IEEE Geoscience and Remote Sensing Magazine - March 2017 - 62

is of particular interest to both humanitarian and financial
aid efforts. However, usually the necessary damage-related
information is difficult to attain from ground sources, as
in situ data collection faces challenges including time constraints, area accessibility, and the physical safety of surveyors [49], [70]. RS provides an applicable and credible source
of data that, when acquired and analyzed properly, can
provide invaluable knowledge about loss distribution [71].
Earth-observation technology offers a variety of information sources that characterize
the objects' spectral, spatial,
and geometrical properties.
CONSEqUENTLy, RS
Diverse types of information
HAS MADE TANgIbLE
can be extracted remotely from
CONTRIbUTIONS IN ALL
optical and RADAR sensors
PHASES OF THE DISASTER
that work on different ranges
RISK-MANAgEMENT CyCLE
of the electromagnetic specAT DIFFERENT SCALES,
trum and with changeable
wITH OFTEN ACCEPTAbLE
spatial resolution. A useful apINFORMATION ACCURACy.
plication for SAR technology
is the detection of both single-building and urban-block
damage. This task is made easier with the use of RADAR,
which has the superior ability to capture information about
the 3-D geometric features of any observed target objects
[69], [72], [73].
RS IN RECOVERY
RS imagery offers low-cost monitoring and evaluation of
the recovery and reconstruction actions, which are often
required by national governments and donor agencies [67],
[74]. The monitoring is carried out through the analysis of a
time series of satellite imagery that can be used to track the
postdisaster building reconstruction, the number of tents
in the displacement camps, or the vegetation recovery after
a wild fire [75]-[77].
DATA SUITABILITY
Data suitability defines the main outlines of the most desirable RS data sources, considering sensor type and scale.
The geographical scale, as a product of the image resolution, determines the level of the spatial information of the
mapping product. Data details are derived in four different
ways: spatial (objects' detectability and separability), radiometric (intensity differences), spectral (spectral distinction), and temporal (acquisition repeatability) resolution.
In the literature, several efforts have been made to investigate the properties of the main data types of different
resolutions [78] as well as the effect of selecting suitable
sensors that best serve the imagery objective [79]. Others
investigated the importance of the relationship between
scale and spatial resolution in producing thematic maps
[80]. However, in multirisk analysis, several objective-dependent factors must be considered. For instance, in response
and emergency activities, weather conditions add a constraint on optical data acquisition, and the proper usage of
62

complementary technologies like RADAR-based acquisition offers a back-up data source in such situations.
DATA INTEGRATION
The data-fusion process of spatial data types is an ongoing
activity in different scientific fields [78], [81]. In geoinformatics, the integrated spatial data types are often archived
either as vector data (e.g., points, lines, or polygons) or as
raster data (e.g., satellite images). The data integration can
be made among the RS acquisition systems where the fusion process occurs among data with different spectral and/
or spatial resolution (e.g., different RADAR scales, panchromatic with multispectral, LIDAR and multispectral, and
RADAR and multispectral [81]-[83]). However, the data
integration process is not restricted to a single data source
(e.g., RS) but often extends to include diverse data sources
(e.g., tabular data, thematic data, topographic maps, and RS
data) [84].
CONCLUSIONS
RS as a geospatial data source offers low-cost, broad-coverage
data acquisitions. Consequently, RS has made tangible
contributions in all phases of the disaster risk-management cycle at different scales, with often acceptable information accuracy.
Multirisk approaches are designed according to stakeholders' objectives, the data scale, and data availability.
Literature on multirisk assessment can be classified into
two main branches. The first approach estimates the multihazards and their associated vulnerabilities separately and
integrates the obtained overall hazard and overall vulnerability components to present a multirisk assessment. The
second approach presents the multirisk assessment as the
integration of a set of individual risks assessed distinctly.
However, the single-risk integration approach appears to
have an evident privilege for introducing a unified risk
metric of the elements at risk. Thus, it reduces the ambiguity in establishing the common units of hazard intensity
and vulnerability from the different risk types. Moreover,
the recent scientific trends, for both methods, are in creating generic and flexible multirisk evaluation systems as
well as in considering the dependency of both the hazards
and the vulnerabilities on the performed computations.
However, the current approaches often remain applicable
in the context they were designed for and require further
adjustments to be transferable.
Finally, the estimation of multirisk losses often demands extensive and multidisciplinary spatial analysis that requires
integrity among RS and in situ data sources. Thus, a risk manager has to consider availability, suitability, and integration
of the data sources used to extract risk-related information.
ACKNOwLEDgMENT
Financial support from the European Union FP7 project
Rapid Analysis and Spatialisation of Risk (RASOR), grant
606888, is gratefully acknowledged.
ieee Geoscience and remote sensing magazine

march 2017

Table of Contents for the Digital Edition of IEEE Geoscience and Remote Sensing Magazine - March 2017