Available online at www.sciencedirect.com

Journal of Ocean Engineering and Science 2 (2017) 83–89 www.elsevier.com/locate/joes

Review Article

A brief review on tsunami early warning detection using BPR approach

and post analysis by SAR satellite dataset

Sudhir Kumar Chaturvedi a , ∗, Pankaj Kumar Srivastava

b , Ugur Guven

a

a Department of Aerospace Engineering, University of Petroleum & Energy Studies, Dehradun 248007, India b Department of Petroleum Engineering & Earth Sciences, University of Petroleum & Energy Studies, Dehradun 248007, India

Received 20 July 2016; received in revised form 11 December 2016; accepted 21 December 2016 Available online 7 April 2017

Abstract

Tsunami early warning systems have provided to be the extreme importance after the tsunami that hit Japan in March 2011. This research article presents a case study based on the tsunami detection using Bottom Pressure Rate (BPR) measurement and the post the analysis using the SAR datasets. A final decision based system using BPR has been studied to carry out the measurements of tsunami wave parameters. SAR based study has also been carried out for the post tsunami studies. Wiener filters are utilized to remove the speckle noise presents in imagery. Future scope of this work has also been proposed. © 2017 Shanghai Jiaotong University. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license. ( http://creativecommons.org/licenses/by-nc-nd/4.0/ )

Keywords: Remote sensing; Tsunami damage detection; Epicenter; BPR.

1

p

t

t

s

s

1

C

o

s

t

a

s

u

w

s

c

n

[

b

d

g

i

k

w

t

N

W

t

a

t

l

a

h2(

. Introduction

The early detection and warning systems have shown androven an ultimate importance, especially after the destruc-ive tsunami that hit Japan in March 2011. The purpose ofhis research is to notify and enhance the existing tsunami re-ults for the detection and early warning prediction with theuitable accuracy [8] .

In the Pacific Ocean, a basin-wide tsunami occurred in960. The main cause of this generation was by the gainhilean earthquake, which was recorded as magnitude of 9.5ver the Richter scale. Tsunami caused more than 1500 ca-uals over the Chilean coast. Following this mega disaster, aeam has formed known as Tsunami commission by IUGGnd they develop tsunami warning system for the scientifictudies and investigation of the tsunamis by means of thenderwater explosions.

Furthermore, December 26, 2004 Indian Ocean tsunamisas again the worst disaster in the history and was

∗ Corresponding author. E-mail addresses: sudhir.avionics@gmail.com ,

udhir.chaturvedi@ddn.upes.ac.in (S.K. Chaturvedi).

c

m

t

e

ttp://dx.doi.org/10.1016/j.joes.2016.12.001 468-0133/© 2017 Shanghai Jiaotong University. Published by Elsevier B.V. This http://creativecommons.org/licenses/by-nc-nd/4.0/ )

aused by giant Sumatra-Andaman earthquake with mag-itude of 9.3 devastated the shores of Indian Ocean17,19,18] .

International coordination group was initiated and formedy Intergovernmental Oceanographic Commission (IOC) un-er United Nations Educational, Scientific and Cultural Or-anization (UNESCO) after the huge 1960 Chilean Tsunamin 1960. Again in 1965, another coordination group which isnown as International Tsunami Information Centre (ITIC)ith the support from United States of America, Na-

ional Oceanographic and Atmospheric Administration (USA-OAA). The International Coordination Group for Tsunamiarning Systems (ITSU) in Pacific was also formed and es-

ablished under IOC [14] . Once the tsunami waves generated from the deep water

nd runs up towards the coastal regions, it’s a difficult tasko prevent the life. In order to resolve these kinds of prob-ems, early warning systems are implemented in operation tonalyze and produce the results to disaster team. The basiconcepts of such systems is to detect any kind of unusual seis-ic activities under water and automatically judge whether

he seismic activity has been generated from the cause ofarthquake or some other reason. After this process, bulletins

is an open access article under the CC BY-NC-ND license.

84 S.K. Chaturvedi et al. / Journal of Ocean Engineering and Science 2 (2017) 83–89

c

c

c

p

t

d

b

d

p

c

t

t

t

c

p

o

r

t

w

g

w

c

b

e

t

a

a

i

d

2

g

C

t

e

l

v

w

o

s

n

w

m

s

b

i

c

w

c

c

c

t

m

must be issues to the disaster’s authority in order to take upthe appropriate action [3,5] .

Tsunamis, a critical natural menace, have the power tosource great destructions with damage of lives within mo-ments on shores. Any place having a huge water bodies, largelakes even, can cause a tsunami. From sources of history andscientific observations, the occurrence of Tsunamis can be inany large seas of the world, with almost 85% of tsunamishappening in the Pacific region, causing damage within hoursacross a complete ocean basin. Thereby, devastating tsunamishappen in geologically less lively oceans like the Atlantic, theIndian Ocean or the Mediterranean [10] .

Since the seismic activity of tectonic plates cause tsunami,they are often found in the Earth’s most restless fields aroundthe Pacific Rim along the "Pacific Ring of Fire", a regionof high tectonic activity. According to statistics, 17 tsunamisoccurred, in the Pacific, in between the year 1992 and 1996,taking nearly 1700 lives.

Numerous reports of newspaper along with various moviesabout undersea earthquakes and meteor inflicted tsunamishave bestowed to public awareness of the threat. The out-comes of tsunami can be catastrophic at times, for exam-ple, 2004 tsunami in Indian Ocean, the entire Indian Oceanwas propagated and caused immense destruction to at least12 countries. They have the strength to knock down infras-tructures, crush and flip vehicles, lift giant rocks, demolishhouses and cause failures of local/regional/international com-munication network and emergency response systems, causingdamage worth of millions or even billions of dollars. Some ofwhich could not even be addressed for months. They result indecline of any economy. In order for detecting the tsunamisthere is a need to determine the magnitude and epicenter ofthe earthquake.

One way to do so, is to verify the earliest arrivals, thetime of arrival and the wave amplitude of the tsunami, forwhich a former understanding of modeled propagation of thetsunami, is required by the system. The system is then ca-pable to respond to a warning. Such as, system subscriberswill reliably acquire a telephonic warning, as and when athreatening tsunami is unrolled [1] .

However, the large scale of tsunami destruction makes ithard to comprehend total tsunami impact in the whole Ocean.Latest developments of technologies of remote sensing crushmany problems and guides to detecting the elaborated charac-teristics of tsunami detriment. Another way to do so, as manyresearches propose, is an overview in developing a way tolook for and discover the impact of tsunami damage by inte-grating numerical modeling, technologies of GIS and remotesensing. Section of this method is carried out to few tsunamievent, such as, in the year 2007, on Solomon Island, tsunamicaused by earthquake, to find the affected regions and knowthe structural destruction, using the above method and theanalysis of satellite imagery with high-resolution optical.

Presently tsunami watch systems are built on computerprograms of modeling which notify against the likelihoodof the impacts of earthquake-originated tsunami, and try toforecast their strength along with their arrival times vs . lo-

ation based on the earthquake properties. These models ofomputer consist ocean-scale bathymetry and geometries ofoastline, with the input criterions modernized by ocean baseressure as the wave of tsunami moves over after a poten-ially threatening earthquake. A protocol lives for the quickissemination of seismic data and tsunami model notificationsetween alien governments but there is no system for localiscovery of a real incoming wave having a valuable alertingower.

Another process describes the sea level tide-gauges atoastal positions, nearer to the epicenter, able to convey vi-al quantitative data for locations further downstream, as inhe year 2011, in Japan, the tsunami signals were detectedhrough numerous HF radars all around the Pacific Rim withorrect outcomes from sites in Japan, US and Chile. An em-irical way for the automatic notification of a tsunami basedn pattern-recognition in time series of tsunami-generated cur-ent velocities, using information taken by fourteen radars onhe coasts of Japan and USA. Presently the HF radar systemsorks without break from many coastal locations around thelobe, observing the currents on the surface of the ocean andaves to distances up to 200 km from shore [7,13,12] . Basi-

ally the tsunami arrival and detection researches are limitedut the post processing has been carried in some of the lit-ratures using satellite image analysis for the affected area,sunami damage level classification using image processingnd manual techniques.

The main components for an end-to-end system of tsunamire to yield real-time surveying, seismic and tsunami activ-ties alert, punctual decision production and advisories, andissemination of warnings and information.

. Scientific advancements

Most of the current tsunami under water seismological al-orithms has been developed since the 1960s when the gianthilean earthquake generated in Pacific Ocean. Plate tectonic

heory was also introduced in the same year, numerous math-matical models of earthquake source were developed to re-ate the seismic moment and size of the fault. For an obser-ational sides, various tide gauge sensors, seismic networksas also deployed in 1960s. In 1970s, using the obtained the-ries and observed datasets, fault attributes of the larger ormaller earthquake studies have been carried out. The mag-itude moment scale which is well known as Richter scaleas also introduced. Theoretical and computational develop-ents made to estimate and compute the seafloor shift, fault

izes and tsunami wave propagations models on the actualathymetry. In 1980s, seismograms have been recorded dig-tally, which improved the data quality and reduces the pro-essing times. The large scale tsunami propagation modelsere also employed and studied in this decade and it be-

ame more popular. The developments and advancements ofomputer networking have made it possible and reliable toarry out the researches in seismic wave analysis and realime measurements of the tsunami wave parameters. Further-ore, the tsunami hazard development tools have also been

S.K. Chaturvedi et al. / Journal of Ocean Engineering and Science 2 (2017) 83–89 85

i

[

s

i

t

o

b

H

s

o

g

4

a

l

d

(

o

h

c

9

m

c

s

h

o

m

d

a

f

L

d

i

q

9

d

h

t

A

b

(

a

0

t

b

r

(

p

a

3

e

Table 1 Tsunami generation risk bulletin (NOAA Forecast manual, 2011) [11] .

Earth-quake depth Location Magnitude Tsunami potential

< 100 Km Very near sea > 7.9 Ocean wide destruction 7.6 to 7.8 Regional destruction 7.0 to 7.5 Local destruction 6.5 to 7.0 Very small scale

Inland > 6.5 No tsunami potential > 100 Km > 6.5 No tsunami potential

b

b

to

p

s

m

i

g

c

s

t

t

o

p

i

t

t

w

p

a

i

i

p

mplemented and discovered to analyze the seismic activities16,17] .

In the present era, the globally observed seismological andea-level datasets are available within the minutes through thenternet [21] . Using these datasets and information’s, moresunami studies could be carried out in a very short durationf time. Numerous scientific reports and research articles haveeen published regarding the 2004 tsunami earthquake [22] .ere, we only refer the special issues of some of the reputed

cientific journals in which most of the seismological aspectsf tsunami and earthquakes are explained [20] .

As under the discussion and part of the tsunami workingroup, the tide gauge data was collected and compiled about0–50 tide gauge records in to region of Indian Ocean tonalyze maximum amplitude and spectral components . Sea-evel monitoring of the oceanographic activities have beeneveloped known as Global Sea Level Observation SystemGLOSS) and is located in Australia, The pacific and Atlanticceans recorded the 2004 tsunami [16] .

Tsunami execution and propagation was also measured onydrophones or seismometers in which the analysis have beenarried out to provide a wide range of time of arrival about0 to 3000s. The aspect of Global Positioning Systems (GPS)easurements of tsunamis and its surveys have also been

arried out this research categories based on the societal re-ponses [9,15] .

The real-time observations and monitoring of a tsunamiave been limited to deep-water pressure-sensor observationsf variation in the sea level changes. The coastal based radaronitoring systems are implemented in various countries to

etect the tsunami wave’s arrival near to the coast and to an-lyze and present the report to the disaster management teamor the quick and sudden action to save various lives. Belindaipa et al. [30] have suggested an empirical model for theetection of the initial arrival of a tsunami, and demonstratets use with results from data measured by fourteen high fre-uency radar sites in Japan and USA following the magnitude.0 earthquakes off Sendai, Japan, on 11 March 2011. Theistance offshore at which the tsunami can be detected, andence the warning time provided, depends on the bathymetry:he wider the shallow continental shelf, the greater this time.rrival times measured by the radars preceded those at neigh-oring tide gauges by an average of 19 min (Japan) and 15 minUSA) The initial water-height increase due to the tsunamis measured by the tide gauges was moderate, ranging from.3 to 2 m. Thus it appears possible to detect even moderatesunamis using this method. Larger tsunamis could obviouslye detected further from the coast. We find that tsunami ar-ival within the radar coverage area can be announced 8 min i.e ., twice the radar spectral time resolution) after its first ap-earance. This can provide advance warning of the tsunamipproach to the coastline locations.

. Proposed detection and warning systems

In this article, we study and analyze the tsunami caused byarthquakes which can further be defined as a sudden distur-

ance of tectonic plates on the earth core. A system muste employed in universe which correlates the relation be-ween sudden tectonic plate’s displacements and propagation

f tsunami waves from deep to coastal regions of the waveropagation [2,4] .

Tsunami warning systems starts with the monitoring ofeismic events and corresponding wave patterns and deter-ining the earthquakes magnitude and epicenter, then further

t detects the tsunami waves. Such system detects the propa-ation of tsunami waves before it strikes on shoreline [6] .

A comprehensive tsunami detection and warning systemonsists of:

1 Seismic data, marine data collection using in-situ methodor satellite remote sensing imageries

2 A secured sea to ground surface and space based telecom-munication network.

3 An observing system which is effectively a virtual networkknown as Global Seismic Network (GSN), to measure andrecord all seismic vibrations

4 Regional satellite or other telecommunication based net-work to provide efficient budget link analysis

5 Additional notification system through mobile based dis-semination technique to support the government for thequick action.

Table 1 indicates the bulletin content based upon the deci-ion about tsunami generation risks, In general it depicts thathe local tsunamis effects are within 100 km of the epicen-er, regional tsunamis are limited to 100 km of epicenter andcean-wide tsunamis are across the entire ocean basin.

The case study of Japan tsunami is presented in thisaper, and the acquired datasets for the same is proposed

n Fig. 1 . The further step of the detection process is to obtain bot-

om pressure recorder (BPR) Measurement h d . This informa-ion is required to measure the expected run-up of the tsunamiaves towards the coast from deep water. Furthermore, de-ends on the flow chart mentioned in Fig. 2 , the bulletinsre generated and sent the warning to the concerned author-ties using e-mail for emergency purpose. The measurements used to estimate the expected run-up as per the equationrovided in Eq. (1) [23] .

h s

h d =

(H d

H s

)1 / 4

(1)

86 S.K. Chaturvedi et al. / Journal of Ocean Engineering and Science 2 (2017) 83–89

Fig. 1. Acquired datasets for the tsunami detection and arrival monitoring [23] .

Fig. 2. Decision for monitoring using BPR measurement [23] .

s

b

c

i

3

a

s

(

s

e

e

t

d

The bottom pressure recorders play the important role tofind and detect the actual cause of the tsunami generationpoint. It can be seen that if BPR is greater or equal to 0.03 m,the tsunami run-up starts. If the magnitude reaches more than2 m, the warning is to be issued to the concerned authority forthe further action. Hence, if BPR design should be so accuratein such a way that it should provide the accurate thresholdcommand values so that the quick action can be taken. Thelast and final step of the tsunami detection is to obtain thesea level fluctuation recorded by the tide gages located at theshore line and the actual run-up and the maximum wave run-up inundation can further be estimated using the mathematicalrelationship as mentioned in Eq. (2) .

H max = 2. 83 h s 1 . 25 cot B (2)

B is the slope of the seabed (in degrees)

Finally the bulletin is to be circulated depends on the deci-ion provided. The actual and acquired run-up values shoulde matched for the further validation. The updates about theonfirmation or cancellation are done based on the chart asllustrated in Fig. 3 .

. SAR image processing

The remote sensing tools are now a day on demand to an-lyze the tsunami affected regions. One of the important andophisticated well known sensor Synthetic Aperture RadarSAR) sensors can be used to observe the datasets. SARatellite sensor has been designed and, partially, put into op-ration, leading to an important breakthrough in Earth Sci-nce studies. The common characteristics of such new sys-ems are, indeed, a reduced revisit time (as short as a feways) and, in most cases, an improved spatial resolution (as

S.K. Chaturvedi et al. / Journal of Ocean Engineering and Science 2 (2017) 83–89 87

Fig. 3. Decision tree for confirmation, cancellation and update the database.

s

d

h

t

a

c

u

f

p

f

t

m

o

p

r

T

e

t

t

a

r

m

r

p

r

u

a

u

e

i

m

t

t

c

c

c

a

o

n

h

t

fi

m

v

F

b

P

S

l

q

P

t

l

t

a

b

t

b

p

l

i

d

s

l

n

dt

a

T

g

mall as a few meters), providing scientists with unprece-ented data for the mapping and monitoring of natural anduman-induced hazards [12-15] . At present due to lack ofhe datasets, we could not able to show how the SAR imagesppear for tsunami affected region. The simulation has beenarried out in MATLAB interface. A Wiener filters can besed to enhance SAR image by removing the speckles andurther post tsunami analysis can be carried out using imagerocessing technique. Wiener filter that is especially suitableor speckle and noise reduction in multilook synthetic aper-ure radar (SAR) imagery. The proposed filter is nonpara-etric, not being based on parametrized analytical models

f signal statistics. Instead, the Wiener-Hope equation is ex-ressed entirely in terms of observed signal statistics, with noeference to the possibly unobservable pure signal and noise.his Wiener filter is simple in concept and implementation,xactly minimum mean-square error, and directly applicableo signal-dependent and multiplicative noise. We demonstratehe filtering of a genuine two-look SAR image and show how no negatively constrained version of the filter substantiallyeduces ringing [28] . As illustrated in Fig. 4 , the images areapped into black and white binary, where white region rep-

esented the damaged fields. A pixel differential rule is ap-lied between the final obtained image and enhanced image,esultant of both provides the black region which indicatesnaffected fields. Once these images re resulted out, percent-ge damage can be determined.

A very few examples of detection, determination, and eval-ation of the damaged areas affected by the March 11, 2011arthquake and tsunami in east Japan. Due to very lim-ted time and resources, we used only conventional analysisethodologies; in spite of this limitation, we have been able

o show the very promising potential of PolSAR in disas-er observation, especially for scattering mechanism analysis,ompared to conventional single polarization SAR and opti-al remote sensing Detection of damaged areas after a disaster

an be done using only a single PolSAR observation. This is great advantage compared to the widely used techniquesf SAR interferometry. In particular, this advantage is sig-ificant for airborne PolSAR systems, which usually do notave archived data sets. In addition, though the resolution ofhe spaceborne ALOS/PALSAR full polarimetric mode is notne enough to identify each target, it could provide largescaleonitoring and understanding of the damaged area by re-

ealing the polarimetric scattering mechanisms on the whole.inally, observations of paddy fields demonstrated the possi-ility of quantitative evaluation of the flooding effect usingolSAR images [24] .

The use of full polarimetric synthetic aperture radar (Pol-AR) images for tsunami damage investigation from the po-

arimetric viewpoint. The great tsunami induced by the earth-uake of March 11th, 2011, which occurred beneath theacific off the northeastern coast of Japan, is adopted as

he study case using the Advanced Land Observing Satel-ite/Phased Array type L -band Synthetic Aperture Radar mul-itemporal PolSAR images. The polarimetric scattering mech-nism changes were quantitatively examined with model-ased decomposition. It is clear that the observed reduction inhe double-bounce scattering was due to a change into odd-ounce scattering, since a number of buildings were com-letely washed away, leaving relatively a rough surface. Po-arization orientation (PO) angles in built-up areas are alsonvestigated. After the tsunami, PO angle distributions fromamaged areas spread to a wider range and fluctuated moretrongly than those from the before-tsunami period. Two po-arimetric indicators are proposed for damage level discrimi-ation at the city block scale. One is the ratio of the dominantouble-bounce scattering mechanism observed after-tsunami o that observed before-tsunami, which can directly reflect themount of destroyed ground-wall structures in built-up areas.he second indicator is the standard deviation of the PO an-le differences, which is used to interpret the homogeneity

88 S.K. Chaturvedi et al. / Journal of Ocean Engineering and Science 2 (2017) 83–89

Fig. 4. Tsunami post damage extent analyses [28] .

o

i

d

S

o

T

w

t

t

m

t

u

t

fl

e

t

p

b

c

S

o

p

a

i

a

reduction of PO angles. Experimental results from after- andbefore-tsunami comparisons validate the efficiency of theseindexes, since the built-up areas with different damage levelscan be well discriminated. In addition, comparisons betweenbefore-tsunami pairs further confirm the stability of the twopolarimetric indexes over a long temporal duration. These in-teresting results also demonstrate the importance of full po-larimetric information for natural disaster assessment [25] .

A quick response to a large-scale natural disaster such asearthquake and tsunami is vital to mitigate further loss. Re-mote sensing, especially the spaceborne sensors, provides thepossibility to monitor a very large scale area in a short timeand with regular revisit circle. Damage ranges and damagelevels of the destructed urban areas are extremely importantinformation for rescue planning after an event. Rapid map-ping of the urban damage levels with synthetic aperture radar(SAR) is still challenging. Compared with single-polarizationSAR, fully polarimetric SAR (PolSAR) has a better poten-tial to understand the urban damage from the viewpoint ofscattering mechanism investigation [26] .

The great earthquake and tsunami occurred in March 2011at Japan, the various parameters have been analyzed and re-sulted out such as the marine debris in the region of japancoastal area which contains the post analysis activity usingSARdatasets over the various processing cycle of the satellitemovements [27] . Earthquake & Tsunami of 2011, the number

f confirmed deaths is 15,891 as of April 10, 2015, accord-ng to Japan’s National Police Agency. Most people died byrowning. More than 2500 people are still reported missing.

The major outcomes of the analysis of the COSMO-kyMed (CSK) synthetic aperture radar (SAR) observationsf the area hit by the 2011 Japan tsunami are presented.he height of the tsunami waves was such as to cause aidespread inundation of the coastal area. The SAR acquisi-

ions have been performed on March 12 (i.e., one day afterhe tsunami occurred) and March 13, 2011 in interferometricode, so that not only the information on the intensity of

he radar signals, but also the complex coherence has beensed. The interpretation of the available data has allowed uso detect the flooded areas, as well as the receding of theoodwater from March 12 to March 13, 2011 and the pres-nce of the debris floating above the water surface. Moreover,hanks to the high spatial resolution of the CSK images, theresence of floodwater in some urban areas in the Sendai har-or has been revealed by exploiting the information on theoherence [29] . Since tsunamis readily manifest themselves asea Level Anomalies (SLA) it’s necessary that one satelliteverflies the wave almost immediately after it originates. Theresence in orbit of several satellites (constellation), instead,llows to improve the frequency of observation and accord-ngly to have a better possibility in surveying the phenomenons soon as it occurs. It is worthwhile to remember the ac-

S.K. Chaturvedi et al. / Journal of Ocean Engineering and Science 2 (2017) 83–89 89

c

S

s

C

v

l

a

m

e

4

t

b

i

T

a

u

t

i

c

t

t

R

[

[

[

[[

[[[[

[[[[[

[

[

[

[[

[

[

omplishment of COSMO-SkyMed (Constellation of Smallatellites for Mediterranean basin Observation) satellite con-tellation, the first global constellation for Earth Observation.OSMO-SkyMed data can be used to exploit the most ad-anced remote sensing technology with the four SAR satel-ites. The satellite constellations would provide more timelynd comprehensive data and the ability to support disasteranagement and information on the evolution of disaster ar-

as.

. Conclusions

In this article, we carried out a case study of existingsunami system for the detection and monitoring. A decisionased matrix has been prepared to provide the early warningssues based upon the bottom pressure rate measurements.he model was then followed up by the enhancing SAR im-ge processing techniques with the removal of speckle noisesing wiener filters. The filters are more suitable for the postsunami analysis. For the future work, we could propose themproving communication network using the fiber optics mi-rowave cables in the ocean for the selected region. The real-ime monitoring can be observed based on the data acquitionechniques to provide BPR to the coastal based radars.

eferences

[1] W.M. Adams , J.M Jordaan , Hydraul. Struct. Equip. Water Data Acquis.Syst. 1 (2011) 238–261 .

[2] Australian Government Bureau of Meteorology. Tsunami Facts and In-formation. Sept 18. 2011

[3] G Bugharello , Syst. Chall. Global Scale 35 (2) (2005) 435–445 . [4] D. Chang-Seng (2006) “Seychelles progress report towards a Tsunami

warning system in the Indian ocean”. [5] DART. “Deep–Ocean Assessment and Reporting of Tsunamis Descrip-

tion”. National Data Buoy Center. 2011. [6] Envirtech. “Technical specifications of underwater module”. 2011. [7] K. Horsburgh , L. Bradley , M. Angus , D. Smith , F. Wijeratne , P. Wood-

worth , High Frequency Sea Level Recording for Tsunami Warning and

Enhanced Storm Surge Monitoring at UK Sites, National EnvironmentalResearch Council Open Research Archive., 2010 .

[8] W Morissay , Tsunamis Monitoring, Detection and Early Warning Sys-tems, The Library of Congress Congressional Research Service, 2005 .

[9] R. Stosius , G. Beyerle , M. Semmling , A. Helm , A. Hoechner , J. Wick-ert , J. Lauterjung , in: Proceedings of the Geoscience and Remote Sens-ing Symposium (IGARSS), 2010, pp. 3047–3050 .

10] Y. Teshirogi , J. Sawamoto , N. Segawa , E. Sugino , in: Proceedings of theIEEE International Advanced Information Networking and ApplicationsWorkshops, 2009, pp. 890–895 .

11] Forecast Inundation Models. NOAA Center foe Tsunami Research. nctr.pmel.noaa.gov/forecast _ inundation _ models.html. 2011.

12] J. Dubois , O. Abdul-Latif , Trans. Eng. Comput. Technol 12 (2005)139–143 .

13] J.S. Daba , M.R. Bell , Opt. Eng. 42 (1) (2003) 211–227 . 14] International Federation of Red Cross and Red Crescent Societies , World

Disaster Report, Kumarian Press, 2005, p. 246 . 15] Y. Fujii , K Satake , Bull. Seism. Soc. 97 (2007) 192–207 . 16] E.A Okal , Pure Appl. Geophys. 164 (2007) 355–378 . 17] K Abe , Pure Appl. Geophys 164 (2007) 454–463 . 18] A.B. Rabinovich , R.E Thomson , Pure Appl. Geophys 164 (2007)

261–308 . 19] Smith, W.H.F. and Sandwell, D.T (1997). Science. 277: 1956–1962. 20] S. Stien , E.A Okal , Nature 434 (2005) 581–582 . 21] R Bilham , Science 308 (2005) 1126–1127 . 22] W.D. Iwan , Earthq. Spectr. 22 (2006) 11–14 . 23] J.O. Dubois , J.S. Daba , H. Karam , J. Abdallah , Int. J. Comput. Inf.

Syst. Control Eng. 8 (7) (2014) . 24] Motoyuki Sato , Si-Wei Chen , Makoto Satake , Proc. IEEE 100 (10)

(2012) 2861–2875 . 25] Si-Wei Chen , Motoyuki Sato , IEEE Trans. Geosci. Remote Sens. 51 (4)

(2013) 1985–1997 . 26] Si-Wei Chen, Xue-Song Wang, Motoyuki Sato, Urban damage level

mapping based on scattering mechanism investigation using fully polari-metric SAR data for the 3.11 east Japan earthquake, IEEE Transactionson Geoscience and Remote Sensing.

27] M. Arii , M. Koiwa , Y. Aoki , IEEE JSTARS 7 (5) (2014) 1729–1744 . 28] RS Caprari, AS Goh, EK Moffatt, 2000. Noise and speckle reduction

in synthetic aperture radar imagery by nonparametric Wiener filteringhttps:// www.ncbi.nlm.nih.gov/ pubmed/ 18354677 .

29] C. Marco , P. Luca , P. Nazzareno , IEEE Geosci. Remote Sens. Lett. 3(9) (2012) 938–942 .

30] B. Lipa , I. James , N. Bruce , B. Donald , Remote Sensing 4 (2012)1448–1461 .