IGARSS2011_final.pptx

Caldera Rims of Northeast Honshu Extracted from Gravity Anomalies and Aeromagnetic Data

Oky Dicky Ardiansyah, PRIMA (Prima, O.D.A.)Faculty of Software & Information Science, Iwate Prefectural University152-52 Sugo, Takizawa, Iwate 020-0193, JAPAN

Takeyoshi, YOSHIDA (Yoshida, T.)Institute of Min., Pet. & Eco. Geol., Grad. School of Sciences, Tohoku Univ.6-3 Aoba, Aramaki, Sendai 980-8578, JAPAN

Takeshi, KUDO (Kudo, T.)Department of Natural Science and Mathematics, Chubu University1200, Matsumoto-cho, Kasugai-shi, Aichi 487-8501, JAPAN

2011 IEEE International Geoscience and Remote Sensing Symposium

FR2.T09.5 - GIS Techniques III

Contents

• Background• Research aim• Data• Algorithm• Results• Conclusion

Background

Northeast Honshu, buried calderas, hypocenters

Hypocenters

Onikobe Caldera Akakura Caldera

Sendai

Japa

n Se

a

Background

Buried caldera causes landslide

Sumikawa landslide (1997/5/11) : pre-Yakeyama caldera (a buried caldera) covered by the younger Yakeyama volcano had played an essential role for the landslide (Oyagi, 2000).

★

800 m380 m

The first acknowledged landslide caused by buried calderas in Japan

Yakeyama Mt.

Background

Buried caldera causes landslide

Head cliff : 140m

Length : 1,400m

Width : 810m

The most recent landslide of buried caldera(Iwate-Miyagi Nairiku Earthquake, 2008/6/14)

Background

Hypocenters of Iwate-Miyagi Nairiku Earthquake, calderas, and seismic tomography

Okada et. al(2008)

Age of calderas (Ma)

Background

Caldera’s appearance on gravity anomalies and aeromagnetic data (convex and concave forms)

Gravity anomalies Aeromagnetic data

Concave forms Convex forms Convex forms

Research Aim

To automatically extract the topographic rim of buried caldera (with diameters ranging from 5 to 30 km) from gravity anomalies and aeromagnetic data using “watershed delineation” in GIS.

Ex: Watershed delineated from a DEM Gravity anomalies

Lowest point (sink)

Data

Gravity anomalies(published by the Geological Survey of Japan, 2000, 2004) (published by the Geological Survey of Japan, 2005)

Grid spacing 1 km　

Projection UTM

Grid Spacing 200m

Projection UTM

Aeromagnetic data

Algorithm

Gravitational acceleration

Calculating flow direction to the lowest points

Watershed delineation

Candidates for caldera rims

Aeromagneticdata

Calculating flow direction to the lowest points

Watershed delineation

Candidates for caldera rims

Pre-processing

Gridding using tension spline

2-D FFT

2-D iFFT

Data filtering

Gravity anomalies

Calculating Terrain Corrected Bouguer

Gravity

For calderas with concave forms in the data

Inverting the input data will allow the algorithm to delineate calderas with convex forms

Data filtering for gravity anomalies

km

Typical spectral distribution of gravity anomaly (Nozaki, 1997)

Eliminate noises using 2D FFT(Cut-off wavelength is empirically determined)

← long Wavelength short →

Four

ier

ampl

itud

e sp

ectr

um

(lo

g)

Trend

Signal

Noise

Watershed DelineationDelineating boundaries of concave (or convex) in gravity and aeromagnetic data is similar with that of watersheds from a digital elevation model (DEM) using hydrological modeling but without “depression-filling” process.

0

2

4

6

8

10

12

14

16

184 7 7 6 6 8 12

8 5 10 8 2 4 8

10 7 8 1 1 2 6

12 10 6 0 1 2 6

14 8 4 2 2 4 8

18 12 8 6 6 8 12

20 18 14 12 12 14 15

A concave form in the data

Rim extraction with ArcGIS model builder

Input data Reproject data Clip region of interest

Calculate flow directionCalculate watershed

Vectorize the boundary of watershed

Results

Concave and convex forms in gravity anomalies

Concave forms Convex forms

Vol

cani

c fr

ont

Vol

cani

c fr

ont

Results

Caldera rims improved by data filtering

No filteringCut off : 2 kmCut off : 3 kmCut off : 4 kmCut off : 5 kmCut off : 8 kmCut off : 10 km

Blue line: Caldera rims (Yoshida et al., 2005)Yellow line: Automatically delineated caldera rims

MukaimachiOnikobe

NarukoAkakura

Hanayama

Cut off : 4 kmSeamless Geological Map

kmConcave forms

Sendai

Results

Concave and convex forms in aeromagnetic data

Concave forms Convex forms

Vol

cani

c fr

ont

Vol

cani

c fr

ont

Results

Extracted rims from aeromagnetic data

Concave forms

Concave forms

Convex forms

Ushitaki

Tohgatta

Kawafune

Results Observable calderas by each geophysical data

Conc. : Concave Conv.: Convex

Conclusion

In this study, we have extracted caldera rims of Northeast Honshu, Japan depending on their forms: concave and convex, from gravity anomalies and aeromagnetic data.

Gravity anomalies seem to be superior on extracting caldera rims. However, there are some caldera rims that were extracted only from aeromagnetic data.

Thank you for your attention