wichmann, 1918)

Banda Neira
island Banda Api
volcano

Tsunami wave
reportedly hit base
of hill below fort
Modelling the 1852 event Cummins et al., 2020
Nature Geoscience

Based on distribution of MMIs based on historical accounts (Wichmann, 1918)

Modelling the 1852 event Cummins et al., 2020
Nature Geoscience

Earthquake on ”megathrust" Earthquake on Banda Detachment

20 min
travel t

to Banda
Neira

JAGURS tsunami
modelling code
(Baba, 2019)
Modelling the 1852 event Cummins et al., 2020
Nature Geoscience

Earthquake on ”megathrust" Submarine landslide

(40 km long; 15 km wide; 50 m thick)

20 min
travel t

to Banda
Neira

JAGURS tsunami
modelling code
(Baba, 2019)
Modelling the 1852 event
Submarine landslide
(40 km long; 15 km wide; 50 m thick)

LOTS of vertical exaggeration

Modelling the 1852 event Cummins et al., 2020
Nature Geoscience

Weber Deep 20 min

Slump travel t

Banda
Detachment

Tanimbar
Trough
NEGATIVE 1st wave ”megathrust”
To summarise…
• Slab rollback drove
extension of the Weber
Deep along the Banda
Detachment

• Historical tsunamis in the

Banda Sea are explained 20 min
by a combination of… travel t
• Large (c. Mw 8) earthquakes
within (or related to) the
Banda Detachment shear
zone
• Submarine landslides down
the eastern ‘wall’ of the
Weber Deep
Outline of rest of talk
1. (Ultra-)high-T metamorphism and
melting in the northern Banda Arc
2. The Weber Deep and the Banda
Detachment (and historical tsunami)
3. When? New 40Ar/39Ar results for Buru
and the Banda Sea region
2.

3.
1.
Extended oceanic crust + continental slivers
North
Banda
Basin:
12–7 Ma
South
Banda
Basin:
6.5–3.5 Ma
Hinschberger et al., 2005
New 40Ar/39Ar dating • ANU Argon Lab
(Marnie Forster)
• Furnace step-heating
method
• Dual furnaces coupled
to Argus VI mass spec.
Somewhere in the
eastern Banda Sea

Thanks Ato Trihatmojo!

Buru
• Samples from
WAHLUA COMPLEX

• ? Equivalent to Tehoru
Formation on Seram

• Staurolite-garnet
micaschists to
greenschists

• Two phases of
intense, inclined to
recumbent folding

Pownall et al., IN PREP.

Buru
Pownall et al., IN PREP.

6–4 Ma
40Ar/39Ar

ages 4.58 Ma 4.23 Ma

6.07 Ma 5.92 Ma 5.13 Ma 4.11 Ma 4.71 Ma

Buru
Pownall et dating
Previous al., IN PREP.
of the Wahlua Complex
(Linthout et al., 1989, 1991)

• Whole-rock K–Ar ages:

6–4 Ma 4.5 ± 0.2 to 4.2 ± 0.8 Ma
40Ar/39Ar
• Biotite Rb–Sr ages between
ages 3.8 ± 0.2 and 3.0 ± 0.3 Ma 4.58 Ma 4.23 Ma

6.07 Ma 5.92 Ma 5.13 Ma 4.11 Ma 4.71 Ma

Pownall et al., IN PREP.

+ c. 90–60 Ma
40Ar/39Ar ages
69.9 Ma 89.7 Ma 59.6 Ma ?? Related to mid-
Cretaceous suturing in
central Sulawesi?

atmosphere

6.07 Ma 5.92 Ma 5.13 Ma

 • Quartz-mica mylonite lens, south of Wai Apo valley
Buru • Hosted by gneiss; limited partial melting? Pownall et al., IN PREP.

Range of Late
Jurassic to
Cretaceous ages
 Pownall et al., IN PREP.
Banda Ridges 12–11 Ma
40Ar/39Ar ages
On micaschists dredged by the
R/V Kana Keoki (Schwartz, 1985; Silver et al.,
1985; Cook and Charlton, 1986)

12.22 Ma 11.16 Ma
Seram (west / central)
Pownall et al., 2017

16–3.5 Ma 40Ar/39Ar ages

12.22 M 11.16 Ma
Seram (west / central)
Pownall et al., 2017 5.68 Ma 4.66 Ma
16–3.5 Ma 40Ar/39Ar ages

12.22 M 11.16 Ma
Seram (west / central) 16–3.5 Ma 40Ar/39Ar ages
Pownall et al., 2017

3.5–3.0 Ma:
Extension
on Ambon
4.5–4.4 Ma:
Kawa SZ
movement
c. 16 Ma:
UHT
5.8–5.6 Ma: metamorphism
Extension & melting
in W Seram
Seram (east)
Wai Leklekan Mountains
• Samples from KOBIPOTO COMPLEX
• Ultramafic rocks
• Crd diatexites (similar to W Seram)
• Upper-amphibolite facies pelites
• Mafic granulites (ES16-111, below)
Seram (east)
Wai Leklekan Mountains
16–14 Ma 40Ar/39Ar ages

14.8 Ma 16.0 Ma 14.0 Ma 14.4 Ma

Watubela Islands
Kasiui and Tioor
16–14 Ma 40Ar/39Ar ages

Tioor Kasiui

16.6 Ma 15.9 Ma 14.5 Ma

 Previous dating of bt gneiss on Kur
Watubela Islands (Honthaas et al., 1997)

• Biotite K–Ar age: 16.93 ± 0.39 Ma

Kur • Ksp K–Ar age: 17.64 ± 0.41 Ma
16 Ma
40Ar/39Ar ages

16.67 Ma 16.25 Ma
Watubela Islands
Fadol
16 Ma
40Ar/39Ar ages

16.45 Ma 16.17 Ma
Dai
Amphibolite

Sampled by
A. Richardson
in 1991
30.8 Ma
All Banda geochronology results Pownall et al., 2017
Pownall et al., IN PREP.

3.5–3.0 Ma:
Extension
on Ambon
4.5–4.4 Ma:
Kawa SZ
movement
c. 16 Ma:
UHT
5.8–5.6 Ma: metamorphism
Extension & melting
in W Seram
The c. 16 Ma UHT event…
E Seram
Exposed Kobipoto
Complex rocks

…but where
were theseFadol
rocksIsland
at 16 Ma?

W 7 km-deep basin  E
How would this look at 16 Ma? Pownall et al., 2017

Hall (2012) 16 Ma
How would this look through time? Pownall et al., 2017
Pownall et al., IN PREP.

16 Ma 7 Ma 5.7 Ma

4.5 Ma 3.4 Ma
2 to 0 Ma
Insights from E Indonesia

Rollback can drive

rapid extension causing:
• Exhumation of hot mantle to shallow
depths;
• Short-lived (ultra-)high temperature
metamorphism + melting;
• Rapid exhumation of the resultant
• high–T complex;
• Extension of deep oceanic basins along
major detachments.