Role of Big Data Technologies in Water Information System

Hicham Jamil

, Bouabid El Mansouri

and Aniss Moumen

Natural Resources Geosciences Laboratory, Kenitra Faculty of Sciences, Ibn Tofail University, Morocco

Systems Engineering Laboratory, Kenitra ENSA, Ibn Tofail University, Morocco

Keywords: Information System, Big Data, Water Resources

Abstract: Water management is an essential vector that Morocco has adopted as an integral part of its government policy

since the dawn of independence. It also initiated an important project for the collection and analysis of data

relating to water resources.

Data and systems (Water Department, 2021). Necessary for water-related problems. The Hydraulic Basin

Agencies, which constitute the body producing data relating to water resources, through measurement stations

produce hydrological data, which will be consolidated at the Directorate General for Water level, thus making

it possible to make decisions. The volume of these data and the diversity of the actors leads us to think of a

water information system that integrates the concept of Big data to ensure good governance of these water

resources by exploiting various data from several databases.

In this article, we will compare existing architectures of water information systems according to a review of

systemic literature and propose an architecture by exploiting Big data technology.

1 INTRODUCTION

Water is considered an essential component of the

Moroccan economy. The combination of growth in

demand for water, climate, and hydrological gap

pushed water resource managers to search for

strategies for the management of water resources

(Loi, 2016).

So the management of water resources is a

strategy initiated by the Kingdom of Morocco since

independence.

The Directorate General for Water, relying on the

Hydraulic Basin Agencies, is required to measure and

collect data relating to water resources, to process,

consolidate it, to be able to make decisions, and ''

ensure perfect governance of water resources.

The Water Act of 1995 contributes to watershed

agencies' (ABH) several functions, including

collecting data and information on the situation of

water resources in a watershed (Loi, 2016).

Since its creation, the hydraulic basin agency has

not ceased to achieve strategic objectives by its

attributions;

 Operation and maintenance of hydraulic

structures, including dams

 water resources management at the hydraulic

basin level

 Issue authorizations and concessions for the

use of the public hydraulic do-main

 Taking measures relating to water resources

 Develop the Integrated Management Plan for

Water Resources (PDAIRE)

The Hydraulic Basin Agency (ABH) produces

and manages several Data, Has several applications

and databases for data management. By its Mandate

to take decisions based on the results of its

Information system (Moumen et al., 2015).

Based on this, the Hydraulic Basin Agency plays

a vital role in producing data relating to water

resources before their processing, modelling, and

analysis, allowing decision-making: the part of

decision-making information systems in this process

is essential.

Also, the volume of data produced at the HBA

level makes their exploitation difficult, leading us to

think of using the latest technologies relating to big

data.

The following points, therefore, represent the

problem:

Jamil, H., El Mansouri, B. and Moumen, A.

Role of Big Data Technologies in Water Information System.

DOI: 10.5220/0010735800003101

In Proceedings of the 2nd International Conference on Big Data, Modelling and Machine Learning (BML 2021), pages 419-425

ISBN: 978-989-758-559-3

419

 Which decision-making information system is

adequate for water management

 What are the components of this system?

 How to integrate the notion of big data into this

system, and at what level

2 METHODOLOGY

The literature review is an essential step for rich

documentation of the research. It is explained as

follows:

Figure 1: Working Methodology

In the end, we found that 93 references were

identified in the various databases and were stored in

the Zotero library. This number of references is

created using the following method

Figure 2: Constitution of the Zotero Library

Based on searches using the keywords used:

Information System, Water Resources, Big Data.

3 COMPARATIVE STUDY

Systematic Literature Review (SLR) has enabled us

to find articles that offer water information systems

incorporating the concept of big data. Indeed, several

works have dealt with this problem, which shows the

interest and importance of this subject, but no

universal architecture is final apart from these works.

We will try to present some actual results and try to

compare them:

The choice of these architectures based on those

related to the processing and management of data

relating to water resources

3.1 Big Data Open Platform for Water

Resources Management

By this architecture, the authors explain how we can

exploit Big Data concepts for environmental sciences.

(Driss et al., 2015)

The authors, in figure 3, presents the architecture

of a big Data Open Platform used for supporting

Water Resources Management; the architecture is

based on nine blocks:

1. Decision Support Tools: It is a model that

allows you to select the best decision module

2. Knowledge-Based System: This block

concerns the collection and storage of data

relating to water recooling and its exchange

with stakeholders

3. Geographic Information System (GIS):

Concerns the manipulation and analysis of

geographic data

4. Big Data Analysis System: This block relates

to big data processing and analysis for various

data relating to water resources

5. Simulation Models: This module contains

simulation models related to the GIS module

6. Computation and Processing: This part

contains calculators and simulators for

forecasting water resources

7. Communication System: The purpose of this

block is to ensure communication between the

different remote or local systems

8. Search Engine: This module ensures an

indexed search at the big data level

9. Users Interfaces: Provide the user's

capabilities of communicating with the

platform

ConstitutionoftheZOTEROlibrary

RISFileExport

NVIVOtreatment(Pillaietal.,2019)

Analysis

Referencesfromindexed

databases(133)

Articlesafterremoving

duplicates(110)

Referencesafter

verificationofeligibility(93)

BML 2021 - INTERNATIONAL CONFERENCE ON BIG DATA, MODELLING AND MACHINE LEARNING (BML’21)

420

Figure 3: Conceptual architecture of Big Data Open

Platform For Water Resources Management (Driss et al.,

2015).

3.2 A Framework for Processing Water

Resources Big Data and

Application

The development of technologies and the addition of

several aspects to water resources data makes the

dynamic analysis of this data more complex. The

authors' object in this paper is to present the

application of big data in this process and offers a

framework for processing water resources big data

and application (Ping and Zhao, 2014).

This framework, presented in Figure 4, mainly

consists of four layers:

1. Data acquisition layer:

 Collecting sufficient quantity, density, and

variety of real-time water resources data

 These collected data are divided into

structured data, semi-structured and

unstructured data

2. Resource organization layer

 Organization data

 Using SQL and NoSQL tools for data

extraction, integration, and transformation, to

form the Master database finally

3. Data analysis layer

 The core of big data processing

 Support the big data analysis and application

4. Application service layer

 Based on the data analysis layer

 Provide comprehensive information services

mainly for Water Re-sources System

Figure 4: A framework for processing water resources big

data (Ping and Zhao, 2014).

3.3 Big Data Technology in

Establishment and Amendment of

Water Management Standard

Exploiting new technologies for water resources

management, the authors build the integrated solution

of intelligent water management standards based on

big data in this paper.

In Figure 5, the authors present another

architectural approach based on the following steps:

1. Primary data resource platform:

 The data are stored in the backend

database of all business systems

 Support the regular stable operation of

the business system

2. Unified data platform:

 Data resources are managed in a

unified way

 Consolidate the data

 Data warehousing of intelligent water

management standards

3. Data resource utilization :

 Provides multidimensional data

analysis

 Assisting business decisions and

project approvals

Role of Big Data Technologies in Water Information System

421

Figure 5: The technical framework of information resource

(Bai et al., 2017).

3.4 Big Data Analytics for Water

Leakage in Bangalore

In this article, the authors discuss water leaks that

cause a significant water crisis in Bangalore (India).

They propose using big data analytics for water

management and the prediction of possible leaks

(Pillai et al., 2019).

For this, they present this platform, in figure 6,

based on:

Data acquisition - Scada model: This layer allows

to acquire the water data from the sensors

1. Data transformation & storage - Database:

The database that stores & transform the

data

2. Analytics layer - Prediction & Forecast:

Provides visualization of the data present in

the database and allow to make possible

predictions based on data processing

3. Business layer - Decision: The layer that

allows reporting the data and monitoring

the water leakage.

4. Presentation layer: this layer is at the top of

the platform and allows the visualization of

all the data present in the database

5. Presentation layer: this layer is at the top of

the platform and allows the visualization of

all the data present in the database (Pillai et

al., 2019)

Figure 6: Architecture overview of water leakage data

(Pillai et al., 2019).

In this article published in 2019, the authors

limited themselves to designing an architecture by

declaring that the implementation has not yet been

carried out.

4 DISCUSSION

The examination of the different architectures in the

previous phase allowed us to draw up the following

table, highlighting the advantages and weaknesses of

each architecture:

Presentation layer

Business layer - Decision

Analytics layer - Perdiction &

Forecas

Data transformation & storage -

Database

Data acquisition - Scada model

BML 2021 - INTERNATIONAL CONFERENCE ON BIG DATA, MODELLING AND MACHINE LEARNING (BML’21)

422

Table 1: Comparaison of big datas architectures.

Modules Status

Architecture 1

Chalh Ridouane

(2015)

 Decision Support Tools

 Knowledge-Based System

 Geographic Information System (GIS)

 Big Data Analysis System

 Simulation Models

 Computation and Processing

 Communication System

 Search Engine

 Users Interfaces

 The proposed framework is in the

design stage

Architecture 2

Ai Ping

(2014)

 Data acquisition layer

 Resource organization layer

 Data analysis layer

 Application service layer

 The platform remained at the

theoretical proposal stage

Architecture 3

Bai Y (2017)

 Basic data resource platform

 Unified data platform

 Data resource utilization

 The architecture was initially

proposed for the management of

water resources in China but posed

problems in terms of

implementation

Architecture 4

Deepthipriya

R Pillai

(2019)

 Data acquisition - Scada

 Data transformation & storage - Database

 Analytics layer - Prediction & Forecast

 Business layer - Decision

 Presentation layer

 The proposed framework is in the

design stage

Table 2: Comparison of the strengths and weaknesses of big data architectures.

+ -

Architecture 1

Chalh

Ridouane

(2015)

 Distributed on the web

 The presence of a module specialized in

decision making

 Integration of the geographic module

 Presence of a particular research

 Architecture does not follow a

process of data flow

Architecture 2

Ai Ping

(2014)

 Classification of input data

 Architecture follows the pre-flow of

information flow

 Presence of a layer for decision making

 Absence of a layer dedicated to

GIS data

 data consolidation based on SQL

only

Architecture 3

Bai Y (2017)

 Input data in several formats

 Architecture follows the pre-flow of

information flow

 Standardization of data

 Possibility of multidimensional analysis

 Absence of a particular module for

decision-making

 Lack of a layer dedicated to GIS

data

Architecture 4

Deepthipriya

R Pillai

(2019)

 The presence of a lawyer specialized in

decision making

 Presence of water prediction modules

(leaks)

 Big data technology is not

explained

 Platform designed for water leaks

Role of Big Data Technologies in Water Information System

423

Based on the comparison of the previous

architectures, the first remark, which one can note, is

that all the proposed designs remained in the state of

design without implementation or real test.

Indeed, the first architecture (Chalh Ridouane,

2015), composed of blocks, contains a block of

hydraulic models for data processing, a block for

simulations, and a central block relating to the big

data core.

In the second architecture (Ai Ping, 2014), the

authors proposed a framework that aims to make

decisions based on acquisition and processing data

from several databases and accommodates the user

from the acquisition phase to decision making,

always passing through a relative layer of big data.

In the third architecture (Bai Y, 2017), the authors

propose a platform that always follows the process

from data acquisition for decision-making and a

centre for the unification and standardization of data,

which come from several business databases.

The authors, in the final design (Deepthipriya R

Pillai (2019), present a model oriented towards water

leaks, with a treatment model and prediction models

but, like the last two architectures, always follow the

process of Acquisition  Treatment  Decision.

Figure 7: Comparison of big data water resources

architecture.

After bibliographic research on the authors of

these architectures, it seems to us that the authors,

without any real implementation or experimentation

in a real or quasi-real environment, proposed these

architectures.

This observation raises questions about the

possible obstacles and challenges to be presented to

successfully introduce big data to the water

information system of public actors in charge of water

resources management

After comparing the different architectures of the

authors, we came to the following conclusions for a

water information decisional system:

 The architecture must be structured according to

the flow of data processing flows

 It includes, as input, mechanisms for the

acquisition of several data formats from several

business databases

 It must have a block for the standardization of

data based on the concept of big data

 It must contain a layer for multidimensional data

analysis

 It must have a decision-making aid module

The treatment of articles through the nvivo tool

allowed us to establish the following degree of

similarity:

Figure 8: Degree of similarity of the articles.

Based on the above, we can propose the following

architecture for a water decisional information

system:

Figure 9: Proposed architecture.

The architecture, therefore proposed in Figure 7,

results from the comparison of the different

architectures studied. It is explained as follows:

 The data from business databases that contain

data relating to water resources (measurement

data, quality, GIS, etc.) are integrated into a

standardization module.

 This module allows the processing and

unification of the data format according to the

big data concept.

BML 2021 - INTERNATIONAL CONFERENCE ON BIG DATA, MODELLING AND MACHINE LEARNING (BML’21)

424

 The analytics module allows us to do our

various analyzes on this data to add a

multidimensional visualization layer

 Next, there is the decision module, which allows

you to define decision-making scenarios based

on the result of the process.

 The user interface is the leading portal for using

the platform, which contains an advanced search

module (metadata search, etc.).

We can thus, based on this functional architecture,

propose the following technical architecture

Figure 10: Technical architecture.

The proposed architecture contains the Hadoop tools

for the organization and analysis of data collected via

ETLs from the water basin agencies before being

exploited by users by Cassandra and Hbase and

interactively analyzed by Zeppelin.

5 CONCLUSIONS

The study showed us that several architectures are

proposed for the realization of an information system

for the management of water resources, but the

implementation, as well as the test, remains a

significant challenge to succeed to be able to present

an exciting result the decisions-makers

The perspectives of our study would be to study

in-depth the possible opportunities and challenges of

the adoption of big data by the administrations in

charge of water resource management in Morocco

REFERENCES

The website of the Water Department

Loi 36-15, Chapitre X (October 2016)

Aniss Moumen, Bouabid El Mansouri, Hassane Jarar

Oulidi, Lamiaa Khazaz. ‘Système d'Information sur

l'Eau au Maroc: Etat d'art, Problématique, Approche et

Prototype’, (Conference Paper ꞏ November 2015)

Hasnaoui Moulay Driss, Ouazar Driss, Bakkoury Zohra,

Chalh Ridouane 'Big data open platform for water

resources management (2015 International Conference

on Cloud Technologies and Applications (CloudTech),

Pages 1-8, June 2015

Ai Ping, Yue Zhao' A Framework for Processing Water

Resources Big Data and Application' Bai (Applied

Mechanics and Materials 519-520, 2014)

Bai Y, Bai X, Lin L, Huang J, Fang H.W, Cai K 'Big data

technology in establishment and amendment of water

management standard' (Applied Ecology and

Environmental Research 263-272 2017)

Deepthipriya R Pillai, P. Jeyalakshmi, Shwetha S P 'Big

Data Analytics for water leakage in Bangalore' (IJ RAR

Volume 6, Issue 2, May 2019)

ETL

Water Datas

Role of Big Data Technologies in Water Information System

425