Soil Moisture Sensor-Based Landslide Monitoring: A Laboratory-Based
Approach for Guwahati City
MADHUSHREE SHARMA, SHAKUNTALA LASKAR
Electrical and Electronics Engineering Department
Assam Don Bosco University
Azara, Guwahati, Assam
INDIA
Abstract: - Various techniques for landslide mapping, monitoring and modelling are being employed in a
variety of studies to keep people safe from landslides. Guwahati, a city in Assam (India) is surrounded by hills,
with varied slope angles, become prone to landslide during monsoon season. Relative increase in the moisture
content of soil is a major parameter for determining the occurrence of landslides that are induced by rainfall.
An experimental model with varying slope angles is demonstrated to witness some proportionality behaviour of
soil moisture value for the collected soil sample from landslide prone areas. The soil moisture sensor value
increases with increase in slope angle. The toe position of moisture value also shows a significant display of
data during landslide. This early warning module can be incorporated with the help of Blynk Application to
send messages to the residents of landslide prone areas. This study would be a cost effective alternative for
landslide early warning hazard monitoring and fast emergency response process and the model may be
considered as a miniature version of real-life slope conditions for the hills of Guwahati city, Assam, India.
Key-Words: - Breakdown, Slope Angle, Toe of Slope Area, Top of Slope Area, Display, Early Warning
Received: January 19, 2023. Revised: November 12, 2023. Accepted: December 15, 2023. Published: February 13, 2024.
1 Introduction
Natural catastrophes (landslide, flood, erosion etc.)
cannot be prevented, but we can prepare ourselves
by learning how to either mitigate them or set up an
effective early warning system. Landslide is one
such natural hazard that affects a small region ([1],
[2], [3]). Landslide mitigation techniques involve
mainly two methods- firstly, some advanced
techniques like installing of abutment and anchor
piles. However, the second option of installing a
suitable alarm and warning system is a more
economical option.
The unplanned construction due to rapid
urbanization has pushed the city of Guwahati to the
brink by destroying the natural balance of the city
[4]. Along with gravity, heavy rainfall, earthquake
and a slope that was cut, may also induce and
trigger landslide like conditions. As per report,
during June 14-June 21, 2022, around 72 landslides
took place across the hills of Guwahati owing to the
rain and as many as 266 families residing in
locations vulnerable to mud slips have been asked
by the District Administration to shift to the safer
places. Hence an urgent need for landslide early
warning process for the city of Guwahati is the
objective of the study. The early warning system of
landslide has 3 steps viz. landslide mapping,
monitoring and modeling ([5], [6], [7]).
There are mainly two causes of rainfall induced
landslides- reduction in soil’s shearing resistance
due to an increase in soil moisture content and
increase in unit weight of the soil [8]. Landslides are
further caused by changes in the slope angle, surface
erosion, and an increase in pore water pressure in
faults and joints [9].
Therefore, a rise in soil moisture content brought on
by precipitation infiltration plays a crucial role in
causing slope failure.
Additionally, slope collapses are usually caused by a
specific area at the slope's toe where the soil
moisture content nearly reached full saturation, even
when other areas of the sliding mass were still
International Journal of Environmental Engineering and Development
DOI: 10.37394/232033.2024.2.3
Madhushree Sharma, Shakuntala Laskar
E-ISSN: 2945-1159
27
Volume 2, 2024
partially saturated [10]. Here, a suitable
experimental setup was developed to analyze the
sensor readings. The results obtained during the
experiment, can be extended well to the real-life
landslide situations. Here, soil moisture sensors are
installed at different slope areas (upper and lower
set of sensors) and also at different distance from
the top and toe of the slope. It is feasible to forecast
when rainfall will cause a slope to fail by keeping an
eye on the percentage of soil moisture content on
the slope.
Numerous studies on landslide hazard zonation
mapping were identified during the literature
survey. There are several steep slopes in Hongkong
which become prone to landslide during heavy
seasonal rainstorms [11]. A relation between rainfall
and landslide was deduced with a threshold of 70
mm/hour. In [12], susceptibility map for landslides
has been created using neural network and
analytical hierarchy process (AHP) and fuzzy
methodology. A total of 5 risk categories were
found out which are very low, low, moderate, high
and very high. These categories were based on 5
parameters. These parameters directly influence the
process of landslide. Several researches on landslide
hazard zonation mapping were also found for the
research area. In Italy ([13], [14]), a study was
performed for landslide early warning with the help
of drones which are equipped with optical cameras.
Fuzzy interface System (FIS), Artificial Neural
Network (ANN), Genetic Algorithm (GA),
Unmanned Aerial Vehicle (UAV) photography may
also be used for landslide simulation and prediction
([15], [16], [17]).
To study landslide/debris flow, it is always advised
to prepare a landslide model in the laboratory ([18],
[19], [20]). Hence, numerous laboratory based study
in landslide/debris flow were also found in the
literature. While developing a framework for
community resilience from natural hazards in some
landslide prone areas in Afganistan, it was found
that some sort of early warning system may be
incorporated with the aid of sms, mosque
announcement etc. [21].
However, there is still a requirement for an
experimental landslide early warning system to
ensure effective execution of emergency responses
to landslides for the city of Guwahati, Assam
(India). Thus, this study aims to get a relationship
among all the parameters that induces landslide. In
Guwahati, out of 18 hills, 8 are found to be landslide
prone. The soil collected is from such a landslide
prone area of Guwahati City. Here, soil moisture
sensors are placed at various places of the
experimental set up and eventually tested with
varied induced rainfall amount.
Key benefits and utility of the study named “Soil
Moisture Sensor-Based Landslide Monitoring: A
Laboratory-Based Approach for Guwahati City” are
as follows:
a) Early detection of landslide
b) Loss in property and human casualty can also be
minimized.
c) Cost effective alternative for landslide early
warning system.
The research can also be combined with Artificial
Intelligence. With the help of Python, which is the
most simple, flexible and readable language, an AI
themed early warning system with the sensor data
obtained in the experimental study can be
developed. A simple flow chart may be seen as
below:
Previous data
Early warning threshold setting
Real time data
Threshold reached
Landslide early warning
2 Methodologies
2.1 Materials Used
This region's hill slopes are home to two distinct
types of soil. Mature residual soils are defined as the
highest layers, which are lateritic in origin and range
in thickness from a few centimeters to a few (1-2)
meters. They are distinguished by low permeability
values, a high percentage of clay size particles, and
a high plasticity index. Poorly graded silty sand can
be used to describe the saprolitic soils that lie
beneath the lateritic soils. ([22], [23]).
International Journal of Environmental Engineering and Development
DOI: 10.37394/232033.2024.2.3
Madhushree Sharma, Shakuntala Laskar
E-ISSN: 2945-1159
28
Volume 2, 2024
The experimental model (dimension 170 cm*
80cm*100cm), with the facility for changing the
slope, was filled with the sample soil collected from
Rani Region of Guwahati region. Method of
compartmentalization or quartering is used for soil
collection. Quartering process involves the division
of a well mixed sample into four equal parts [24].
Two opposite quarters are discarded and the
remaining two quarters are remixed until the
required amount of soil is obtained. After
compaction, the soil is ready for performing
experiment with the experimental set up (Figure 1).
A solar powered water pump is used for inducing
artificial rainfall which can sprinkle water uniformly
to the slope section of the set up. Soil Moisture is
one of the most important parameter when we talk
about landslide. Arduino compatible Soil moisture
sensor YL-38 (Figure 2) is utilized to measure the
electrical resistance of soil between the two
conductors of the probe. A variety of materials
make up the soil, some of which include minerals
and salts. These minerals and salts function as
electrolytes—which can conduct electricity—when
water is added and soil moisture sensor values are
displayed.
2.2 Study Area
Rani (a Block under Kamrup (Rural) District,
Assam, India) is chosen as the study area and the
soil sample is collected from the hills of that area
[25] (Figure 3). Rani area is located 93 kilometers
from the district headquarters in Amingaon.
Figure 1: Experimental set-up with the sprinkler
system
Figure 2: YL 38 soil moisture sensor
Figure 3: Study area map
2.3 Methods
The experiment was carried out with the induced
rainfall which is powered by a solar powered water
pump. The soil moisture sensors were placed at
various places. The rainfall is measured with the
help of a pre-calibrated water storage system. With
the addition of water, the conductivity of the soil
increases or resistivity of the soil decreases. A
change in the moisture content would alter the
voltage on the sensor’s analog output. Moisture
content scale would reflect these changes. The slope
angles can be altered with the flexibility provision in
the set up. Seepage system is also included in the set
up to perform experiment with the seepage water
value. Two sets of experiment were performed. First
soil moisture sensor values were found for different
slope angles. Second, by keeping the slope angle
constant, moisture sensors were placed at different
positions in the slope area.
International Journal of Environmental Engineering and Development
DOI: 10.37394/232033.2024.2.3
Madhushree Sharma, Shakuntala Laskar
E-ISSN: 2945-1159
29
Volume 2, 2024
3 Experimental Results and Discussion
3.1 Mathematical Formulation
Soil moisture index is a very important factor for
rainfall induced landslide model development [26].
Landslide data obtained from the Executive
Summary of Rapid Visual Screening for Potential
Landslide Areas of Guwahati city prepared by
District Disaster Management Authority are taken
for finding the threshold of the rainfall intensity and
duration. The intensity-duration (ID) threshold has
the general form [27],
 (1)
Where, I- Mean or average rainfall intensity
D- Rainfall Duration
c, α, β- Parameters that vary from places to places.
Before, rainfall is introduced to the experimental set
up, soil moisture index M0 (an important parameter
in determining the landslide occurrence) is assumed
to be negative. Let this value be FC (mm) where FC
is the field capacity.
Therefore,
= - (2)
Now, soil Moisture index at time t,
 (3)
Here, Mt-1 is the soil moisture extent at time t-1,
Rt is the effective rainfall
Dt is the drainage
And Et is the daily evaportranspiration
For calculation of drainage amount, drums of known
capacity may be installed in the experimental set up.
3.2 Experimental Results
As stated earlier, the experimental set up has
provision to change the slope angle of the set up,
hence by changing the slope angle, with same
amount of induced artificial rainfall, soil moisture
sensor values are found out. The soil moisture
sensor (YL-38) values are displayed in the 7
segment display and are plotted accordingly (Figure
4). As confirmed from the graph, soil moisture value
(in percentage) for 40 degree slope angle (grey
colored curve in the graph), is more than 30 degree
slope angle (blue colored curve in the graph).
To verify the toe and top region dependency on
landslide, the soil moisture sensors were placed at
two different positions. Figure 5 and Figure 6
pictorially illustrate the readings obtained in Table 1
and Table 2. As evident from the graph, (x-axis
mentions the rainfall amount (in L) and y-axis
mentions the soil moisture sensor value (in
Percentage)) the toe region plays an important role
in the determination of landslide event occurrences.
As the sensors are placed more towards the toe or
top, the lower sensor’s displayed data are
comparatively more than the upper sensors
displayed data.
The saturation point before breakdown (after about
50 L of induced rainwater), the soil moisture value
does not change much for all the following cases.
The result shows an increase in the soil moisture
sensor value as the position of the sensor is coming
near the toe region. The breakdown is achieved at
about 70 L of induced rainfall. Initial condition was
captured at time 12:14 pm and the final breakdown
was captured at 01:01 pm (Figure 7). Whereas, the
initial figure was clicked at time 11:12 am and the
final breakdown was clicked at 11:50 pm (Figure 8).
Figure 4: Soil moisture sensor value for different
slope angles
Figure 5: Upper sensor and lower sensor value for
sensors placed at half distance from top and toe
International Journal of Environmental Engineering and Development
DOI: 10.37394/232033.2024.2.3
Madhushree Sharma, Shakuntala Laskar
E-ISSN: 2945-1159
30
Volume 2, 2024
Figure 6: Upper sensor and lower sensor value for
sensors placed at one third distance from top and
toe.
Table 1: Upper and lower soil moisture sensor value
for half distance position
Table 2: Upper and lower soil moisture sensor
value for one third distance position
Figure 7: Initial and breakdown conditions
when sensors are placed at half distance
International Journal of Environmental Engineering and Development
DOI: 10.37394/232033.2024.2.3
Madhushree Sharma, Shakuntala Laskar
E-ISSN: 2945-1159
31
Volume 2, 2024
Figure 8: Initial and breakdown conditions when
sensors are placed at one third distance
4 Sensitivity Analysis
The phenomenon of landslide depends on various
parameters such as rainfall, humidity, temperature
and most importantly soil moisture value. The above
study was performed by considering increase in soil
moisture sensor as the main parameter for
occurrence of landslide. The occurrence of the
landslide event may be identified with the help of an
accelerometer sensor MPU 6050 that will provide
the x-axis and y-axis displacement of the soil.
Regression analysis was performed by taking rain
fall, humidity and soil moisture data as input and
displacement reading from accelerometer as the
output. The Regression Statistics shows that R2
value is the highest for soil moisture sensor data
(Figure 9) proving the fact that soil moisture data is
the most crucial factor in determining the landslide
event occurrence i.e. rainfall induced landslides are
most sensitive towards the change in soil moisture
value.
Regression Statistics for
Rainfall Data
Multiple R
0.823017
R Square
0.677357
Adjusted R Square
0.631265
Standard Error
0.253621
Observations
10
Multiple R
0.925273
R Square
0.856129
Adjusted R Square
0.835576
Standard Error
0.16936
Observations
10
Regression Statistics for Soil
Moisture Sensor
Multiple R
0.967163113
R Square
0.935404488
Adjusted R
Square
0.927330049
Standard Error
0.185028704
Observations
10
Figure 9: R2 value displayed on graph
5 Conclusions
The landslides in this region of North East is rainfall
induced, therefore the sprinkler system that was
used in the experimental set up is more advisable
than the seepage technique used in. Another
observation from the tables and their corresponding
graphs is that at first the change in sensor value is
substantial compared to the marginal change after a
particular value for equal amount of rainfall. This
result can be integrated with wireless sensor
network or artificial intelligence. Arduino based
YL-38 soil sensor system provides a cost-effective
solution for this natural/ human made hazard. The
project can be extended to offer additional
improvements for various industrial applications,
including but not limited to construction sites,
infrastructure projects, and precision agriculture
practices [28] and oil and natural gas pipelines.
International Journal of Environmental Engineering and Development
DOI: 10.37394/232033.2024.2.3
Madhushree Sharma, Shakuntala Laskar
E-ISSN: 2945-1159
32
Volume 2, 2024
Acknowledgement:
Authors are thankful to all the staff and faculty
members of Departments of Electrical and
Electronics Engineering (EEE) and Civil
Engineering (CE) Assam Don Bosco University,
Azara, Guwahati-781017
References:
[1] Hidayat R., Jonson S., Hidayah A., Ridwan B.,
Mulyana A., Development of Landslide Early
Warning System in Indonesia, Geoscience, Vol
9, No 10, 2019,
https://doi.org/10.90/geoscience9100451.
[2] Bai, S., Wang, J., Bell, R., Glade, T.,
Distribution and Susceptibility Assessments of
Landslide Triggered by Wenchuan Earthquake
at Longnan. In Proceedings of the International
Conference on Informatics, Cybernetics, and
Computer Engineering (ICCE2011),
Melbourne, Australia, 19–20 November 2012.
[3] Froude, M.J.; Petley, D.N., Global Fatal
Landslide Occurrence from 2004 to 2016. Nat.
Hazards Earth Syst. Sci. 2018, pp 2161–2181.
[4] Sarma P.C., Landslide Hazard Assessment of
Guwahati Region using Physically Based
Models. 6th Annual Conference of the
International Society for Integrated Disaster
Risk Management (IDRIM-TIFAC), New Delhi,
India, 2015.
[5] Aleotti P., A Warning System for Rainfall-
Induced Shallow Failures. Engineering
Geology, Vol 3, No 73, pp 247–265.
[6] Ramesh M., Pullarkatt D., Geethu T.H., and
Rangan P., Wireless Sensor Networks for
Early Warning of Landslides: Experiences from
a Decade Long Deployment, Landslides,
Vol 13, No 4, 2017, pp 833-838.
[7] Sharma M., Laskar S., Landslide Mapping,
Monitoring and Modelling Techniques: A New
Approach using DOFS, 2017 International
Conference on Circuits, Controls, and
Communications (CCUBE), Bangalore, India,
2017, pp. 21-24, doi:
10.1109/CCUBE.2017.8394149.
[8] Chaulya S., Slope Failure Mechanism and
Monitoring Techniques. Sensing and
Monitoring Technologies for Mines and
Hazardous Areas , Elsevier, 2014.
[9] Montrasio L., Shallow Landslides Triggered by
Rainfalls: Modelling of Some Case Histories in
the Reggiano Apennine (Emilia Romagna
Region, Northern Italy). Natural Hazards,
2014, pp 1231–1254.
[10] Piciullo, L., Calvello, M., Cepeda, J.M.,
Territorial Early Warning Systems for Rainfall-
Induced Landslides. Earth Sci. Rev. 2018, 179,
228–247
[11] Brand E.W., Premchitt J., Phillipson H.B.,
Relationship between Rainfall and Landslides
in Hong Kong. Proceedings of 4th
International Symposium on Landslides,
Toronto, pp 377-384
[12] Bernardo E., Palamara R., Boima R., UAV and
Soft Computing Methodology for Monitoring
Landslide Areas, WSEAS Transactions on
Environment and Development, doi:
10.794/22015.2021.17.47.
[13] Rossi G., Tanteri L., Tofani V. et al.,
Multitemporal UAV Surveys for Landslide
Mapping and Characterization, Landslides,
Vol.15, 2018, pp. 1045-1052.
https://doi.org/10.1007/s10346-018-0978-0
[14] Segoni, S.; Lagomarsino, D.; Fanti, R.; Moretti,
S.; Casagli, N. Integration of Rainfall
Thresholds and Susceptibility Maps in the
Emilia Romagna (Italy) Regional-Scale
Landslide Warning
System. Landslides 2015, 12, 773–785.
[15] Jiacheng Z., Chonglong W and Xinglin G, A
Dynamic Simulation Algorithm based on
Multitask Spatiotemporal Data Model, WSEAS
Transactions on Computers, Vol 14, 2015.
[16] Ma S., Xu C., Shao X., Zhang P., Liang X,
Tian, Y., Geometric and Kinematic Features of
a Landslide in Mabian Sichuan, China, derived
from UAV Photography, Landslides, 16, 2019,
pp. 373- 381
[17] Glenn N.F., Streutker D.R., Chadwick D.J.,
Thackray G. D., Dorsch S.J., Analysis of
LiDAR-Derived Topographic Information for
Characterizing and Differentiating Landslide
Morphology and Activity. Geomorphology
73(1): 2016, pp 131–148
[18] Evangelista S., Marinis G., Cristo C., Leopardi
A., Dam Break Dry Granular Flows:
Experimental and Numerical Analysis, WSEAS
Transactions on Environment and
Development, Vol 10, 2014.
[19] E Yuliza et al., Study of Soil Moisture Sensor
for Landslide Early Warning System:
Experiment in Laboratory Scale, 2016, J. Phys:
Conf. Ser. 739012034.
[20] Osanai, N., Shimizu, T., Kuramoto, K.,
Kojima, S., Noro, T., Japanese Early-Warning
for Debris Flows and Slope Failures using
Rainfall Indices with Radial Basis Function
Network. Landslides 2010, 7, pp 325–338.
[21] Mohanty A., Mishra M.,Hussain M., Kattel D.,
Exploring Community Resilience and Early
Warning Solutions for Flash Flood, debris
International Journal of Environmental Engineering and Development
DOI: 10.37394/232033.2024.2.3
Madhushree Sharma, Shakuntala Laskar
E-ISSN: 2945-1159
33
Volume 2, 2024
Flow and landslides conflict prone Villages of
Badakhshan, Afganistan, International Journal
of Disaster Risk Reduction, January 2019, Vol-
33, pp 5-15.
[22] Sarma H., Granitization of the Gneissic Rocks
in the Rani - Pamohi Area, Kamrup Metro,
Assam, India. International Journal of
Research and Analytical Reviews Vol 6: 2018,
pp 112-124
[23] Maswood M., Pathak, R., Migmatites Around
Maliata and Dakhola,Kamrup, Assam.
Jour.Ass.Sci.Soc., 1983, Vol.25,No.2. P.70-75.
[24] Carter M. R., Gregorich E. G., Soil Sampling
and Methods of Analysis, Second Edition (New
York: Taylor and Francis Group), 2008.
[25] Mazumder D., Benchmark Survey of
Rajapanichandra Village in Rani Block of
Kamrup District in Assam, Economic Affairs,
Vol 60, No 2, 2015, pp 237-241
[26] Guzzetti F., Peruccacci S., Rossi M., Stark C.
P., Rainfall Thresholds for the Initiation of
Landslides in Central and Southern Europe
Meteorology atmospheric physics 98 239-67,
2017
[27] Irwan A., Virgianto R., Safril A.,Munawar,
Gustono S.,Putranto N., Rainfall Threshold and
Soil Moisture Indexes for the Initiation of
Landslide in Banjarmangu sub district central
Java, Indonesia, IOP Conf. Series Earth and
Environmental Science 243 (2019) 012028.
Doi: 10.1088/1755-115/24/012028
[28] Sireesha T., Kalyani M., Gowthami D., Design
of Autonomous Vehicle for Precision
Agriculture using Sensor Technology, WSEAS
Transaction on Environment and Development,
Vol 14, 2018.pp 155-158.
Contribution of Individual Authors to the
Creation of a Scientific Article (Ghostwriting
Policy)
The authors equally contributed in the present
research, at all stages from the formulation of the
problem to the final findings and solution.
Sources of Funding for Research Presented in a
Scientific Article or Scientific Article Itself
This research is supported by Assam Science
Technology and Environment Council
(Department of Science, Technology and Climate
Change, Govt. of Assam) under Student Science
Project Scheme vide letter number ASTEC/S &
T/206/2019-2020/1243-1287 dated 05-05-2020.
Conflict of Interest
The authors have no conflicts of interest to declare
that are relevant to the content of this article.
Creative Commons Attribution License 4.0
(Attribution 4.0 International, CC BY 4.0)
This article is published under the terms of the
Creative Commons Attribution License 4.0
https://creativecommons.org/licenses/by/4.0/deed.en
_US
International Journal of Environmental Engineering and Development
DOI: 10.37394/232033.2024.2.3
Madhushree Sharma, Shakuntala Laskar
E-ISSN: 2945-1159
34
Volume 2, 2024