European Journal of Remote Sensing

ISSN: (Print) 2279-7254 (Online) Journal homepage: https://www.tandfonline.com/loi/tejr20

Earthquake anomalies recognition through

satellite and in-situ monitoring data

Maria Zoran, Roxana Savastru & Dan Savastru

To cite this article: Maria Zoran, Roxana Savastru & Dan Savastru (2016) Earthquake anomalies
recognition through satellite and in-situ monitoring data, European Journal of Remote Sensing,
49:1, 1011-1032, DOI: 10.5721/EuJRS20164952

To link to this article: https://doi.org/10.5721/EuJRS20164952

© 2016 The Author(s). Published by Taylor &

Francis.

Published online: 17 Feb 2017.

Submit your article to this journal

Article views: 1275

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=tejr20
European Journal of Remote Sensing - 2016, 49: 1011-1032 doi: 10.5721/EuJRS20164952
Received 15/09/2016 accepted 08/11/2016

European Journal of Remote Sensing

An official journal of the Italian Society of Remote Sensing

www.aitjournal.com

Earthquake anomalies recognition through

satellite and in-situ monitoring data
Maria Zoran*, Roxana Savastru and Dan Savastru

Remote Sensing Department, National Institute of R&D for Optoelectronics,

MG5 Magurele-Bucharest, 077125 Romania,
*Corresponding author, e-mail address: [email protected]
Abstract
Cumulative stress energy in seismic active regions under operating tectonic force manifests
various earthquake precursors. For seismic hazard analysis in Vrancea area placed in the
Eastern Carpathians, Romania have been selected the earthquake presignals detectable from
space: land (LST) and air (AT) surface temperature anomalies, outgoing longwave radiation
(OLR) provided by time-series satellite NOAA AVHRR and MODIS (Terra/Aqua) data.
This study analyzed the presignal anomalies of geophysical parameters for some strong and
medium recorded earthquake events in Vrancea geotectonic active area: 1977, March 4th,
Mw = 7.4; May 30ty , 1990, Mw = 7; 1986 August , 30ty , Mw= 7.1; 2004, October 27th , Mw =5.9.
Keywords: Seismic precursors, land surface temperature, air surface temperature,
outgoing longwave radiation, AVHRR (NOAA) and MODIS, Vrancea.

Introduction
Earthquake prediction is one of the major imperative tasks of modern geophysics, with
two potentially compatible, but distinctly different objectives: (a) phenomena that provide
information about the future earthquake hazard and (b) phenomena causally related to the
physical processes governing failure on a fault that will improve our understanding of those
processes. According to the classical earthquake theory, small earthquakes should continue
to grow into large earthquakes until they spread all along the fault line. The mechanical
processes of earthquake preparation are always accompanied by deformations, afterwards
complex short- or long term precursory phenomena can appear. Cumulative stress energy
in seismic active regions under operating tectonic force manifests various earthquake
precursors (defined as changes in the Earth physical-chemical properties that take place
prior to an earthquake). The most important earthquake precursors are changes of electric
and magnetic fields, gas emissions (radon, methan, etc.), water level and quality (wells),
land and air surface temperature days or weeks prior to strong, even moderate earthquakes
[Freund, 2011; Varotsos et al., 2011; Erdal et al., 2016]. Also have been recorded ground
surface deformations (elevations) increase in earthquake-prone areas. It seems that Vrancea
region in Romania is fitting such a model [Zoran, 2012; Zoran et al., 2014].

1011
Zoran et al. Earthquake anomalies geospatial signatures

With the development of space-based technologies to measure surface geophysical

parameters and deformation at the boundaries of tectonic plates and large faults, earthquake
science has entered a new era. Different criteria can be used to select the remote sensed
earthquake precursors for which appear a clear evidence of the geophysical observable
anomalies. Observations from Earth orbiting satellites are complementary to local and
regional airborne observations, and to traditional in field measurements and ground-based
sensor networks. Remote sensing plays a critical role in operational seismotectonics
monitoring due to the remote locations of active fault systems and the large spatial extent
of potential seismic presignals. In order to select the main geophysical parameters to be
surveyed by geospatial and in-situ measurements in an integrated monitoring system, this
paper focused on the analysis of major earthquake activities of momemt magnitude Mw >5
recorded in the Vrancea tectonic active area in Romania for which specific precursors can
be reliably detected using satellite sensors, and establishes an important critical energy
threshold for lithosphere and atmosphere coupling.
One of the earthquake short-term presignals that have gained a lot of attention and support
from the scientific community across the world is thermal anomaly, defined by sudden rise
in land and air surface temperature a few days or weeks before the earthquake occurrence.
Due to possibly changes in physical properties that reflect the dilated or undilated state of
the region, the occurrence of a large or moderate earthquake may well change in response
to tectonic stress in a region. Thus, observed changes of some geophysical and geochemical
parameters in that area are attributed to changes in the crustal elasticity of the region [Dey,
et al., 2003]. The micro-crack formation strongly suggests unusual stress accumulation
in the fracture zone of an active fault. This energy transformation may result in enhanced
transient thermal infrared (TIR) emission, which can be detected through satellites
equipped with thermal sensors like MODIS Terra/Aqua (Moderate Resolution Imaging
Spectroradiometer), NOAA AVHRR (Advanced Very High Resolution Radiometer) TRMM
TMI (Tropical Rainfall Measuring Mission Microwave Imager, AMSR-E (Advanced
Microwave Scanning Radiometer). Regarding pre-earthquake ‘‘thermal infrared” (TIR)
anomalies, have been considered CO2 emanation leading to a local greenhouse effect or
latent heat due to condensation of water on air ions formed as the result of radon emanation
[Tronin, 1996; Tronin et al., 2004; Cervone et al., 2006; Pulinets et al., 2006; Freund, 2011].
Earthquake forecasting depends, in general, on the ability to detect precursory phenomena
and understand its relation to the nature of on-going physical processes in active tectonic
areas. As seismic events are some of the most catastrophic natural disasters to affect
mankind, the threat of earthquakes will probably increase because of global urbanization,
and thus millions of people are exposed to earthquakes. Remote sensing techniques, both
spaceborne and airborne, can bring an effective contribution for earthquake prediction.
In this paper, changes before and after the Vrancea earthquakes in the atmosphere-land
parameters have been investigated on the basis of time-series geospatial and field data
analysis. The detected changes show a complementary behavior in terms of the various
land, and atmospheric parameters, further showing strong evidence of coupling between
lithosphere-land surface- atmosphere associated with the Vrancea’s earthquakes. Continuous
monitoring and surveillance of these potential presignals helps us in differentiating
earthquake related variations from seasonal changes and atmospheric effects. Strong and
even moderate earthquakes are exceptional events reflecting very long-term deformation

1012
 European Journal of Remote Sensing - 2016, 49: 1011-1032

processes in a subduction zone [Takeshi et al., 2011; Chmyrev et al., 2013]. To develop a
comprehensive understanding of seismic events beside short-term precursors we must also
consider geodetic, geological and geomorphological information about crustal deformation.
A better understanding of the physics of earthquakes and an increase in the knowledge about
the space-time variation of the crustal process (i.e., seismicity and strain accumulation)
would allow geoscientists to make useful statements on long-term behavior of the crust , so
called “intermediate and long-term earthquake prognosis”, which are very important tools
for long-term seismic hazard reduction measures such as development of realistic building
codes, retrofitting existing structures, and land-use planning.

Vrancea study test site

Vrancea zone in Romania bounded by latitudes 45.6 °N and 46.0 °N and longitudes 26.5 °E
and 27.5 ° E (Fig.1) is one of the most active intracontinental tectonic areas in Europe , with
high potential of seismic hazard associated to a few strong intermediate depth earthquakes
(10 November 1940, Mw = 7.7, H = 150 km; 4 March 1977, Mw = 7.4, H = 94 km; 30 August
1986, Mw = 7.1, H = 131 km; 30 May 1990, Mw = 6.9, H = 91 km; 31 May 1990, Mw = 6.4,
H = 87 km; 27 October 2004, Mw =5.9, H =96 km) [Ardeleanu et al., 2005; Oth et al., 2007].

Figure 1 - Location of the Vrancea seismic zone in Romania.

The seismic activity of Vrancea is concentrated at the contact between the main tectonic
units being represented by the intermediate depth earthquakes concentrated in a narrow
area of some 3000 km2, near-vertical volume in the depth range of 85-220 km placed at the
Eastern Carpathian Arc Bend. Because earthquake hypocenters are concentrated within a
small volume, the Vrancea Zone is often described as unique. The existence of two kinds of
major tectonic units- the orogenic units (the Carpathian Orogen and the North Dobrudjan
Orogen) and platform units (the Moldavian Platform, as the South-Western margin of the
East-European Platform, and the Moesian Platform) leads to the hypothesis of a mobile and
seismic active contact between these units. Based on the hypothesis that the lithosphere
blocks are separated by comparatively thin, weak, less consolidated fault zones such as

1013
Zoran et al. Earthquake anomalies geospatial signatures

lineaments and tectonic faults, and that major deformation and most earthquakes occur in
such fault zones, Vrancea seismically active region is considered as a system of absolutely
rigid blocks divided by infinitely thin plane faults which interact between themselves and
with the underlying medium. The interaction of the blocks along the fault planes is elastic-
viscous (normal state) while the ratio of the stress to the pressure is below a certain strength
level. When the critical level is exceeded in some part of a fault plane, a stress-drop (a failure)
occurs (in accordance with the dry friction model), and it can cause a failure in other parts of
the fault planes. These failures produce earthquakes. Immediately after the earthquake and
for some time, the corresponding parts of the fault planes are in the creep state. This state
differs from the normal state because of the fast growth of the inelastic displacements and
lasts until the stress falls below some other threshold. According to such model, the strains
are accumulated in fault zones, which reflect strain accumulation due to deformations of
plate boundaries [Martin et al., 2005]. According to a 3D magneto-telluric tomographic
images model, the trigger action of the intermediate depth earthquakes in the Vrancea zone
may be interpreted as the rock response to active torsion processes induced by the complex
interplay among the structure of the litho-sphere and the surrounding asthenosphere, in
this sector of the Eastern Carpathians [Stanica et al., 2010]. This may explain the increase
shear stress and drive faulting system and re-shearing within the rigid slab in the Carpathian
Arc Bend and Vrancea zone. Located at Eastern Carpathians bend, intermediate-depth
seismogenic Vrancea zone is a complex tectonic environment, the heat flux budget in the area
playing a decisive role in defining the lithospheric rheological properties. Sedimentation,
palaeoclimate, under compaction and heat refraction effects, on one hand, and the heat
generated in the upper crust, on the other, combine to explain the observed subsurface
temperature field, and in particular the pronounced curvature of the vertical temperature
profiles. Due to the interplay of thermal effects of high sedimentation rate episodes (13-
12.5 Ma) and subsequent thermal relaxation and new sedimentation episodes, the overall
effect of the Neogene-Quaternary sedimentation is rather uniform along the study zone. In
the investigated area, both the thermal effects of rapid sedimentation and palaeoclimatic
effects contribute to the temperature field in the upper part of the sedimentary section.
The sedimentation process induces a significant time-dependency of the temperature field
of the underlying lithosphere with repercussions on its thermal thickness, metamorphic
state and rheological behavior. Temperature variations as large as 70-100◦C occurred in the
crystalline crust immediately under the sedimentary pile. Progressively smaller variations
were present up to 40-50 km depth. The temperature increases in the upper crust caused
by sedimentation. Thermally defined lithosphere thickness is considered to be of about 160
km. The lithosphere has not reached yet the thermal steady - state, the present temperatures
being about 80-90% of the equilibrium temperatures, depending on depth and on location
along geologic profiles [Demetrescu et al., 2007].
This study investigated the selected intermediate-depth seismic events presented in Table
1. The strength of an earthquake is usually measured on different magnitude scales, but the
moment magnitude Mw is regarded as the most representative value of the seismic source.
Earthquake data for four earthquakes with moment magnitude Mw between 5.9 and 7.4
and focal depth between 91km and 131 km are provided by the National Institute of Earth
Physics Catalog and USGS Catalog.

1014
 European Journal of Remote Sensing - 2016, 49: 1011-1032

Table 1 - Selected intermediate-depth Vrancea earthquakes.

Origin Location Focal Earthquake
Earthquake
number Date Time Latitude Longitude Depth Moment
yyyy/mm/dd (UTC) (oN) (oE) (km) Magnitude (Mw)
1. 1977/03/04 19:21:54.1 45.77 26.76 94.0 7.4
2. 1986/08/30 21:28:35.7 45.52 26.49 131.0 7.1
3. 1990/05/30 10:40:06.4 45.83 26.89 91.0 6.9
4. 2004/10/24 20:34:36 45.79 26.63 96.0 5.9

Methodology and geospatial data used

Pre-earthquake physical and chemical interactions in the earth’s ground may cause
anomalies of land surface temperature, outgoing longwave radiation, air temperature, etc..
Earthquake preparing is a transient dynamic process. Any tectonic activity is accompanied
by energy transfer and material movements, which can change the state of ground thermal
radiation. Based on this assumption is possible to derive present-day tectonic activities
from thermal infrared satellite data regarding thermal radiation state on the ground. As
received satellite infrared information is influenced by many types of factors, the main
problem to be solved is to extract useful information associated with tectonic activities and
to eliminate the effects of non-tectonic factors.
Our methods to find the anomalies of geophysical parameters identified as seismic
precursors have been used in the following steps: firstly, we analyzed thermal infrared
information provided by satellite data, removing data which were not related to tectonic
activity and selecting the data which were correlated with tectonic processes in the Vrancea
zone; secondly, we studied the relationship between tectonic strain and thermal exchange
through theoretical analysis; thirdly, we analyzed in-situ available data provided by the
seismic network in Vrancea area as well as meteorological stations. Finally, we compared
the ground surface and satellite data, and established the possible relationship between
these two kinds of observations. Before a potential anomaly can be nominated as a reliable
precursor, it should pass or be proved by the following tests or analyses: whether or not it
is an artificial anomaly, whether or not it correlates with an investigated event, whether or
not it is a random anomaly.
Prior to strong and moderate seismic events the earthquake precursors appear at different
distances and heights over the active seismogenic areas. It seems that the earthquake
preparation area on the ground can be estimated according to the relation:

R = 10M w0.43M [1]

where R is the radius of the preparation zone and Mw is the earthquake moment magnitude
[Dobrovolsky et al., 1979; Dobrovolsky et al., 1989]. Most of the precursory phenomena
may exhibit different patterns in which systematic growth takes place during the pre-seismic
period, corresponding to the local stress build-up. After that is followed by a relatively
rapid return to near equilibrium conditions following the seismic event. As examples, for 4
March 1977 Vrancea strong earthquake with Mw =7.4, the corresponding calculated radius

1015
Zoran et al. Earthquake anomalies geospatial signatures

of preparation zone would be R ≈ 1520.5 km and ≈ ± 13.7 degrees in the latitudinal and
longitudinal direction from the epicenter, while for moderate earthquake from 27 October
2004 with Mw =5.9, the corresponding calculated radius would be R ≈ 344 km and ≈ ±3.1
degrees in the latitudinal and longitudinal direction from the epicenter.
Function of their occurrence elevation the seismic presignals can be grouped into three
classes: on the surface (land surface temperature, radon gas, etc.); in the lower atmosphere
(outgoing long wave radiation, surface latent heat flux, and air temperature; and in the
higher atmosphere -ionosphere (total electron content -TEC, etc), [Dey et al., 2003; Tronin
et al., 2004; Ozunov et al., 2004; Tramutoli et al., 2005]. Some of the seismic presignals
have been observed by satellite remote sensing, while others by in-situ monitoring
techniques. Different criteria can be used to select remotely sensed earthquake presignals
for which there is evidence of anomalies in geophysical observables. Observations from
Earth-orbiting satellites are complementary to local and regional airborne observations, and
to traditional in-field measurements and ground-based sensor networks [Saraf et al., 2005;
Zoran et al., 2012a; Zoran et al., 2012b].
During last decades the remote sensing data have been widely used for the assessment
of prior and after strong earthquakes changes as well as for the detecting and mapping
of post-earthquake damages based on the earth observation data like (ALOS, PALSAR,
SPOT, Landsat TM/ETM, TerraSAR, RADARSAT-1/2, as well as some other remote
sensing imagery) [Singh et al., 2001; Yusuf et al., 2001]. With the advent and continuous
enlargement of space-based observations, researchers extended that use of geospatial
data in different geosciences applications which enables new solutions to determining
interrelations among various geophysical-geochemical phenomena like are earthquake
precursory phenomena [Zoran et al., 2007]. The relevant parameters (Outgoing Longwave
Radiation, Land Surface Temperature and the mean daily Air Temperature) related to
investigated earthquakes in Table 1 have been provided by the NOAA Climate Prediction
Center, National Environmental Satellite Data and Information Service (NESDIS), National
Centers for Environmental Prediction (NCEP) and the NOAA/ESRL Physical Sciences
Division, Boulder Colorado, USA from their web site at (http://www.ncep.noaa.gov/),
(www.nesdis.noaa.gov), (http://www.esrl.noaa.gov/psd/).
For Vrancea earthquake recorded on 27 October 2004 have been used also the available
time series of MODIS Terra/Aqua Land Surface Temperature/Emissivity ( LST ) 8-Day
L3 Global 1km SIN Grid MOD11A2 LST_Day_1km and LST_Night_1km data and
MODIS/Aqua Land Surface Temperature/Emissivity ( LST ) 8-Day L3 Global 1km SIN
Grid MYD11A2 LST _Day_1km and LST_Night_1km data, over different periods of time
provided by Oak Ridge National Laboratory Distributed Active Archive Center (ORNL
DAAC) (http://daac.ornl.gov/MODIS/modis.html). The available MODIS Terra/Aqua
satellite data starting from 2000/2002 years make it possible to resolve land surface thermal
anomalies globally with better than 1oK resolution. Meteorological data around the Vrancea
region in Romania: air temperature and anomaly on the base period of normals 1975-1995
and 1981-2010, have been provided by the Romanian Meteorological Administration
(www.meteoromania.ro) and (www.wunderground.com). Also, in-situ meteorological data
have been compared with satellite data.
Surrounding Vrancea, the several seismic stations belonging to the Romanian Seismic
Network are recording seismic and other geophysical, geoelectromagnetic, geodynamic

1016
 European Journal of Remote Sensing - 2016, 49: 1011-1032

and meteorological parameters. ENVI 5.0 and IDL 6.3 software for satellite data processing
have been used.

Results and discussion

In order to set accurate, precise, and stand-alone criteria for the earthquake prediction,
simultaneous integration of several precursor parameters which reduce the parameter
uncertainty must be considered. Using time-series remote sensing observations, this
paper examines spatio-temporal variations of land surface temperature, outgoing long-
wave radiation, and mean air temperature several weeks to several days before the
studied earthquakes. Earthquake is a dynamic phenomenon, usually accompanied by crust
displacements attributed to an energy transfer due to a breakdown between source and
environment. The changes prior to the earthquake or along with it may have different physical
and chemical effects on the lithosphere, atmosphere and ionosphere, therefore making it
detectable. These variations of lithosphere, atmosphere and ionosphere parameters before
the main earthquakes are considered as earthquake precursors which are used as hints for
impending earthquakes. As earthquakes are large-scale fracture phenomena, associated
fracture-induced physical fields allow a real-time monitoring of electromagnetic (EM)
emissions in a wide frequency spectrum ranging from kHz to MHz which are produced by
opening cracks. Those emissions can be considered as the so-called precursors of general
fracture.
These presignals suggest the existence of a strong coupling between the dynamics of
lithospheric processes land- atmospheric anomalies associated with seismic processes
[Pulinets et al., 2006]. Prior to a strong or moderate earthquake event in the areas close
to the earthquake epicenter, due to high tectonic stresses can be released radon, and
greenhouse gases (CO2, CH4, N2, etc.) that escape to the lower atmosphere and create a
localized greenhouse effect and thus augment the LST of the area. There is a great deal of
interest in the emissions of various gases from the earth prior to earthquakes.
Another possible mechanism responsible of the temperature rise preceding an earthquake
is the presence of positive hole-type charge carriers in rocks which become activated when
rocks are subject to high levels of stress prior to an earthquake, the appearance of these
electronic charge carriers within a finite rock volume leads to enhanced infrared emission
from the rock surface in the 8-12 mm region and hence an increase in LST [Dobrovolsky
et al., 1989; Ouzounov et al., 2004]. Pre-earthquake spatio-temporal thermal anomalies
are controlled by various factors like as earthquake moment magnitude and its focal depth,
geological setting, topography and land covers. Have been reported several observations of
such thermal anomalies preceding several major earthquakes in different parts of the world
[Tronin, 1996; Tronin et al., 2004; Saraf and Choudhury, 2004; Saraf and Choudhury, 2005;
Saraf et al., 2009].

Outgoing Longwave Radiation

Outgoing Longwave Radiation which is the emission to space of terrestrial radiation from
the top of the Earth’s atmosphere is controlled by the temperature of the earth and the
atmosphere above it, the water vapor content in the atmosphere, and the clouds. Long wave
radiation of the Earth is a major driver of the Earth climate system. OLR is the thermal
radiation flux emergent from the top of the atmosphere and connected with the earth-

1017
Zoran et al. Earthquake anomalies geospatial signatures

atmosphere system in general, being dependent on cloud and land surface temperature.
This energy has been measured at the top of the atmosphere by the National Oceanic and
Atmospheric Administration satellites and includes all of the emission from the ground,
atmosphere and clouds formation. The possible link of transient thermal fields on the ground
with pre-earthquake processes establishes the rationale to explore the radiation budget prior
to four major earthquakes in Vrancea area. OLR is a NOAA polar-orbiting satellite derived
measurement of the radiative character of energy radiated from the warmer earth’s surface
to a cooler space in the 10-12 μm infrared windows.
The interpolated OLR data are continuous spatially as well as temporally. The estimates
of interpolated OLR values (W/m2) originally observed by polar orbiting NOAA are based
on dedicated developed algorithms. Maximum and minimum OLR value ranges including
other parameters were defined in the analyzed earthquake cases. The high stress field prior
to earthquakes is responsible for surface and subsurface deformation in the epicentral
region, which, in turn, is responsible for changes in the thermal, electrical and magnetic
fields [Zoran et al., 2012]. Daily OLR anomalies have been investigated in detail for
selected Vrancea earthquakes because these can reflect the energy variation in the earth-
atmosphere system and record the earthquake thermal anomalies in different time scales
over the seismic monitoring field. Based on NOAA 1981-2010 climate data, for strong
Vrancea earthquake 4 March 1977 Mw =7.4, starting with the 16 February 1977 till 20
February 1977 has been developed a high positive daily OLR anomaly in the range of
10- 35 Watt/m2, with a maximum positive value on 18 February 1977 (Fig. 2). After that,
between 21 -24 February 1977 was recorded a strong negative OLR anomaly between -10
Watt/m2 to -40 Watt/m2. Other weak positive daily OLR anomalies have been registered on
27 -28 February and 1-3 March 1977, followed by another negative daily OLR anomaly on
28 February-1 March 1977 and 4 March 1977.
In case of 30th August 1986 earthquake with Mw =7.1 in the Vrancea tectonic active zone,
from NOAA derived data and based on 1981- 2010 climate have been recorded some
positive daily OLR anomalies between 17-19 August and 22-24 August 1986 in the range
of 10- 30 Watt/m2, but a very high positive anomaly between 26- 30 August 1986 in the
range of 15- 35 Watt/m2 with an intense maximum one day before the main shock in value
of 35 Watt/m2 (Fig. 3). The same the temporal variation of daily mean OLR anomaly,
NOAA derived data over the Vrancea area (Fig. 4) have been observed for 30th May 1990
Mw = 6.9 earthquake based on 1981- 2010 climate data. Prior almost 3- 4 weeks (1-10
May, 15-18 May, 21-24 May 1990) before the main shock appeared significant daily OLR
positive anomalies of 20-30 Watt/m2. Three days before earthquake (27-30 May 1990) a
strong negative daily OLR anomaly (-20 to -50 Watt/m2) was evidenced over Vrancea area,
which can be interpreted due to contraction the rocks. Similarly, for moderate 27th October
2004 earthquake (Mw = 5.9) recorded in Vrancea region, 3-4 days prior to the main event
have been recorded daily OLR positive anomalies of 20-50 W/m2 higher than the normal,
which evidenced gradually increase of the outgoing long-wave radiation OLR emitted by
land surface in the epicentral area before earthquake registered on 27 October 2004.

1018
 European Journal of Remote Sensing - 2016, 49: 1011-1032

Figure 2 - Daily OLR anomaly variation over Vrancea area before

and after 4 March 1977 earthquake.

Figure 3 - Daily OLR anomaly variation over Vrancea area before

and after 30 August 1986 earthquake.

1019
Zoran et al. Earthquake anomalies geospatial signatures

Figure 4 - Daily OLR anomaly variation over Vrancea area for 30

May 1990 earthquake.

Figure 5 - Daily OLR anomaly variation over Vrancea area for 27

October 2004 earthquake.

Figure 5 presents the temporal variation of daily mean OLR anomaly, NOAA derived
data over Vrancea epicentral area during 20 October 2004- 3 November 2004 expressed

1020
 European Journal of Remote Sensing - 2016, 49: 1011-1032

in W/m2 with an accuracy of 3 W/m2. The mean daily OLR increased starting from 24
October 2004, with a maximum value between 25-26 October and then decreased gradually
till 27 October in the morning, after that increased again between 27 - 30 October. OLR
anomalies covered an extended area described by latitudes 45N - 47N and longitudes 25E
- 27E and were distributed along the fault zone system in the Vrancea region. OLR is
dependent on local meteorological parameters, temperature and humidity and changes in
these variables may be responsible also for anomalous OLR values associated with seismic
events.

Land Surface Temperature

Land surface temperature, a measure that is widely used for thermal anomaly studies in
the epicentral and surrounding areas of strong earthquakes, can be retrieved by a variety
of satellite sensors depending on the spatial and temporal resolution needed. Data from
satellite sensors such as NOAA AVHRR series satellites, MODIS Terra and Aqua Earth
Observation System satellites, AATSR (Advanced Along-Track Scanning Radiometer) on
board the ENVISAT satellite, provide LST products retrieved for a specific area with a
spatial resolution of about 1 km, while SEVIRI (Spinning Enhanced Visible and Infrared
Imager) sensor on board the METEOSAT Second Generation (MSG) series geostationary
satellites, retrieves LST at a lower spatial resolution (about 3 km). Land surface temperature
anomalies (ALST) have been obtained by subtracting the multi-year mean LST from the area
averaged values and divided by multi-year mean values, which can be expressed as:

LST − LST
ALST = [ 2]
LST

From NOAA AVHRR data, generation of all the land surface temperature LST maps
was based on the split window algorithm, which uses the differential absorption effect
in NOAA AVHRR channels 4 and 5 for correcting the atmospheric attenuation mainly
caused by water vapor absorption [NOAA KLM, 2006]. Data calibration and temperature
calculation were based on the method provided in NOAA user’s guide [Gruber et al.,
1984]. Selected digital data sets (scenes) were kept consistent in terms of the time of
acquisition. This is followed by the time-series layouts of these LSTs that are finally
studied to analyze the temperature variation of the ground surface. Cloud-covers were
delineated and avoided from any type of calculation. As TIR radiation cannot penetrate
clouds, the temperature of cloudy areas will be only the temperature of cloud tops and
not the actual LST of the area.
A succession of temporal LST images show the gradual development of TIR anomaly
in epicentral area Vrancea and surroundings. As in case of daily OLR anomalies, mean
daily LST anomalies investigated for selected Vrancea earthquakes recorded the thermal
anomalies in different time scales over the seismic monitoring field. Based on NOAA 1981-
2010 climate data, for strong Vrancea earthquake 4th March 1977 Mw =7.4, starting with
16 February 1977 till 27 February 1977 has been developed a high positive mean daily
land surface temperature in the range of 5-11oC and maximum values of almost 11-13oC
between 20-26 February 1977. In case of daily OLR, the positive anomaly recorded almost

1021
Zoran et al. Earthquake anomalies geospatial signatures

4 days before 4th March 1977 earthquake was followed by a negative mean daily LST
anomaly with values of -1oC to -5oC for 28 February-3 March 1977 period (Fig. 6).

Figure 6 - Mean Daily LST anomaly over Vrancea area for 4

March 1977 earthquake.

Figure 7 - Mean Daily LST anomaly over Vrancea area for 30

August 1986 earthquake.

1022
 European Journal of Remote Sensing - 2016, 49: 1011-1032

Clear mean daily LST positive anomalies were developed for 30th August 1986 earthquake
with Mw =7.1 in Vrancea zone, between 16-20 August in the range of 1.6- 3.2oC with
a maximum value on 19th August 1986 of 4.8 and 26-30 August 1986, in the range of
1-3.2oC (Fig. 7) from NOAA derived data and based on 1981- 2010 climate data. The same
behavior of temporal variation of mean daily LST anomaly has been observed for 30th
May 1990 Mw = 6.9 earthquake based on 1981- 2010 climate and NOAA derived data over
Vrancea area. The pregnant appearance of mean daily LST positive anomaly was recorded
during 23-27 May 1990 period in the range of 2-8oC with 1oC resolution (Fig. 8). Three
days before the main shock was recorded a negative mean daily LST anomaly up to - 4oC.

Figure 8 - Mean Daily LST anomaly over Vrancea area for 30 May
1990 earthquake.

Based on NOAA derived data over Vrancea and 1981- 2010 climate data, for 27th October
2004 Mw = 5.9 earthquake appeared a high mean daily LST positive anomaly almost one
week prior the main event with 5- 10oC with 0.8oC resolution (Fig. 9) and a maximum value
of 12.8oC almost one week after the earthquake. As from 2002 are available time-series
products MODIS/Terra Land Surface Temperature/Emissivity ( LST ) 8-Day L3 Global 1km
SIN Grid, OD11A2/LST_Day_1km data, was possible to represent land surface temperature
variation during 2004 year over the Vrancea region (Fig.10) centered on earthquake of 27
October 2004 epicenter (45.787 N, 26.622 E), 101 km x 101 km surface area . For Land
Surface Temperature (LST) anomaly analysis over the Vrancea area during October 2004
month have been considered LST mean values of MODIS/Terra time-series data during
2000-2012 period and MODIS/Aqua time-series and different aria size polygons centered
on Vrancioaia (VRI) seismic station (Latitude 45.8657 and Longitude 26.7276).

1023
Zoran et al. Earthquake anomalies geospatial signatures

Figure 9 - Mean Daily LST anomaly over Vrancea area for 27

October 2004 earthquake.

Figure 10 - LST variation over Vrancea for 27 October 2004

earthquake from MODIS Terra data.

Time series satellite data analysis revealed a clear increase of land surface temperatures LST
around the epicentral area ranging 5-10oC in good accordance with NOAA AVHRR satellite
data. MODIS classification considered Pixel Aggregation Method (PAM) and found that
3559 of 10201 pixels [34.89%] were belonging to the same class as the center pixel “(5)
Mixed Forests”. In all test cases considered earthquakes in this paper, the increase in land
surface temperature LST and OLR near Vrancea epicentral area can be attributed to enhance

1024
 European Journal of Remote Sensing - 2016, 49: 1011-1032

greenhouse gas emission from the squeezed rock pore spaces and/or to the activation of
p-holes in stressed rock volume and their further recombination at rock-air interface. Due to
high tectonic stresses, prior to a strong earthquake event in the areas close to the earthquake
epicenter from the earth’s surface can be released radon gas, and greenhouse gases (CO2,
CH4, N2, etc.) that escape into the lower atmosphere and create a localized greenhouse
effect and thus augment the LST anomaly of the area [Zoran et al., 2012b].

Air Temperature
In order to study the relationship between the air temperature and Vrancea earthquakes
have been analyzed time-series mean daily air temperature anomalies, on the base period of
normals 1981-2010 around the Vrancea zone in Romania provided by NOAA satellites. For
4th March 1977 strong earthquake was analyzed 10 February -15 March 1977 period. The
positive air temperature anomaly started developing to North West and South East of the
epicentral area, air temperature showing a rise of around 6oC- 10oC during 11- 28 February
1977 (Fig. 11), in good correlation with in-situ measurements.

Figure 11 - Air Temperature anomaly over Vrancea 4 March 1977 earthquake.

1025
Zoran et al. Earthquake anomalies geospatial signatures

Figure 12 - Air Temperature anomaly over Europe before 4 March 1977

Vrancea earthquake.

NOAA/ESRL/Physical Sciences Division results for 10 February-28 February 1977 over

Europe (Fig. 12) revealed a pronounced increase of mean air temperature anomaly over
Romania and neighbor countries of more than 3.5oC, which can be explained through
long distance calculated radius of preparation zone of R ≈ 1520.5 km from the epicenter
according Equation [1] for earthquake 4 March 1977 with Mw =7.4. During 28 February- 4
March 1977 period, was recorded a clear evidence of negative air temperature anomaly
in range of -2oC to -6oC, also in accordance to in-situ measurements and other findings
for strong earthquakes in the world [Zoran, 2012]. The recorded negative air temperature
anomaly over the Vrancea zone and Romania five-six days before strong earthquake 4
March 1977 can be explained by an increased field compression in the epicentral and the
surrounding area. For next ten days period during 5-15 March 1977, again was recorded
a small positive air temperature anomaly in alternating with small negative temperature
anomaly over the Vrancea area and Romania. This observation fits well with the concept
that the thermal anomaly can appear some days to weeks before strong earthquakes and
dissipated after the quake. Time-series analysis of mean daily air temperature anomalies for
test case 30 August 1986 Vrancea earthquake revealed two strong positive air temperature
anomalies recorded about ten days (17- 20 August 1986) and another with three days (27-
30 August 1986) before the main shock in the range of 1.8- 3.6oC with 0.6oC resolution
(Fig. 13).

1026
 European Journal of Remote Sensing - 2016, 49: 1011-1032

Figure 13 - Air Temperature anomaly over Vrancea region for 30 August

1986 earthquake.

Similarly, mean daily air temperature anomalies have been registered before 30th May 1990
earthquake , when air temperature higher than the normal were observed one week prior to
the main event between 23- 26 May 1990 in the range of 1.6-6.4oC with 0.8oC resolution.
Figure 14 presents a succession of positive air temperature anomalies and negative anomaly
which appeared between 27- 31 May 1990 over Vrancea area. It seems that the negative
air temperature anomaly developed 3 days before the main shock of Mw= 6.9 on 30th May,
followed by a second seismic event of Mw= 6.4 on 31st May 1990 can be attributed to a cold
air masses front developed over Romania in that period of time.
In case of the 27th October 2004 earthquake, have been analyzed time-series of mean daily
air temperature and anomaly data for the period of 15 October -15 November 2004, on the
base period of normals 1981-2010 around the Vrancea region provided by NOAA satellites.
The positive air temperature anomaly started developing to North West and South East of
the epicentral area, air temperature showing an increase of around 4.8o C- 8o C during 24-27
October 2004 (Fig. 15), in good correlation with in-situ measurements, which revealed a
pronounced increase of air temperature over the Vrancea region.

1027
Zoran et al. Earthquake anomalies geospatial signatures

Figure 14 - Air Temperature anomaly over Vrancea region for 30 May

1990 earthquake.

Figure 15 - Air Temperature anomaly over Vrancea region for 27

October 2004 earthquake.

1028
 European Journal of Remote Sensing - 2016, 49: 1011-1032

After the main shock of 27th October, during 27 October -3 November 2004, air temperature
recorded a gradual increase with a maximum of 5oC degrees.

Conclusion
This study attempts to acquire information on seismic precursors in the Vrancea geotectonic
active area in Romania from in situ and satellite data (NOAA AVHRR and MODIS Terra/
Aqua time series). This study presents observations made using meteorological satellite
NOAA-AVHRR time-series data-derived outgoing long-wave radiation (OLR), land surface
temperature (LST) and air temperature (AT) in case of three strong Vrancea earthquakes (4th
March 1977, 30ty August 1986, 30ty May 1990) using anomalous TIR signals as reflected in
LST and AT rise and high OLR values which follow similar spatio-temporal growth pattern.
In addition, in order to investigate thermal anomaly pre-signal for a moderate earthquake
(27th October 2004) beside NOAA AVHRR time-series data have been used MODIS Terra/
Aqua time-series satellite data. In all cases, the spatio-temporal statistical analysis revealed
thermal anomalies enhancements of OLR, LST and AT parameters developed some days
to weeks before main seismic events and disappeared after the main shock, having a good
spatial correspondence with the epicenters and the active fault’s location as well as with
moment magnitude and preparation areas. This is in good accordance with the seismogenic
coupling processes between lithosphere-surfacesphere- atmosphere and for recognizing
earthquake anomalies with multiple parameters from integrated Earth observation system.
The analysis of all available geospatial data shows evidence of a thermal build up near the
epicentral area related to the investigated earthquakes in the Vrancea zone in Romania.
NOAA AVHRR time-series data analysis evidenced a transient thermal infrared rise in
land surface temperature ranging 5oC-10oC and outgoing longwave radiation increase
ranging 20-50 W/m2 around Vrancea epicentral area, function of latitude and longitude
around epicenters. Air temperature data recorded by meteorological in-situ observatories
and NOAA AVHRR time-series satellite data also support the development of short-term
thermal anomaly prior to earthquakes.
The results of pre-earthquake thermal anomalies recorded prior strong and moderate
earthquake analyzed in this paper are promising tools towards the future short -term
forecasting of the impending earthquakes in tectonically active region Vrancea as well as in
other seismogenic areas in the world.
As Earth science is multidisciplinary, it is very important as surveillance and research of
tectonic active areas to integrate all available satellite and observational information with
different temporal and spatial scales and review all the seismic, geodetic, geomorphological,
geological, geospatial data. Other precursory phenomena such as foreshocks and
nonseismological anomalies (electromagnetic, atmospheric, ground-water gas emissions,
etc.) must be well considered.

Acknowledgements
This research is supported by National University Research Council under Romanian
Ministry of Education and Research, PN II, PCCA Contract 86/2014 VRAGEO and
Program STAR Contract 73/2013 VRAFORECAST.

1029
Zoran et al. Earthquake anomalies geospatial signatures

References
Ardeleanu L., Leydecker G., Bonjer K.-P., Busche H., Kaiser D., Schmitt T. (2005)
- Probabilistic seismic hazard map for Romania as a basis for a new building code.
Natural Hazards and Earth Systems Sciences, 5: 679-684. doi: https://doi.org/10.5194/
nhess-5-679-2005.
Cervone G., Maekawa S., Singh R.P., Hayakawa M., Kafatos M., Shvets A. (2006) -
Surface latent heat flux and nighttime LF anomalies prior to the Mw = 8.3 Tokachi- Oki
Earthquake. Natural Hazards and Earth Systems Science, 6 (1): 109-114. doi: https://
doi.org/10.5194/nhess-6-109-2006.
Chmyrev V., Smith A., Kataria D., Nesterov B., Owen C., Sammonds P., Sorokin V.,
Vallianatos F. (2013) - Detection and monitoring of earthquake precursors: TwinSat, a
Russia-UK satellite project. Advance in Space Reserch, 52 (6): 1135-1145. doi: https://
doi.org/10.1016/j.asr.2013.06.017.
Demetrescu C., Wilhelm H., Tumanian M., Nielsen S.B., Damian A., Dobrica V., Ene M.
(2007) - Time dependent thermal state of the lithosphere in the foreland of the Eastern
Carpathians bend- Insights from new geothermal measurements and modeling results.
Geophysical Journal International, 170: 896-912. doi: https://doi.org/10.1111/j.1365-
246X.2007.03408.x.
Dey S., Singh R.P. (2003) - Surface Latent Heat Flux As An Earthquake Precursor.
Natural Hazards and Earth System Sciences, 3: 749-755. doi: https://doi.org/10.5194/
nhess-3-749-2003.
Dobrovolsky I.P., Zubkov S.I., Miachkin V.I. (1979) - Estimation of the size of earthquake
preparation zones. Pure Applied Geophysics, 117: 1025-1044, doi: https://doi.
org/10.1007/BF00876083.
Dobrovolsky I.P., Gershenzon N.I., Gokhberg M.B., (1989) - Theory of electrokinetic effects
occurring at the final stage in the preparation of a tectonic earthquake. Physics of Earth
Planetary Interiors, 57: 144-156. https://doi.org/10.1016/0031-9201(89)90224-0.
Erdal Y., Knížová P.K., Georgieva K., Ward W. (2016) - A review of vertical coupling in
the Atmosphere-Ionosphere system: Effects of waves, sudden stratospheric warmings,
space weather, and of solar activity. Journal of Atmospheric and Solar-Terrestrial
Physics, 141: 1-12. doi: https://doi.org/10.1016/j.jastp.2016.02.011.
Freund F. (2011) - Seeking out Earth’s warning signals. Nature, 473 (7348): 452. doi:
https://doi.org/10.1038/473452d.
Gruber A., Krueger A.F. (1984) - The status of the NOAA outgoing long wave radiation
data set. Bulletin American Meteorology Society, 65 (9): 958-962. doi: https://doi.
org/10.1175/1520-0477(1984)065<0958:TSOTNO>2.0.CO;2.
Martin M., Ritter J.R. (2005) - High-resolution teleseismic body-wave tomography
beneath SE Romania. Implication for three-dimensional versus one dimensional crustal
correction strategies with a new crustal velocity model. Geophysical Journal, 162: 448-
460. doi: https://doi.org/10.1111/j.1365-246X.2005.02661.x.
NOAA KLM user’s guide (2006) - Available online at: http://www2.ngdc.noaa.gov/docs/
klm/html/c7/sec71.htm.
Ouzounov D., Freund F., (2004) - Mid-Infrared Emission Prior To Strong Earthquakes
Analyzed By Remote Sensing Data. Advances in Space Research, 33: 268-273. doi:
https://doi.org/10.1016/S0273-1177(03)00486-1.

1030
 European Journal of Remote Sensing - 2016, 49: 1011-1032

Oth A., Wenzel F., Radulian M. (2007) - Source parameters of intermediate-depth Vrancea
(Romania) earthquakes from empirical Green’s functions modeling. Tectonophysics,
438: 33-56. doi: https://doi.org/10.1016/j.tecto.2007.02.016.
Pulinets S.A., Ouzounov D., Ciraolo L., Singh R., Cervone G., Leyva A., Dunajecka M.,
Kotsarenko A. (2006) - Thermal, Atmospheric and Ionospheric Anomalies around the
Time of the Colima M7.8 Earthquake of 21 January 2003. Annales Geophysicae, 24:
835-849. doi: https://doi.org/10.5194/angeo-24-835-2006.
Saraf A.K., Choudhury S. (2004) - Satellite detects surface thermal anomalies associated
with the Algerian earthquakes of May 2003. International Journal of Remote Sensing,
26: 2705-2713. doi: https://doi.org/10.1080/01431160310001642359.
Saraf A.K., Choudhury S. (2005) - NOAA-AVHRR detects thermal anomaly associated with
the 26 January 2001 Bhuj earthquake, Gujrat, India. International Journal of Remote
Sensing, 26: 1065-1073. doi: https://doi.org/10.1080/01431160310001642368.
Saraf A.K., Rawat V., Choudhury S., Dasgupta S., Das J. (2009) - Advances in understanding
of the mechanism for generation of earthquake thermal precursors detected by satellites.
International Journal of Applied Earth Observation and Geoinformation, 11: 373-379.
doi: https://doi.org/10.1016/j.jag.2009.07.003.
Singh R.P., Bhoi S., Sahoo A.K. (2001) - Surface manifestations after the Gujarat
earthquake. Current Science, 80: 1376-1377.
Stanica D., Stanica D.A. (2010) - Constraints on Correlation between the Anomalous
Behavior of Electromagnetic Normalized Functions (ENF) and the Intermediate Depth
Seismic Events Occurred in Vrancea Zone (Romania). Terrestrial Atmospheric and
Oceanic Sciences, 21: 675-683. doi: https://doi.org/10.3319/TAO.2009.09.09.01(T).
Takeshi S., Kanamori H., Yagi Y., Yamada M., Mori J. (2011) - Rebuilding seismology.
Nature, 473: 146-148.
Tramutoli V., Cuomo V., Filizzola C., Pergola N., Pietrapertosa C. (2005) - Assessing
the potential of thermal infrared satellite surveys for monitoring seismically active
areas. The case of Kocaeli (Ýzmit) earthquake, August 17th, 1999. Remote Sensing of
Environment, 96 (34): 409-426. doi: https://doi.org/10.1016/j.rse.2005.04.006.
Tronin A. (1996) - Satellite thermal survey-a new tool for the study of seismoactive
regions. International Journal of Remote Sensing, 17: 1439-1455. doi: https://doi.
org/10.1080/01431169608948716.
Tronin A.A., Molchanov O.A., Biagi P.F. (2004) - Thermal anomalies and well observations
in Kamchatka. International Journal of Remote Sensing, 25: 2649-2655. doi: https://doi.
org/10.1080/01431160410001665812.
Varotsos P.A., Sarlis N.V., Skordas E.S. (2011) - Scale-specific order parameter fluctuations
of seismicity in natural time before main shocks. Europhysics Letters, 96: 59002. doi:
https://doi.org/10.1209/0295-5075/96/59002.
Yusuf Y., Matsuko M., Yamazaki F. (2001) - Damage assessment after 2001 Gujarat
earthquake using Landsat-7 Satellite Images. Journal of the Indian Society of India, 29:
17-22. doi: https://doi.org/10.1007/bf02989909.
Zoran M., Weber C. (2007) - Use of multi-temporal and multispectral satellite data for
urban change detection analysis. Journal of Optoelectronics and Advanced Materials,
9 (6): 1926-1932.
Zoran M., (2012) - MODIS and NOAA-AVHRR land surface temperature data detects

1031
Zoran et al. Earthquake anomalies geospatial signatures

thermal anomaly preceding 11 March 2011 Tohoku earthquake. International Journal

of Remote Sensing, 33 (21): 6805-6817. doi: https://doi.org/10.1080/01431161.2012.
692833.
Zoran M., Savastru R., Savastru D. (2014) - Surveillance of Vrancea active seismic region in
Romania through time series satellite data. Central European. Journal of Geosciences, 6
(2): 195-206. doi: https://doi.org/10.2478/s13533-012-0158-z.
Zoran M., Savastru R., Savastru D. (2012) - a-Radon levels assessment in relation with
seismic events in Vrancea region. Journal of Radioanalytical and Nuclear Chemistry,
293 (2): 655-663. doi: https://doi.org/10.1007/s10967-012-1712-3.
Zoran M., Savastru R., Savastru D., Chitaru C., Baschir L., Tautan M. (2012b) - Monitoring
of radon anomalies in South-Eastern part of Romania for earthquake surveillance.
Journal of Radioanalytical and Nuclear Chemistry, 293 (3): 769-781. doi: https://doi.
org/10.1007/s10967-012-1780-4.

© 2016 by the authors; licensee Italian Society of Remote Sensing (AIT). This article is an open
access article distributed under the terms and conditions of the Creative Commons Attribution
license (http://creativecommons.org/licenses/by/4.0/).

1032