An Early Flood Detection System Using Mobile Networks
An Early Flood Detection System Using Mobile Networks
Abstract—In this paper, an early flood warning system has the operational managers to decide a flood warning message
been studied and developed. The system consists of an early flood more quickly, precisely in real time. However, this research
warning station powered by solar energy, a monitoring center, did not bring out the way to transmit flood information to
and a notification system. The flood warning station is equipped
with the water level sensor and the precipitation sensor. The citizen. The short message service (SMS) was used in flood
data, which is received from sensors, will be stored in memory alerts in [6]–[8]. These systems used the same types of GSM
card and transmitted to monitoring center via GPRS protocol. modem to send the SMS messages which contain the water
The functions of monitoring center are to receive, analyse data level information measured from wireless water sensor, liquid
from the stations and send results to notification system where sensor or ultrasonic sensor. The users can receive the real time
warning can be generated to the considered residential areas.
The warning levels of flood can be set up easily by using the water level data in case that exceeds the predefined thresholds.
buttons in the circuit board or by using the SMS message from The General Packet Radio Service (GPRS) is a migration
user’s mobile phone. A website was also created to display the step toward the third-generation mobile network [9]. Based on
river water level, rainfall and the predicted information for the the available Global System for Mobile (GSM) network, GPRS
next few hours. technology is relative fast and its service cost is reasonable,
Index Terms—flood, GPRS, sensor, water level, precipitation.
so it is a good candidate in term of transmitting and receiving
data from the flood warning station to data center or to
I. I NTRODUCTION
the Internet. In this paper, GPRS plays as a transmission
Climate change has been a serious problem which attracts method to exchange information between the warning station
the interest of many countries in the recent years. One of the and monitoring unit. The duration between two messages is
clear effect of climate change is flood which has happened set up when the system is installed and its values change
more frequently in many regions and caused the devastating according to the flood warning levels. The correlation between
impacts on human lives and livelihoods. Many flood warning the precipitation and the river water level depends on the
stations have been developed and installed in prosperous location where the system is installed. Based on the measured
countries but the manufacturing cost is usually too high to data statistics during a long period at a considered area, the
be practical in developing countries. Therefore, building an lag time, is determined from the beginning of a rain to the
efficient flood warning system while maintaining reasonable beginning of an increase of the river water level, will be
production cost has been a meaningful mission for many calculated. Thus, we use the data from water level sensor, rain
researchers and manufactures. gauge to predict the warning level, and lag time to estimate
Various flood detection methods have been introduced in the occurrence time of the forthcoming flood.
the past decades. The authors in [1]–[3] analyzed the images The rest of this paper is organized as follows. The system
captured by satellite in different ways to detect the areas description which details the methodology to design and
where flood occurred. These techniques are only useful in implement the early flood detection system is presented in
flood localization but they can not predict whether flood occurs section II. Section III shows the operation of the designed
or not in the next hours. M. Oprea proposed a prototype system and the procedure to set up the initial parameters. The
intelligent system for flood warning and alert in real time [4]. results of system testing are described in section IV. Finally,
That system uses an ARM microcontroller, Marvell 88F6281 section V concludes the paper.
and the Unix FreeBSD for interface and data integration. To
make decision, the data will be transmitted from hydrological II. S YSTEM D ESIGN
station in real time via radio communication. E.Tate and K. The system model is described in figure 1, which consists
Cauwenberghs introduced a flood warning system relied on of three main parts: station flood warning, flood center and
web-based approach for the Demer basin in Belgium [5]. notification system. The flood warning station contains rain
In this system, a cluster of servers will collect and process gauge sensor, water level sensor and uses solar battery as
data from the hydrological observation stations in real time main power supply. The data received from sensors will be
and the available results can be displayed on client computer stored in memory card and transmitted through GSM module.
by remote access. The advantage of this system is to help Besides, the system has a flood center with a computer for
FP
62/$5%$77(5<
FP
92/7$*(5(*8/$725
FP
9 9 9
FP
&855(17
92/7$*( 6'&$5' (a) WL400 Sensor (b) Water Level Testing
&219(57(5
600
2015 International Conference on Advanced Technologies for Communications (ATC)
high sustainability. To supply reliable and low-cost rainfall consumption current is about 80mA. So this system can work
monitoring, sensor is built based on high impact ultraviolet continuously at least 87.5 hours without solar energy charging.
(UV) protected plastic. For each certain amount of rainfall,
III. S YSTEM O PERATION
a sealed reed inside the sensor will generate a mechanism
contact closure.
/ŶŝƚŝĂů
WĂƌĂŵĞƚĞƌ
^ĞƚƚŝŶŐ
tĂƌŶŝŶŐ>ĞǀĞůϭ
ĂƚĂĨƌŽŵ ŽŶƚƌŽůůĞƌ tĂƌŶŝŶŐ>ĞǀĞůϮ
^ĞŶƐŽƌƐ
tĂƌŶŝŶŐ>ĞǀĞůϯ
601
2015 International Conference on Advanced Technologies for Communications (ATC)
where H is the river water level, ζ is the lag time, Γ is using GSM module as in figure 8. The content of each
the rainfall, and α is the coefficient which describes the message includes the warning level, the measured data, and
correlation between water level and rainfall. The important the predicted occurrence time of flooding. Whenever the water
parameter for flood warning, ζ, depends on the geographical level exceeds the threshold, the station immediately sends the
and hydrological conditions of the considered area. This value data to the Internet that flood control unit could recognize that
is approximately 2 hours for Tuy Loan river which locates flood will occur. The users can view the necessary information
in central of Vietnam [10]. Next, we set up the thresholds easily by accessing the website in which data is displayed in
corresponding to warning water levels. Whenever the water bar graphs in figures 9 and figure 10.
level exceeds a predefined threshold, a warning message will
be released. Finally, set the measuring interval Ts which de-
termines the duration between two successively measurements
for each warning level in the following equation:
Ts = Ts − ki Δt, (2)
where Δt is the increment unit, ki ∈ {0, 1, 2, 3} is the warning
level. That equation means the higher water level, the shorter
duration between two measurements, so we will have more
accurate predicted results.
IV. R ESULT A ND D ISCUSSION
In this section, the results of our application test are shown
and discussed. We set up the conditions for testing in the
laboratory because it takes several months for having the
test results in reality. The figure 7 shows the overall PCB,
which has compact size 110x176 mm. All the modules and
a sequence of header pins for connecting with the sensor are
integrated on the board. The FT232 and RS232 are embedded
to interface with computer and text LCD to display system
status or sensor data values.
30
25
20
Rainfall (Cm)
602
2015 International Conference on Advanced Technologies for Communications (ATC)
50 R EFERENCES
40 [1] C. L. Lai, J. C. Yang, and Y. H. Chen, “A real time video processing
based surveillance system for early fire and flood detection,” in GIn-
30 strumentation and Measurement Technology Conference Proceedings,
Warsaw, Poland, May 2007, pp. 1 – 6.
20 [2] A. D Addabbo, A. Rece, G. Pasquariello, F. Bovenga, M. T. Chiaradia,
and D. Nitti, “A Bayesian network for flood detection,” in Geoscience
10 and Remote Sensing Symposium (IGARSS), Quebec, Canada, 2014, pp.
3594 – 3597.
0
[3] A. Scarsi, W. J. Emery, G. Moser, F. Pacifici, and S. B. Serpico, “An
11:20 11:30 11:40 11:50 12:00 12:10 12:20 12:30 12:40 12:50 13:00 13:10 13:20 13:30 13:40 13:50 14:00 automated flood detection framework for very high spatial resolution
Time imagery,” in Geoscience and Remote Sensing Symposium (IGARSS),
Quebec, Canada, 2014, pp. 4954 – 4957.
Fig. 10: Predicted Water Level Graph [4] M. Oprea, “A microcontroller-based intelligent system for real time
flood alerting,” International Journal of Computers, Communications
and Control, vol. 5, no. 5, pp. 844 – 851, Dec. 2010.
the data statistics of the location where the system is installed. [5] E. Tate and K. Cuawenberghs, “An innovative ood forecasting system
for the Demer basin: A case study,” Intl. J. River Basin Management,
The lag time ζ is 120 minutes which is the duration time that vol. 3, no. 4, pp. 1 – 5, 2005.
river water level start increasing by the effect of the rainfall. [6] E. Kuantama, L. Setyawan, and J. Darma, “Early flood alerts using
That means we have a precious time duration to have suitable short message service (SMS),” in International Conference on System
Engineering and Technology, Bandung, Indonesia, 2012, pp. 1 – 5.
actions to reduce the impact of the forthcoming floods. [7] W. A. H. W. M. Asmara and N. H. A. Aziz, “SMS flood alert system,” in
IEEE Control and System Graduate Research Colloquium, Shah Alam,
V. C ONCLUDING R EMARKS Malaysia, Jun. 2011, pp. 18 – 22.
A low cost early flood detection system has been designed [8] I. Aziz, I. A. Hamizan, N. S. Haron, and M. Mehat, “Cooperative
flood detection using GSMD via SMS,” in International Symposium on
and implemented. This system has high mobility characteristic Information Technology (ITSim), Kuala Lumpur, Malaysia, Aug. 2008,
due to the compact size of control board and solar energy pp. 1 – 7.
usage, so it can be installed easily. The flood warning is [9] A. A. Samjani, “General packet radio service [GPRS],” Potentials, IEEE,
vol. 21, no. 2, pp. 12 – 15, 2002.
divided into three levels depending on the measured data [10] E. Karlsson and K. Persson, “Analysis of hydrometeorological measure-
from sensors. For each warning level, the system will update ments in Tuy Loan river basin, Vietnam,” Master of Science Thesis,
the duration of getting data and sending result via SMS Lund University, 2005.
and Internet. This system can be improved to have better
603