Numerical Simulations of Volcanic Ash Plume Dispersal for Sakura-jima using a Real-time Emission Rate Estimation

[1 ] S.-I. Akasofu, and H. L. Tanaka, “Urgent issue of developing volcanic ash tracking

model,” Kagaku Asahi, No.5, 121-124 (in Japanese), 1993.

[2 ] C. Chatfield, “The analysis of time series: An introduction,” Chapman and Hall, 286

pp, 1984.

[3 ] K. G. Dean, S. I. Akasofu, and H. L. Tanaka, “Volcanic hazards and aviation safety:

Developing techniques in Alaska,” FAA Aviation Safety Journal, Vol. 3, No. 1,

11-15, 1993.

[4 ] M. Iguchi, “Prediction of volume of volcanic ash ejected from Showa crater of Sakurajima volcano, Japan,” Disaster Prevention Research Institute, Annual Report B,

University of Kyoto, 55, 169-175, 2012.

[5 ] M. Iguchi, “Method for real-time evaluation of discharge rate of volcanic ash: Case

study on intermittent eruptions at the Sakurajima volcano, Japan,” J. Disaster Res.,

11, 4-14, 2016.

[6 ] Japan Meteorological Agency (JMA), Home Page:

http://www.data.jma.go.jp/ svd/vois/data/kouhai/, 2018.

[7 ] R. Kelleher, “Atlantic conference on Eyjafjallajokull and aviation,” UK volcanic ash

safety regulation, Keflavic Airport, Iceland, 2010.

[8 ] J. Kienle, A. W. Woods, S. A. Estes, K. Ahlnaes, K. G. Dean, and H. L. Tanaka,

“Satellite and slow-scan television observations of the rise and dispersion of ash-rich

eruption clouds from Redoubt volcano, Alaska,” EOS, Vol. 72, 2, 748-750, 1991.

[9 ] S. Leadbetter, P. Agnew, L. Burgin, et al., “Overview of the NAME model and its

role as a VAAC atmospheric dispersion model during the Eyjafjallajokull eruption

18

April 2010,” Proc. EGU General Assembly, Vienna, Austria, p 15765, 2010.

[10 ] S. Onodera, “Volcanic activity and flight operations,” Aviation Meteorological

Notes, 45, 13-30, 1997.

[11 ] C. Searcy, K. Dean, and B. Stringer, “PUFF: A volcanic ash tracking and prediction

model,” J. Volc. and Geophys. Res., 80(1-2), 1-16, 1998.

[12 ] H. L. Tanaka, “Development of a prediction scheme for the volcanic ash fall from

Redoubt volcano,” First Int’l. Symp. on Volcanic Ash and Aviation Safety, Seattle,

Washington. U.S. Geological Survey, Circular 1065, 58, 1991.

[13 ] H. L. Tanaka, K. G. Dean, and S. I. Akasofu, “Predicting the movement of volcanic

ash clouds,” EOS, Vol. 74, No .20, 231-231, 1993.

[14 ] H. L. Tanaka, “Development of a prediction scheme for volcanic ash fall from Redoubt volcano, Alaska,” Proc. First International Symposium on Volcanic Ash and

Aviation Safety, U.S. Geological Survey, Bulletin 2047, 283-291, 1994.

[15 ] H.L. Tanaka, and K. Yamamoto, “Numerical simulations of volcanic plume dispersal

from Usu volcano in Japan on 31 March 2000,” Earth, Planets and Space, 54, 743752, 2002.

[16 ] H.L. Tanaka, M. Iguchi, and S. Nakata, “Numerical simulations of volcanic ash

plume dispersal from Kelud volcano in Indonesia on February 13, 2014,” J. Disaster

Res., 11, 31-42, 2016.

[17 ] H.L. Tanaka, and M. Iguchi, “Numerical simulations of volcanic ash plume dispersal

from Kuchinoerabu-jima volcano on 29 May 2015,” J. Natural Disaster Sci., 37, 7990, 2016.

19

[18 ] Tokyo Aviation Weather Service Center, “Volcanic ash advisory service, Japan

Meteorological Agency,” Geophys. Maga. Ser. 2, Vol. 4, Nos. 1-4, 2001.

[19 ] P. Wessel, W. H. F. Smith, R. Scharroo, J. F. Luis, and F. Wobbe, “Generic Mapping

Tools: Improved version released,” EOS Trans, AGU, 94, 409-410, 2013.

20

Figure legends

Fig. 1 Time series of the emission rate (ton/min) for local time on 16 June 2018 estimated by Eq. (3). There is an explosive eruption at 7:20 JST showing the emission

rate of 1000 ton/min.

Fig. 2 Time series of the emission rate converted to 5 min (ton/5min) and the corresponding plume height (m) estimated by Eq. (4) as the input for the PUFF model.

The simulation started at 7:10 JST (22:10 UTC) and ended at 13:10 JST (4:10

UTC). The simulation interval is noted from 0 to 6 hours in the abscissa.

Fig. 3 Distribution of wind vectors at 925 hPa and 500 hPa levels at 00:00 UTC on 16

June 2018 (9:00 JST). The real-time GPV data is provided by Japan Meteorological

Agency through the Center for Computational Science, University of Tsukuba.

Fig. 4 PUFF model simulation of the ash plume dispersal for (a) 7:30, (b) 8:00, (c) 8:30,

and (d) 9:00 JST, respectively. The particle colors indicate different plume heights.

Fig. 5 Ash plume dispersal in zonal-height (X-Z) and meridional-height (Y-Z) cross sections for every 5 min starting from 7:10 JST.

Fig. 6 Zonal-height (X-Z) and meridional-height (Y-Z) cross sections of ash plume dispersal for (a) 7:30, (b) 8:00, (c) 8:30, and (d) 9:00 JST, respectively. The colors of

particles indicate different plume heights.

Fig. 7 PUFF model simulation of the 3D perspective image of the volcanic ash plume

dispersal at 8:30 JST on 16 June 2018. The figure is for 70 min after the beginning

of the eruption. The colors of particles indicate a different plume height, and the

projection onto the ground is marked by black dots.

21

Fig. 8 Particle distribution of ash fallout over 6 hours from the onset of the eruption on

16 June 2018.

Fig. 9 The estimated concentration of ash fallout (g/m2 ) in common log-scale, i.e., 1.0

denotes 10 g/m2 . The contours are calculated by counting the number of fallout

particles in Fig. 8 using 1 km grid meshes.

Fig. 10 As in Fig. 9, but the contours are calculated using 100 m grid meshes. The

contour 2.0 denotes 100 g/m2 .

Fig. 11 Distribution of airborne ash density (mg/m3 ) for (a) 7:30, (b) 8:00, (c) 8:30, and

(d) 9:00 JST, respectively. The values are in common log-scale, i.e., 1.0 denotes 10

mg/m3 .

Fig. 12 Time series of the emission rate converted to 5 min (ton/5min) and the corresponding plume height (m) as in Fig. 2. The simulation started at 22:00 JST (13:00

UTC) and ended at 04:00 JST (19:00 UTC).

Fig. 13 PUFF model simulation of the ash plume dispersal on 13 November 2017 for

(a) 22:30, (b) 23:00, (c) 23:30, and (d) 24:00 JST, respectively, as in Fig. 4. The

particle colors indicate different plume heights.

Fig. 14 Ash plume dispersal in zonal-height (X-Z) and meridional-height (Y-Z) cross

sections for every 5 min as in Fig. 5, starting from 22:00 JST.

Fig. 15 PUFF model simulation of the 3D perspective image of the volcanic ash plume

dispersal at 23:30 JST on 13 November 2017, as in Fig. 7. The figure is for 70 min

after the beginning of the eruption.

Fig. 16 Particle distribution of ash fallout over 6 hours from the onset of the eruption

as in Fig. 8, but on 13 November 2017.

22

Fig. 17 The estimated concentration of ash fallout (g/m2 ) in common log-scale, as in

Fig. 9 but on 13 November 2017. The contours are calculated using 1 km grid

meshes.

Fig. 18 As in Fig.17, but the contours are calculated using 100 m grid meshes.

Fig. 19 Distribution of airborne ash density (mg/m3 ) for (a) 22:30, (b) 23:00, (c) 23:30,

and (d) 24:00 JST, respectively, as in Fig. 11 but on 13 November 2017.

23

Figure 1 Time series of the emission rate (ton/min) for local

time on 16 June 2018 estimated by Eq. (3). There is an

explosive eruption at 7:20 JST showing the emission rate of

1000 ton/min.

Figure 2 Time series of the emission rate converted to 5 min

(ton/5min) and the corresponding plume height (m) estimated

by Eq. (4) as the input for the PUFF model. The simulation

started at 7:10 JST (22:10 UTC) and ended at 13:10 JST (4:10

UTC). The simulation interval is noted from 0 to 6 hours in the

abscissa.

Figure 3 Distribution of wind vectors at 925 hPa and 500 hPa

levels at 00:00 UTC on 16 June 2018 (9:00 JST). The real-time

GPV data is provided by Japan Meteorological Agency.

Figure 4 PUFF model simulation of the ash plume dispersal on

16 June 2018 for (a) 7:30, (b) 8:00, (c) 8:30, and (d) 9:00 JST,

respectively. The particle colors indicate different plume heights.

Figure 5 Ash plume dispersal in zonal-height (X-Z) and

meridional-height (Y-Z) cross sections for for every 5 min

starting from 7:10 JST.

Figure 6 Zonal-height (X-Z) and meridional-height (Y-Z) cross

sections of ash plume dispersal for (a) 7:30, (b) 8:00, (c) 8:30,

and (d) 9:00 JST, respectively. The colors of particles indicate

different plume heights.

Figure 7 PUFF model simulation of the 3D perspective image

of the volcanic ash plume dispersal at 8:30 JST on 16 June

2018. The figure is for 70 min after the beginning of the

eruption. The colors of particles indicate a different plume

height, and the projection onto the ground is marked by black

dots.

Figure 8 Particle distribution of ash fallout over 6 hours from

the onset of the eruption on 16 June 2018.

Figure 9 The estimated concentration of ash fallout (g/m2) in

common log-scale, i.e., 1.0 denotes 10 g/m2. The contours are

calculated by counting the number of fallout particles in Fig. 8

using 1 km grid meshes.

Figure 10 As in Fig. 9, but the contours are calculated using

100 m grid meshes. The contour 2.0 denotes 100 g/m2.

Figure 11 Distribution of airborne ash density (mg/m3) for (a)

7:30, (b) 8:00, (c) 8:30, and (d) 9:00 JST, respectively. The

values are in common log-scale, i.e., 1.0 denotes 10 mg/m3.

Figure 12 Time series of the emission rate converted to 5 min

(ton/5min) and the corresponding plume height (m) as in Fig. 2.

The simulation started at 22:00 JST (13:00 UTC) and ended at

04:00 JST (19:00 UTC).

Figure 13 PUFF model simulation of the ash plume dispersal

on 13 November 2017 for (a) 22:30, (b) 23:00, (c) 23:30, and (d)

24:00 JST, respectively, as in Fig. 4.

indicate different plume heights.

The particle colors

Figure 14 Ash plume dispersal in zonal-height (X-Z) and

meridional-height (Y-Z) cross sections for every 5 min as in Fig.

5, starting from 22:00 JST.

Figure 15 PUFF model simulation of the 3D perspective

image of the volcanic ash plume dispersal at 23:30 JST on 13

November 2017, as in Fig. 7. The figure is for 70 min after the

beginning of the eruption.

Figure 16 Particle distribution of ash fallout over 6 hours

from the onset of the eruption as in Fig. 8, but on 13 November

2017.

Figure 17 The estimated concentration of ash fallout (g/m2) in

common log-scale, as in Fi.g 9 but on 13 November 2017. The

contours are calculated using 1 km grid meshes.

Figure 18 As in Fig.17, but the contours are calculated using

100 m grid meshes.

Figure 19 Distribution of airborne ash density (mg/m3) for (a)

7:30, (b) 8:00, (c) 8:30, and (d) 9:00 JST, respectively, as in Fig.

11 but on 13 November 2017.

...