IEEE Geoscience and Remote Sensing Magazine - March 2020 - 144

tion can provide important insights into the stability
conditions of landslides [57]-[59]. Eventually, such
geomechanical analysis may allow us to anticipate failure conditions prior to the pronounced accelerations of
the tertiary phase [60].
◗◗ Step 2: The potential impacts of active landslides are
assessed at a local scale. After the potential landslide
initiation hazard is identified for specific locations,
field investigations help assess the geological setting
of the landslide. A landslide dynamics model [61], [62]
can be applied to predict the speed and run-out extent
of potential landslide events. The potential landslide
sites identified in Step 1 can be simulated to determine the likely impact on human settlements for each
landslide. Topographic and sociospatial data can be
collated to model landslides and assess their impact.

◗◗ Step 1: Spaceborne InSAR is employed to comprehen-

sively detect active slopes (i.e., clusters of points that
exhibit certain deformational activity [53]) to find potential landslides at a regional scale. The archived and
newly acquired SAR images (e.g., ESA's Sentinel-1) are
interferometrically processed and then analyzed in
time series. An automatic feature-detection algorithm
(possibly relying on machine learning approaches,
e.g., [54] and [55]) should be developed to detect potential landslides based on the regional deformation
rate maps and displacement time series. Time series
analysis can be used to determine the sensitivity of
landslide motion to external factors such as seasonal
precipitation and seismic shaking [23], [56]. First-order
geomechanical modeling of landslide behavior based
on critical-state soil mechanics or rate-and-state fric-

II

III

Primary Creep

Secondary Creep

Tertiary Creep

Accumulated Displacement

I

Failure

Decreasing Rate

Steady Rate

Increasing Rate
Time

(a)
Typical Sampling Intervals

Velocity Scale

Phase

Temporal
Requirement

Satellite
InSAR

GNSS +
In Situ Sensors

Phase

Speed

Primary

Weekly→Monthly

Yes

Yes

Primary

Extremely
Slow to
Very Slow

Yes

Yes

Secondary

Weekly→Monthly

Yes

Yes

Secondary

Very Slow
to Slow

Yes

Yes
Yes

Daily→Hourly

Partly

Yes

Tertiary

Slow
Moderate
Rapid

Yes

Tertiary

No
No

Yes
No

(b)

Satellite
GNSS +
InSAR In Situ Sensors

(c)

FIGURE 5. An EO feasibility analysis of the three stages of a landslide. (a) The idealized displacement-time curves for the three stages of
creep [6], [41], [42]. (b) and (c) The typical sampling intervals and velocity scale analysis for satellite InSAR and in situ sensors in three creep
phases. The landslide speeds in (c) are defined according to [51] and [52], i.e., extremely slow (<16 mm/year), very slow (1.6 m/year), slow
(13 m/month), and moderate (1.8 m/h).

144

IEEE GEOSCIENCE AND REMOTE SENSING MAGAZINE

MARCH 2020
IEEE Geoscience and Remote Sensing Magazine - March 2020

Table of Contents for the Digital Edition of IEEE Geoscience and Remote Sensing Magazine - March 2020
