Content uploaded by Sk Ajim Ali
Author content
All content in this area was uploaded by Sk Ajim Ali on Jun 30, 2020
Content may be subject to copyright.
Mapping of mosquito-borne diseases in Kolkata Municipal
Corporation using GIS and AHP based decision making approach
Sk Ajim Ali
1
•Ateeque Ahmad
1
Received: 4 November 2018 / Revised: 5 January 2019 / Accepted: 8 January 2019
Korean Spatial Information Society 2019
Abstract Mosquito-borne diseases are those which trans-
mitted through the bite of an infected mosquito. Stagnant
water bodies are often preferable for breeding sites of
mosquitos. But from breeding eggs to final stage, there are
many factors responsible for its incubation, maturation and
growth enough to bite and transmit diseases. The main aim
of present study is to focus on associated environmental
factors that provide suitable breeding sites and suscepti-
bility mapping of mosquito-borne-diseases through
geospatial technique and decision making approach. Ana-
lytic hierarchy process as a decision making approach was
integrated with geographic information system to map of
mosquito-borne diseases in Kolkata Municipal Corpora-
tion. Choice based various ranking was used to decide the
weights of selected factors through establishing pairwise
comparison matrix. Initially, 10 relevant environmental
factors were chosen to determine their weight through
pairwise comparison matrix. Concomitantly, weight of
each causative factor was used as geo-database to support
overlay analysis. Consistency ratio was calculated to check
the decision process and significance measurement. The
consistency ratio of decision factors was calculated as
0.0470, which is \0.1 and considered as consistent and
acceptable. The study analysis shows that proximity to
water bodies is a major responsible factor and subsequently
moisture content, water index, availability of shadow area
and presence of vegetation are also influential factors in
prevalence of mosquito-borne diseases. The present result
shows the high applicability of satellite data and spatial
technique in epidemic diseases zonation.
Keywords Mosquito-borne diseases KMC
Environmental factors GIS AHP Detection of
vulnerable zones
1 Introduction
The prevalence of mosquito-borne diseases (MBDs) are
associated with towns, metropolitan and mega cities in
contrast to rural areas. It is because of demographic and
many environmental factors [1]. This epidemic situation is
totally dependent on the local factors such as households,
population, level of literacy, constructional activities, sit-
uation of waste generation and storage, slum clusters,
proximity to drain and water bodies locations etc. which
play vital role in the transmission of mosquito-borne dis-
eases [1–3]. Effort has been made to understand the spatial
distribution of MBDs and level of risk with areal extension
in Kolkata Municipal Corporation (KMC) based on pos-
sible and relevant decision making parameters.
The outbreaks of mosquito-borne diseases depend on
three important factors, specific parasite (malaria/dengue),
human or animal host and the carrying agents (mosquitos).
The process starts when the microbes or virus enter to
parasites as soon as it bite an infected person. The parasite
drinks up such infected blood along with its germs, and the
germs start growing in the body of a parasite with an
average time period 7–10 days [4]. This time period is
called incubation period. After growing such germs while
Electronic supplementary material The online version of this
article (https://doi.org/10.1007/s41324-019-00242-8) contains sup-
plementary material, which is available to authorized users.
&Sk Ajim Ali
skajimali.saa@gmail.com; skajimali@myamu.ac.in
1
Department of Geography, Faculty of Science, Aligarh
Muslim University (AMU), Room No. 21, S. M. E. (Upper),
S. S. Hall (North), Aligarh, U.P. 202002, India
123
Spat. Inf. Res.
https://doi.org/10.1007/s41324-019-00242-8
the same carrying agent bites a healthy person; it transfers
some of the germs into the new host and causing infection.
As malaria and dengue are the major recognised mosquito-
borne disease in the study area, thus mosquito is considered
as the chief parasite in this regards. Hence, the present
study only highlights on parameters suitable sites for
mosquito’s breeding and considered for mosquito-borne
disease.
The distribution of mosquito-borne diseases is associ-
ated with different lively factors like demographic, social
and environmental. Unexpected urbanization, urban growth
and climatic factors including climate change, variation in
temperature and rainfall, variation in moisture contents can
make influence on mosquito’s breeding and growth [5,6].
Moreover, various environmental factors such as changes
in local land use, local sanitation conditions, drainage
facility in and around households can also affect the
transmission of mosquito-borne diseases due to getting
suitable sites for breeding and can render large populations
in towns and cities at risk of viral diseases spread by
mosquitoes [7–9].
About 2.5 million malaria cases are reported annually
from South-east Asia, out of which India alone contributes
76% [10]. Over 2.5 billion people (about 40%) of the
world’s population—are now at risk from dengue [11].
WHO recently estimates that there may be 50–100 million
dengue infections worldwide every year. Kolkata as a
metropolitan city, outbreak of malaria and dengue cases are
seen every year with varying frequency. The record of
malaria cases in KMC has shown a decline number from
42,200 cases in 2011 to 5564 cases in 2016 and in this year
more than 2874 cases till August 2018. The record of
dengue cases shown that the number of dengue infections
fluctuating in different year. While 1852 dengue cases were
reported in 2012 but only 238 cases were reported in 2013.
Again peak dengue incidences were seen in 2016 when
1686 people were reported with dengue positive. But on an
average every year near about 400–500 dengue cases are
found in KMC.
Recently, different initiatives has launched by KMC
authorities to control and reduce mosquito-borne diseases
from the capital city including establishment of mosquito
research laboratory, establishment of dengue detection
centre, dissemination of dengue-report through SMS Alert,
disease surveillance system strengthened, plying of
speedboats along canals for destruction of mosquito larvae,
larvicidal spray along canals by using rowing-boats as
transport, formation of 21 rapid action teams for vector
control, efficient non-medical staff assigned with vector
control responsibilities etc. Along with such type of ini-
tiatives some micro level initiatives are also taken includ-
ing localised vector detection centres, multilingual leaflets,
multi-coloured flex-banners, multi-coloured hoardings,
awareness meeting, comprehensive booklet, documentary
film, auto-miking and house visit etc. for awaking the
residents to lessen the extent of risk of mosquito-borne
diseases.
These all initiatives are considered as good but the best
way ever considered for prevention of mosquito-borne
disease is to detect and destruct the suitable breeding site of
vector carrying bloodsuckers or mosquitos. Recently,
remote sensing (RS) data and techniques of geographical
information system (GIS) proves its efficiency in the
detecting and recognising the suitable breeding sites of
vector carrying parasites, vulnerable risk zones and area
under prone to risk using spatial analysis tool [12]. Dif-
ferent studies have been designed using RS data and GIS
application in malaria vulnerability, dengue risk zonation
and susceptibility zonation of other vector-borne diseases
over time [1,3,12–15]. The application of satellite images,
digital elevation model and GPS location with superior
spatial analysis has great significance for current environ-
ment strategy in urban areas [16]. Use of Geospatial
technology has amazing result to identify and demarcate
malaria vulnerable areas [17]. Application of satellite
images with various resolutions i.e. IRS LISS I, LISS II
and WiFS also provides suitable idea in recognising vector
habitats and vulnerable areas [18]. It is always required to
reliability of authentic and accurate data sets for vector-
borne diseases control and management in order to predict
vulnerable areas and project risk mapping [19,20]. Mos-
quito producing fields were identified in California through
correlating Anopheles larva density with reflectance of
canopy growth in early season by using Landsat TM
imagery and GIS technique [21]. Detection of many vector
diseases is not possible directly through field visit and
observations, thus decision making approach along with
GIS and AHP technique was used for malaria risk mapping
in West Singhbhum district of the Jharkhand, India [1].
Integration of various thematic layers of socio-economic,
geographic and epidemic factors was made to familiar with
malarial hotspots [22]. Moreover, there are different stud-
ies in the different geographical location of earth which has
shown the capability of RS and GIS technique in detecting
vector-borne diseases [12,23–29].
Vulnerability and mapping on spatial information of
such mosquito-borne diseases is required to reduce the
prevalence of such diseases, as destruction of vector is best
method to consider ever. In order to understand the spatial
distribution of mosquito-borne diseases, it is required to
use decision hierarchy with closely associated factors
which are considered for growing and developing parasitic
mosquitos. Thus, going through many research findings, it
is considered that application of AHP is the best way to
prepare decision hierarchy in complex situations like
weighting multiple factors in mosquito-borne diseases
123
S. A. Ali, A. Ahmad
(MBDs) and use of GIS technique and high resolution
satellite data to find out vulnerable risk areas. Many studies
have combining both analytic hierarchy process and GIS
for better decision making in public health assessment and
diseases detection. Similar effort was made in this work.
The main objective of present work is to use different
environmental factors as decision criteria for mapping of
Mosquito-borne diseases. The present work also aimed to
highlights the efficacy of remote sensing data and GIS
technique along with analytic hierarchy process as a multi-
criteria decision tool for mapping of mosquito-borne dis-
ease. Such mapping of mosquito-borne diseases can deliver
assistance to planning authorities to comprehend spatial
distribution under risk and planning for better surveillance.
2 Materials and methods
2.1 Description of the study area
Kolkata is the capital of the Indian state of West Bengal
which is located on east bank of the river Hooghly. A
typical riverine city, in the earlier days, it was surrounded
by marshes, tidal creeks, mangroves, swamps and wet-
lands. But now all these have changed with time. Kolkata
Municipal corporation (KMC) is the largest urban
agglomeration of West Bengal and is located in UTM Zone
45N with geographical extension of 222800000 Nto
223703000N and 881403000 Eto882503000E (Fig. 1). The
KMC has an area of 205.07 km
2
which is divided into 16
administrative Boroughs and respectively 144 wards. The
mean elevation of the city is 1.5 m to 16 m above MSL.
Many part of the city was originally a wetland which was
reclaimed over the time to house the bourgeoning popu-
lation. The annual mean temperature over Kolkata
Municipal Corporation is 26.8 C or 80.24 F. Monthly
mean temperature ranges between 19 and 32 C. Summers
are hot and humid with temperature 34 C, during May and
June the temperatures often exceed 40 C. Winter or Mild
winter lasts for highly two and half month with lowest
temperature between 9 and 11 C in December and Jan-
uary. The annual average rainfall is near about 1600 mm
and highest rainfall occurs during the month of August.
According to latest population data (DSH 2014-15) the city
Fig. 1 Location map of the study area
123
Mapping of mosquito-borne diseases in Kolkata Municipal Corporation using GIS and AHP based…
has 4,496,694 populations and 1,024,928 households. The
population density of KMC is about 24,306 persons per
km
2
where, household density ranges between 755 and
23,237 households per km
2
by making KMC as India’s 3rd
largest metropolitan city as well as the world’s 8th largest
urban agglomeration.
2.2 Study design
It is a complex task to take decision regarding growth of
mosquitos and transmission of diseases because always not
one single criterion is responsible for any incidence. Thus
Multi-criteria decision approach was applied in present
study to spatial map of mosquito-borne diseases. Analytic
hierarchy process (AHP) is one of the more commonly
applied multi-criteria based decision making approach.
AHP allows decision maker to structure the decision cri-
teria into a hierarchy of sub-criteria or alternatives which
can then be analysed individually. This approach is
essentially based on pair wise comparisons matrix (PCM)
which allows decision makers to make compare with
individual criteria and alternatives to establish the weight
to calculate the importance or performance score. Hence,
looking the objective of present study, AHP was used. It is
a multi-step plan including selection of relevant crite-
ria/factors, data inputs, data processing, analysis and make
findings. Kolkata Municipal Corporation was chosen to
carry out such type of study because the occurrence of
mosquito-borne diseases like dengue and malaria have seen
every year during early August to late October and claims
many lives year after year. Thus, proper susceptibility
zonation of MBDs can enhance better surveillance to
reduce the effects.
The important factors that watchfully connected to
MBDs occurrence were preferred looking towards the aim
of the study. First of all, 10 associated factors were selected
and then to check the decision accuracy, the consistency
ratio (CR) was calculated using Saaty’s method of each
selected decision factor. For spatial analysis, the spatial
database of each factor was created in GIS Environment.
GIS integration was run using weighted linear combination
(WLC) method to prepare spatial susceptibility map of
MBDs.
2.3 Data collection
As per requirement, different types of satellite data were
used including Landsat 8 OLI and TIRS, Sentinel-L2A,
SRTM Global Digital Elevation Model (DEM), GPS
locations points and other field visit data (Table 1). The
Landsat OLI and TIRS with 30 m resolution and UTM
projection by default (Path: Row 138/44) were collected
from United States Geological Survey (USGS) earth
explorer portal and specific bands were obtained looking
towards purpose of use. Landsat 8 data were mainly used
for extracting land surface temperature (LST) and Pro-
portion of Vegetation (PV). Sentinel-L2A data (45QXE
and 45QXF) with 10 m and 20 m spatial resolution along
with WGS 1984 projection by default were collected from
Sentinel Hub EO browser. The Sentinel-L2A data were
mainly utilised for preparing land use land cover (LULC),
normalised difference vegetation index (NDVI),
Table 1 Data source and purpose of use in mapping of mosquito-borne diseases
Dataset Data type Spatial resolution Source Date Purpose of uses
Landsat 8 (OLI
and TIRS)
Raster Reflective bands (4 and
5) with 30 m; thermal
bands (10 and 11) with
30 m
USGS-earth
explorer portal
14.04.2017 Land surface temperature, proportions of
vegetation
Sentinel-L2A Raster 10 m (band 3, 4 and 8)
and 20 m (band 8A
and 11
Sentinel Hub EO
browser
11.03.2018
and
17.10.2017
Land use land cover, normalised difference
vegetation index, normalised difference
moisture index and normalised difference
water index
Shuttle Radar
Topography
Mission
(SRTM) Global
DEM
Raster SRTM 1 Arc with 30 m earthdata.nasa.gov 2015 Land elevation, slope and aspect map
Water bodies Vector;
converted
to raster
Resized into 10 m Field survey, GPS
locations
2018 Proximate analysis
Base map Vector Resized into 10 m The Kolkata
Gazette, Govt.
of W.B
2011 Geo-reference; delineation of outer
boundary, ward and borough map
123
S. A. Ali, A. Ahmad
normalised difference water index (NDWI) and normalised
difference moisture index (NDMI). SRTM global DEM
(N22E088) with 30 m resolution and WGS 1984 projection
by default was collected from https://earthdata.nasa.gov.
Using SRTM DEM data land elevation, slope and aspect
map were prepared. Excluding such satellite data, location
of water bodies was also considered in this study because
water bodies offer best breeding ground to mosquitos.
Hence, those living nearness to water bodies may be con-
sidered as high risk zones for malaria and dengue cases.
Thus, as far as possible the GPS locations of many water
bodies i.e. small to big ponds, canals and lakes were
recorded and further polygon shape files were created for
proximate analysis. Finally, these all input layers were
reclassified accordingly in GIS environment with equal
projection and cell size for further processing and analysis.
2.4 Data processing: selection and preparation
of decision criteria for MBDs
2.4.1 Land use land cover
Land cover types has significant role in giving sites for
suitable breeding environment. Major portions of land
under water bodies, low area played a significant role on
the incidence of mosquito-borne disease like Chikungunya,
dengue and malaria [30]. Alterations of land use type
including impoundments, dams, irrigation and draining
systems that create shelters for the mosquitoes which have
potential effect in transmitting vector-borne diseases like
malaria, dengue, and filariasis [31]. Thus, preparation of
land use land cover map is crucial to classify major area
under different land cover and their uses in order to
understand those areas vulnerable to mosquitos breeding
site.
The land use land cover (LULC) map was prepared by
compositing bands of 10 m resolution from Sentinel-2A
dataset. Primarily band combination was performed with
Near Infrared (Band 8, 10 m), Red (Band 4, 10 m) and
Green (Band 3, 10 m) to make false color composite (FCC)
infrared. Secondly, object based signature file was created
to run supervise classification and finally, the entire area
was classified into five major land use types (Fig. 2).
2.4.2 Normalised difference vegetation index
Many researchers emphasized that normalised difference
vegetation index (NDVI) is a significant factor for ana-
lysing mosquito borne diseases. Mosquito-borne diseases
were found to be the highest proportion in places with low
land and lowest forest cover [30]. Moreover, present-day
deforestation seems to be associated with the expansion of
Fig. 2 Land use land cover of
Kolkata Municipal Corporation
123
Mapping of mosquito-borne diseases in Kolkata Municipal Corporation using GIS and AHP based…
mosquito distributions and rise in mosquito-borne disease
transmission [31]. Vegetation indices, specially based on
the normalized difference vegetation index, the habitat
suitability for different species of mosquitoes can be
described [32–34]. The amount of precipitation can also be
linked with vegetation index. The green vegetation is
related with higher amounts of precipitation and is con-
sidered as suitable environment for mosquito habitats [33].
The mosquito population is positively correlated with
normalized difference vegetation index [35].
NDVI is a globally accepted criterion and has been in
application for more than 37 years; from it was first
established by Rouse et al. [36]. It is also highly useful and
multi-purpose application. Different remote-sensing satel-
lites hold bands in the red (R) and near infrared (NIR)
which taken for NDVI calculation and most significantly,
the NIR to red wavelength ratio is correlated with absorbed
photosynthetically active radiation (APAR).
In order to know the vegetation index (VI) of KMC,
NDVI was calculated (Fig. 3). It is used for quantifying the
green vegetation. NDVI normalizes the green leaf scatter-
ing in the Near Infrared wavelength and chlorophyll
absorption in the red wavelength. Hence, Near infrared and
Red bands are required. Band 8 (NIR, 10 m) and Band 4
(Red, 10 m) of Sentinel-2A dataset was used for the
purpose and NDVI was calculated using the following
equation;
NDVI ¼NIR RED
NIR þRED
where NIR is Band 8 and RED is the Band 4. As per
Sentinel-2A EO products guideline, the value of NDVI
range in -1to?1. Negative values (values approaching
-1) represent to water bodies. Values close to zero (i.e.
-0.1 to ?0.1) mostly represent to barren areas of rock,
sand, or snow. Low but positive values represent shrub and
grassland (mostly it ranges from 0.2 to 0.4) and high
positive values indicate dense forest.
2.4.3 Land surface temperature
Surface temperature always plays a vital role in the growth
of mosquito larva. Land surface and air temperature effect
the feeding behaviour of the mosquito. The survival and
the life cycle of vectors are also effected by land surface
temperature [37]. Mosquito survival is low at extreme
temperatures i.e. very high or very low temperature is
harmful for mosquito’s growth. The optimum temperature
in which mosquito can survive is found at 28–32 C and
temperatures \5C and or [40 C are deadly for the
mosquito [37].
Fig. 3 Normalised difference
vegetation index to show nature
of vegetation cover in different
areas
123
S. A. Ali, A. Ahmad
The land surface temperature derived from the Radiative
Transfer equation based method using band 10 of Landsat
8 TIRS data has the highest accuracy in comparison to
Qin’s Split window Algorithm and the Single Channel
method which have moderate and low accuracy respec-
tively [38]. Therefore, the land surface temperature (LST)
of KMC was derived from band 10, thermal band of
Landsat 8 dataset (Fig. 4). It is a long process to calculate
LST using radiative transfer equation based method.
Firstly, the conversion of Digital Number to top of atmo-
sphere (TOA) radiance has to do. OLI and TIRS datasets
can be converted to TOA spectral radiance using the
radiance rescaling factor which provided in metadata file of
Landsat dataset. Thus following equation was used;
Lk¼MLQcal þAL
where L
k
= TOA spectral radiance [Watts/(m
2
srad lm)],
M
L
= band specific multiplicative rescaling factor, Q
cal-
= quantized and calibrated standard product pixel value,
A
L
= band specific additive rescaling factor taken.
Now, the top of atmospheric brightness temperature has
to calculate from the reflectance value using the equation;
BT ¼K2
In K1
Lkþ1
273:15
where BT = atmospheric brightness temperature in Kelvin,
which is further subtracted by 273.15 to calculate the
degree Celsius, K2 = band specific thermal conversion
constant, K1 = band specific thermal conversion constant,
L
k
= TOA spectral radiance.
Finally, brightness temperature was converted to land
surface temperature using the following equation;
LST ¼BT=1þkBT=c2ðÞLnðeÞðÞðÞ
where BT = brightness temperature, k= wavelength of
emitted radiance, c2 = h*c/s = 1.4388 910
2
-
m K = 14,388 lm (h, c and s are constant), e = emissivity
which was calculated from Near infrared and Red band of
the same dataset.
2.4.4 Normalised difference moisture index
Many variables are interrelated and correlated with each
other. Like the wet areas have high moisture content and
concomitantly high vegetation. The normalised difference
moisture index (NDMI) was calculated using NIR (20 m)
and SWIR (20 m) Bands of Sentinel-2A (Fig. 5). NIR is
useful for classifying the vegetation and SWIR is good for
measuring the moisture content of soil and vegetation, i.e.
the reflectance in the NIR band is influenced by the leaf
internal structure and the SWIR reflects the changes in
Fig. 4 Land surface
temperature to show the
temperature differentiation in
different part of KMC
123
Mapping of mosquito-borne diseases in Kolkata Municipal Corporation using GIS and AHP based…
vegetation water content. In SWIR, a darker area highlights
higher water content. Thus, following equation was applied
to measure moisture content;
NDMI ¼NIR SWIR
NIR þSWIR
where NIR is for the Near infrared (Band 8A) and SWIR is
the Short-Wave Infrared (Band 11). For Sentinel-2A data,
these bands are with 20 m spatial resolution. Normalised
difference moisture index have values ranging from -1to
?1, where -1 indicates very bad moisture level and ?1
indicated very high moisture level.
2.4.5 Normalized difference water index
Only water bodies can provide breeding sites for parasites
like mosquito. The spatio-temporal variations in water
indices allow for the detection of deficiencies which can in
turn lead to accumulations of disease vectors and resulting
disease outbreaks [39]. Normalized difference water index
(NDWI) can be utilised to explain open water bodies
depending on the reflectance bands [40,41].
To analyse water index of KMC area, Band 3 and Band
8 of Sentinel-2A with 10 m resolution of each were uti-
lised. Band 3 have the capability to contrast between clear
and muddy water and fairly well penetrates clear water.
Man-made features are still visible in this band. On the
other hand Band 8 is well considered for mapping the
biomass content, as well as at detecting and analysing
vegetation. The following equation was used to calculate
NDWI;
NDWI ¼GREEN NIR
GREEN þNIR
where GREEN is Band 3 (10 m) and NIR is Band 8
(10 m). The high positive value of normalized difference
water index represents presence of water bodies and zero or
negative value represents built up areas (Fig. 6).
2.4.6 Proportion of vegetation
High proportion of vegetation means land cover with lot of
leaf, which after decomposition produces litters with high
moisture content and may offers suitable environment for
the mosquitoes to place eggs, growth of larvae, pupae of
vectors to grow in comparison to low vegetated areas [1].
The proportion of vegetation or PV value was calculated
using Landsat 8 dataset (Fig. 7). As it represents the veg-
etation index, thus NDVI values have to calculate first in
order to know the highest and lowest value of NDVI. For
the same NIR and RED Bands of Landsat 8 image was
used and flowing equation was run;
PV ¼NDVINDVImin
ðÞ=NDVImaxNDVImin
ðÞðÞ
2
Fig. 5 Normalised difference
moisture index displays the
moisture content
123
S. A. Ali, A. Ahmad
Fig. 6 Normalized difference
water index of KMC
Fig. 7 Proportion of vegetation
123
Mapping of mosquito-borne diseases in Kolkata Municipal Corporation using GIS and AHP based…
where NDVI
min
and NDVI
max
represents the lowest and
highest value of calculated normalised difference vegeta-
tion index. Generally, PV appears in positive value and
ranges in 0–1. The value 0 or close to it indicates area
under non-vegetation and value 1 or close to 1 indicates
area under high vegetation cover.
2.4.7 Land elevation
In most of the cases mosquito-borne diseases were noted in
the low land areas mainly covered by wetlands and water
bodies i.e. consisting channels, streamlets and lakes, all
these are considered as suitable mosquito breeding sites
[37]. Kolkata has very limited variation in term of topog-
raphy. But land elevation has significant role and consid-
ered to reproduce its influence in risk zonation. The
elevation of landforms in KMC varies between the lowest
point (-28 m) to the highest point ([13 m). The lowest
portions are mainly encircled by water bodies and low
laying areas. The rest areas are varies between 5 and 10 m
in height.
Several studies proposed the significant role of land
elevation in risk zonation of mosquito-borne diseases like
dengue and malaria. The low laying areas always offer
suitable site for mosquito breeding, that’s why in risk
assessment; the low areas are taken as more vulnerable
zones as compared with high areas. Topography plays as a
major role in determination of Breeding Habitats to
Malaria Risk and Vulnerability [42]. Among many other
parameters, land elevation can also be considered in
assessment of dengue risk zonation [12].
2.4.8 Slope
Slope is also another important factor for defining habitats
of mosquito larval growth [42]. The steeper slopes are
always allow water movement faster and disturbs stability
of aquatic habitats [8]. In comparison, the surface water
movement in gentle slopes is nearly stagnant and this
creates favourable condition for mosquito breeding sites
[1]. Therefore, with other environmental and topographic
factors; slope was also considered in present study as an
input layer. The slope of KMC ranges from less than 1to
above 12. The areas under 0–1slope indicates water
bodies and other flat lands which have very gentle slope
and suitable condition for mosquito habits in comparison
with high multi-stair buildings and other built-up areas
which have moderate to steep slope.
2.4.9 Aspect
Aspect is an important parameter to know the area under
shadow. Usually, the shadow areas are considered as low
air temperature and high moisture. Hence, aspect is taken
as important factor for vector-borne parasite transmission
[42]. The northern aspect ranges from NW340to NE70
and is called as shadow area. The shadow areas not receive
direct sunlight, where the content of moisture is too high
and relatively the proportion of vegetation is also found
high as compared with the other aspect. Shadow affects the
larval distribution of mosquitos and hence, more vulnerable
to vector growth [1].
Informations regarding these three factors were col-
lected from global digital elevation model. 30 m resolution
DEM of SRTM 1 arc was used to analysis topographic
features including aspect, slope and elevation which were
considered as significant factors in determining vulnerable
zones of vector-borne diseases. Primarily, DEM data was
collected which is in Geographic Coordinate System
(WGS_1984). Conversion in Universal Transverse Mer-
cator coordinate system (UTM_45N) was done in GIS
environment for further analysis. Finally spatial analysis
tool was used for analysing these three topographic fea-
tures (Figs. 8,9,10).
2.4.10 Proximity to water bodies
The mosquitoes have a typical flight ranges up to 2 km
depending upon the species. Aedes mosquitoes (main car-
rier of dengue virus) breeds in fresh stagnant water and has
a flying range of only 100–200 m, maximum 300–500 m,
while female Anopheles mosquito (main carrier of malaria
virus) has a flying range of maximum 2 km [43,44]. In
urban areas the presence of wastewater points, which serve
as productive mosquito habitat and increase the risk [37].
Thus, the water bodies were considered as an important
factor in this study for identifying risk zonation of mos-
quito-borne diseases. For the same, the proximate analysis
was carried out in GIS environment to create buffer dis-
tance from location of water bodies with common interval
of 200 m (Fig. 11).
The plot of exact location of water bodies is a very tiring
task but their accurate locations are also essential to detect
area under risk and vulnerable assessment of mosquito-
borne diseases because water bodies offer suitable site for
growing of vector carrying parasites. Thus, effort was made
to locate the water bodies (small to large ponds, canals and
lakes) for proximate analysis. The task was performed in
three stages. Initially, during the field survey (June to July
2018) in each Borough of KMC, the GPS locations of such
water bodies from different points were recorded (Fig. 12).
Concomitantly, in second stage, Google Earth Pro was used
123
S. A. Ali, A. Ahmad
Fig. 8 Land elevation
Fig. 9 Slope to show land surface inclination
123
Mapping of mosquito-borne diseases in Kolkata Municipal Corporation using GIS and AHP based…
Fig. 10 Aspect indicates the
shadow and non-shadow areas
Fig. 11 Proximity to water
bodies
123
S. A. Ali, A. Ahmad
to create KML layer of water bodies based on these
recorded GPS locations. Finally, the KML layer converted
into shape file (polygon entity) using conversion tool in
GIS environment and multiple ring buffers were created for
proximity analysis.
2.5 Analytic hierarchy process
In this study, analytic hierarchy process was used as a
multi-criteria decision approach to make final decision
towards developing MBDs susceptibility zonation of
Kolkata Municipal Corporation. The AHP was developed
and first used by Thomas L. Saaty. The AHP was designed
to solve multifaceted problems relating multiple criteria. In
this method, the importance of criteria is compared pair-
wise with respect to the desired goal to develop their
weights and then consistency of judgments is checked in
order to confirm a reasonable level of consistency in cri-
teria based decision making. Thus, the first of all a hier-
archy for the decision have to build which is also called as
decision modelling [45]. As the all selected criteria will not
have the same significance for a particular instance, hence,
the relative priorities or weights for the criteria is to derive
in next step. It is called relative because priorities are
measured with respect to each other based on their
importance in an incidence. Then the pair-wise
comparisons have to perform to prepare comparison matrix
of the criteria to measure the weightage.
PCM is a relative ranking based matrix table used to
calculate the weight value of all alternatives under a
selected criterion in decision making process. Ci and j of
the matrix, is the measure of importance of the item in row
i when compared to the item in column j. AHP finds the
significance ranking of Cj,i by calculating the reciprocal of
Ci,j.
Cj1
Cj2
Cj3
:
:
Cjn
NCS
1=Cij11=Cij1... 1=Cij1
1=Cij21=Cij2... 1=Cij2
1=Cij31=Cij3... 1=Cij3
:: :
:: :
1=Cijn 1=Cijn 1=Cijn
RCij RCij ... RCij
0
B
B
B
B
B
B
B
B
@
1
C
C
C
C
C
C
C
C
A
w1
w2
w3
:
:
:
wn
where C is the selected criteria, j is the rank of that criteria,
i is the reciprocal rank of Cj, NSC is normalised column
sum, w is the weight of that criteria and 1, 2, 3 and n
indicating the No. of alternatives.
The significance rank of all alternatives is compliments
and divides of the same alternatives in the matrix [46,47].
The AHP employs an underlying scale with values from 1
to 9 to rank the relative importance for two items (Table 2).
PCM gives the weights of each criterion with comparison
to all others. Once the comparison rank is fitted, the
Fig. 12 Location of water
bodies in KMC (used for
creating buffer distance)
123
Mapping of mosquito-borne diseases in Kolkata Municipal Corporation using GIS and AHP based…
weightage have to be calculated for each respective alter-
native to judge the consistency for considering into deci-
sion. Therefore, the Eigenvector was calculated by the
considering the following equation:
Ax ¼kmaxX
Calculated through, (w1 * Cij1) ?(w2 * Cij2) ?
(w3 * Cij3) ?(wn * Cijn)
kmax ¼Ax1;Ax2 ...:: Axn=wx1 ;wx2...::wxn
where A is the comparison matrix of size n 9n, for ‘n’
number of criteria, also considered as priority matrix, x is
the Eigenvector of size n 91 which also called as priority
vector, w1…wn are the weight of alternatives, Cij1…..
Cijn are the rank of alternatives and kmax is the
Eigenvalue.
Now, it is required to check that result is either con-
sistent or not. As the numeric ranks are derived from the
subjective preferences of individuals, it is highly possible
to have some personal bias in making the final matrix of
judgments. The question is how much inconsistency is
acceptable. For this purpose, AHP always offers a measure
of the consistency of PCM by computing a consistency
ratio (CR) comparing the consistency index (CI) of the
matrix. Saaty [48] provides the calculated RI value for
matrices of different sizes of alternatives as given here
(Table 3). The ratio is designed in such a way that if the
value of the ratio exceeding 0.10 is considered inconsistent
for judgments and value 0 is considered as perfectly con-
sistent [48]. The value 0 or close to 0 (i.e. 0.02 or 0.05) are
highly acceptable. Consistency ratio (CR) was calculated
through the help of CI and RI, whereas, CI was calculated
using following equation;
CR ¼CI
RI
CI ¼kmax n
n1
where RI = Random index, CI = consistency index,
kmax = Average of Rw1…wn. RI depends on the number
of elements being compared (i.e. number of alternatives in
PCM) and if the value ranges from 0 to 0.09, the matrix
will be considered as reasonably consistent and hence, may
continue the process of decision-making using AHP.
Different researchers have applied AHP as a multi-cri-
teria decision making method for vector-borne diseases
assessment. Analytic hierarchy process was used for
monitoring and early warning system to develop the pre-
paredness and awareness for malaria disease in Kan-
chanaburi, Thailand [49]. The integrated Geospatial and
Multi-Criteria Evaluation (AHP) technique was applied to
identify malaria risk zones in Vadodara district, Gujrat,
India [50]. Vulnerability zones were identified for malaria
epidemic using AHP approach and geospatial tools in
Chakardharpur sub-division of the West Singhbhum dis-
trict of Jharkhand, India [1]. Mapping of malaria hazard
and risk was drawn using spatial Multi-criteria evaluation
technique along with GIS in Tekeze Basin Development
Corridor, Amhara Region, Ethiopia [3]. GIS based Ana-
lytic hierarchy process was used for dengue risk mapping
of Kolkata Municipal Corporation, India [12].
3 Result and discussion
KMC is treated as come in high to medium vulnerable
zones in term of mosquito-borne diseases (MBDs). A
protracted and complex decision making process was
Table 2 Numeric scale to
establish pair-wise comparison
matrix adopted from Saaty
AHP scale of importance for PCM Numeric rank Reciprocal rank (decimal)
Extremely importance 9 1/9 (0.11)
Very strong to extremely importance 8 1/8 (0.12)
Very strongly importance 7 1/7 (0.14)
Strongly to very strongly impotence 6 1/6 (0.17)
Strongly importance 5 1/5 (0.20)
Moderately to strongly importance 4 1/4 (0.25)
Moderately importance 3 1/3 (0.33)
Equally to moderately importance 2 1/2 (0.50)
Equally importance 1 1 (1.00)
Table 3 Numeric value of
random index (RI) to measure
consistency
N23456789101112
RI 0 0.58 0.9 1.12 1.24 1.32 1.41 1.45 1.49 1.51 1.53
123
S. A. Ali, A. Ahmad
carried out through selection, rating, weighting and ana-
lysing of multi-criteria. Present study used 10 decisions
making factors. Thus, the hierarchy was constructed with
10 decision criteria and 47 alternatives or sub-criteria. The
selected environmental factors are associated with sur-
rounding environmental and topographic features. Total 10
factors were chosen looking towards their relative impor-
tance in mosquito growth and diseases transmission these
factors are land use land cover (LULC), normalised dif-
ference vegetation index (NDVI), land surface temperature
(LST), normalised difference moisture index (NDMI),
normalised difference water index (NDWI), proportion of
vegetation (PV), land elevation (LE), slope (S), aspect
(A) and proximity to water bodies (PWBs). Thus, a total 47
sub-factor were given risk value in a scale from 1 to 5, (5
defines very high and 1 defines very low vulnerable zones
of mosquito-borne disease) based on the scale of impor-
tance adopted after Saaty which ranges from 1 to 9.
Table 4shows 47 sub-factors which were attributed by
different importance from 1 (equal important towards
MBDs outbreak) to 9 (Extremely important towards MBDs
outbreak) and subsequently risk values were arranged to
define which sub-factor has greatest vulnerability to MBDs
(Table 4).
AHP is considered as semi-quantitative technique as for
standardizing numeric scale along with qualitative aspect.
The main function of AHP is Pair-wise comparison based
on their relative importance which helps in measuring
quantitative judgment in new fields [48]. Thus, expert
choice based intensity of importance was given to partic-
ular sub-factor looking their association towards mosquito-
borne diseases and pair-wise comparison matrix was
formed to obtain the weightage value. To check the result
and error of decision making and empirical bias during
putting ranks in respective field, the Consistency Index (CI)
and Consistency Ratio (CR) were considered. The weights
of all 47 sub-factors along with CI and CR are summarized
here (Table 5). Saaty suggested that if the ratio of CI and
index for the resultant random matrix are found [0.1, the
decision making process and selection of rank may be
considered as inconsistent [51]. The consistency ratio of all
selected factors including LULC, NDVI, LST, NDMI,
NDWI, PV, LE, SL, AS and PWBs were calculated as
0.0503, 0.0152, 0.0375, 0.0313, 0.0381, 0.0404, 0.0347,
0.0434, 0.00 and 0.0311 respectively. The result shows that
the calculated value of consistency ratio always less \0.1
for all the selected factors. Thus, it may considered that the
selection of rank were always in acceptable choice when
pair-wise comparison matrix was prepared.
Based on weightage calculated from pair-wise compar-
ison as presented in Table 5, shows that moisture index
with value of 0.3268–0.6009, water index with value
0.0247–0.3132, B200 m distance from water bodies and
shadow area have been providing greatest risk of mosquito-
borne diseases because these range of values always con-
sidered for offering suitable sites from breeding and growth
of mosquitos. Along with, gentle slope with 0–1.3877,
proportion of vegetation with value 0.1295–0.2313, vege-
tation index with value 0.4312–0.6898 and land covering
with water bodies are also considered as high risk zones for
mosquito-borne diseases. Similarly, low land i.e. minus
elevation (below surface such as water bodies), were also
found as great risk to MBDs outbreak.
The each sub-factor was arranged with risk scale on the
basis of weightage calculated from PCM. Now, to derive
final output layer in GIS environment, the comparison
matrix of selected 10 factors was established along with
weightage to overlay (Table 6). Each factor is not have
equal importance and not offers equal suitable site for
breeding. Thus, previous research, environmental condi-
tions for Aedes and Anopheles growth were taken into
consideration while choice was made to put risk rank to
respective factor. The result shows that out of all selected
causative factors, proximity to water bodies have the
greater risk with 0.2297 weight value. It means that in
developing mosquito-borne disease, nearness to water
bodies is greater responsible than other selected factor. The
next most responsible factors are NDMI and NDWI with
weight of 0.1593 and 0.1394 respectively. Land surface
temperature (LST) has also significant weight value of
0.1146, as comfortable temperature always required for
mosquito growth. Correspondingly, LULC, NDVI, pro-
portion of vegetation, land elevation, slope degree, aspect
are also playing vital role in prevalence of mosquito-borne
diseases in Kolkata Municipal Corporation with weights of
0.06297, 0.0689, 0.0888, 0.0401, 0.0435 and 0.0522
respectively. The consistency ratio was calculated as
0.0470, which is considered under acceptance for com-
parison matrix.
Finally, the mosquito-borne diseases vulnerable zones
were extracted through weight based overlay of the chosen
factors using the GIS tool (Fig. 13). The result shows that
the proportions of very high and very low vulnerable zones
are limited and maximum areas come under moderate
MBDs vulnerable zones. The southern part of Borough I
and northern part of Borough II come under very high risk
zone of mosquito-borne diseases, along with different part
of Borough VI, VII, VIII, IX, X and XV come in the same
category.
The choice based weight was assigned in present study.
Similar results were also found from several related stud-
ies. Nearness to water bodies and water logged areas are
more prone to mosquito carrying diseases like dengue [12].
Climatic factors especially temperature and moisture con-
tent in the ambient air are associated with mosquito-borne
diseases outbreak although their linkage are inconsistent in
123
Mapping of mosquito-borne diseases in Kolkata Municipal Corporation using GIS and AHP based…
Table 4 Intensity of importance and respective risk value of selected MBDs causative factors
Factors Class Intensity of importance Degree of vulnerability Risk value
Land use land cover (LULC) Water bodies 1.00 Very high 5
Fallow land 0.33 Very low 1
Open space 0.25 Medium 3
Vegetation cover 1.00 Low 2
Built-up area 0.5 High 4
Normalised difference vegetation index (NDVI) -0.2355 to 0.0959 1.00 Very low 1
0.0960–0.1943 2.00 Low 2
0.1944–0.3072 3.00 Medium 3
0.3073–0.4311 4.00 High 4
0.4312–0.6934 5.00 Very high 5
Land surface temperature (LST) 25.90–28.95 1.00 Medium 3
28.96–30.74 3.00 High 4
30.75–32.04 6.00 Very high 5
32.05–33.39 3.00 Low 2
33.40–37.01 1.00 Very low 1
Normalised difference moisture index (NDMI) -0.3439 to 0.0117 1.00 Very low 1
0.0118–0.1044 2.00 Low 2
0.1045–0.2044 3.00 Medium 3
0.2045–0.3267 5.00 High 4
0.3268–0.6009 7.00 Very high 5
Normalised difference water index (NDWI) -0.5484 to -0.3177 1.00 Very low 1
-0.3178 to -0.2058 2.00 Low 2
-0.2059 to -0.1075 3.00 Medium 3
-0.1076 to 0.0248 5.00 High 4
0.0249–0.3166 6.00 Very high 5
Proportion of vegetation (PV) 0–0.1294 1.00 Very low 1
0.1295–0.2313 0.50 Low 2
0.2314–0.3607 0.33 Medium 3
0.3608–0.5333 0.25 High 4
0.5334–0.9999 0.17 Very high 5
Land elevation (LE) -28 to 4 1.00 Very high 5
5–7 0.50 High 4
8–10 0.33 Medium 3
11–13 0.25 Low 2
Above 13 0.17 Very low 1
Slope (SL) 0–1.3877 1.00 Very high 5
1.3878–2.7755 0.50 High 4
2.7756–5.2041 0.25 Medium 3
5.2042–12.6634 0.20 Low 2
12.6635–44.2535 0.17 Very low 1
Aspect (AS) Non-shadow area 1.00 Low 2
Shadow area 3.00 High 4
Proximity to water bodies (PWBs) B200 1.00 Very high 5
200–400 0.50 High 4
400–600 0.33 Medium 3
600–800 0.20 Low 2
C800 0.14 Very low 1
123
S. A. Ali, A. Ahmad
Table 5 Pair-wise comparison matrix along with measure of consistency of all selected factors
LULC Water bodies Fallow land Open space Vegetation cover Built-up area Weight
Water bodies 1 3 4 1 2 0.3176
Fallow land 1/3 1 1/2 1/4 1/5 0.0642
Open space 1/4 2 1 1/3 1/3 0.0913
Vegetation cover 1 4 3 1 1/2 0.2387
Built-up area 1/2 5 3 2 1 0.2880
CI = 0.0560,
CR = 0.0503
NDVI -0.2355 to 0.0959 0.0960–0.1943 0.1944–0.3072 0.3073–0.4311 0.4312–0.6934 Weight
-0.2355 to 0.0959 1 1/2 1/3 1/4 1/5 0.0623
0.0960–0.1943 2 1 1/2 1/3 1/4 0.0985
0.1944–0.3072 3 2 1 1/2 1/3 0.1610
0.3073–0.4311 4 3 2 1 1/2 0.2617
0.4312–0.6934 5 4 3 2 1 0.4162
CI = 0.0170,
CR = 0.0152
LST 25.90–28.95 28.96–30.74 30.75–32.04 32.05–33.39 33.40–37.01 Weight
25.90–28.95 1 1/3 1/6 1/3 1 0.0722
28.96–30.74 3 1 1/3 1/3 2 0.1502
30.75–32.04 6 3 1 2 3 0.4052
32.05–33.39 3 3 1/2 1 4 0.2859
33.40–37.01 1 1/2 1/3 1/4 1 0.08627
CI = 0.0420,
CR = 0.0375
NDMI -0.3439 to 0.0117 0.0118–0.1044 0.1045–0.2044 0.2045–0.3267 0.3268–0.6009 Weight
-0.3439 to 0.0117 1 1/2 1/3 1/5 1/7 0.0498
0.0118–0.1044 2 1 1/2 1/4 1/5 0.0807
0.1045–0.2044 3 2 1 1/3 1/4 0.1278
0.2045–0.3267 5 4 3 1 1/3 0.2639
0.3268–0.6009 7 5 4 3 1 0.4776
CI = 0.0351,
CR = 0.0313
NDWI -0.5484 to
-0.3177
-0.3178 to
-0.2058
-0.2059 to
-0.1075
-0.1076 to
0.0248
0.0249–0.3166 Weight
-0.5484 to -0.3177 1 1/2 1/3 1/5 1/6 0.0530
-0.3178 to -0.2058 2 1 1/3 1/4 1/5 0.0776
-0.2059 to -0.1075 3 3 1 1/2 1/4 0.1516
-0.1076 to 0.0248 5 4 2 1 1/3 0.2448
0.0249–0.3166 6 5 4 3 1 0.4727
CI = 0.0427,
CR = 0.0381
Proportion of
vegetation
0–0.1294 0.1295–0.2313 0.2314–0.3607 0.3608–0.5333 0.5334–0.9999 Weight
0–0.1294 1 2 3 4 6 0.4036
0.1295–0.2313 1/2 1 3 4 5 0.2964
0.2314–0.3607 1/3 1/3 1 3 4 0.1660
0.3608–0.5333 1/4 1/4 1/3 1 2 0.0824
0.5334–0.9999 1/6 1/5 1/4 1/2 1 0.0514
CI = 0.0452,
CR = 0.0404
Land elevation -28 to 4 5–7 8–10 11–13 Above 13 Weight
123
Mapping of mosquito-borne diseases in Kolkata Municipal Corporation using GIS and AHP based…
Table 5 continued
LULC Water bodies Fallow land Open space Vegetation cover Built-up area Weight
-28 to 4 1 2 3 4 6 0.3979
5–7 1/2 1 4 5 6 0.3360
8–10 1/3 1/4 1 2 3 0.1317
11–13 1/4 1/5 0.5 1 2 0.0830
Above 13 1/6 1/6 1/3 1/2 1 0.0511
CI = 0.0389,
CR = 0.0347
Slope 0–1.3877 1.3878–2.7755 2.7756–5.2041 5.2042–12.6634 12.6635–44.2535 Weight
0–1.3877 1 2 4 5 6 0.4345
1.3878–2.7755 1/2 1 3 4 5 0.2868
2.7756–5.2041 1/4 1/3 1 2 4 0.1385
5.2042–12.6634 1/5 1/4 1/2 1 3 0.0917
12.6635–44.2535 1/6 1/5 1/4 1/3 1 0.0482
CI = 0.0486,
CR = 0.0434
Aspect Non-shadow area Shadow area Weight
Non-shadow area 1 1/3 0.25
Shadow area 3 1 0.75
CI = 0.00, CR = 0.00
Proximity to water
bodies
B200 200–400 400–600 600–800 C800 Weight
B200 1 2 3 5 7 0.4262
200–400 1/2 1 2 4 6 0.2712
400–600 1/3 1/2 1 3 5 0.1771
600–800 1/5 1/4 1/3 1 3 0.0836
C800 1/7 1/6 1/5 1/3 1 0.0416
CI = 0.0348,
CR = 0.0311
CR value 0.1 or \0.1 is considered for acceptance
Table 6 Weight value of each MBDs causative factor for final overlay
LULC NDVI LST NDWI NDMI PV LE SL AS PWBs PCM based Weight Ax kmax Weight (%)
LULC 1 1 1/3 1/4 1/4 1 2 2 2 1/4 0.06297 0.6630 10.5296 6.2969
NDVI 1 1 1/2 1/3 1/3 1 2 2 2 1/3 0.0689 0.7262 10.5322 6.8951
LST 3 2 1 1/2 1 2 3 3 1 1/3 0.1146 1.2306 10.7349 11.4641
NDWI 4 3 2 1 1 1 2 3 3 1/3 0.1394 1.5229 10.9191 13.9471
NDMI 4 3 1 1 1 2 4 4 4 1/2 0.1593 1.7086 10.7214 15.9365
PV 1 1 1/2 1 1/2 1 2 3 3 1/3 0.0888 0.9444 10.6290 8.8856
LE 1/2 1/2 1/3 1/2 1/4 1/2 1 1/2 1/2 1/3 0.0401 0.4224 10.5205 4.0154
SL 1/2 1/2 1/3 1/3 1/4 1/3 2 1 1 0.25 0.0435 0.4525 10.4020 4.3504
AS 1/2 1/2 1 1/3 1/4 1/3 2 1 1 1/3 0.0522 0.5484 10.4890 5.2292
PWBs 4 3 3 3 2 3 3 4 3 1 0.2297 2.4875 10.8250 22.9790
CI = 0.0700, CR = 0.0470
CR value 0.1 or \0.1 is considered for acceptance
LULC land use lands cover, NDVI normalised difference vegetation index, LST land surface temperature, NDWI normalised difference water
index, NDMI normalised difference moisture index, PV proportion of vegetation, LE land elevation, SL slope, AS aspect, PWBs proximity to
water bodies
123
S. A. Ali, A. Ahmad
some cases [6]. Comfortable temperature and presence of
humidity have direct and indirect influences on the heri-
table aspects of mosquitoes and on maturation periods
within mosquitoes [5,52]. Different type of land cover
such as water logged areas; water bodies, water stagnant
agricultural land, and congested human settlement have
been recognized as vulnerable areas for vector-borne dis-
eases [53]. The breeding habits of Aedes aegypti (Aedes
albopictus) mosquitoes are found in congested urban and
peri-urban environment which provides most suitable sites
for breeding [14]. Therefore, it can be suggested that with
the increasing population many environmental conditions
will alter including housing types, mode of livings, space
for land use, local neighbourhood etc. which may directly
or indirectly related DF risk and likely increase the risk of
DF in those areas which already an endemic of this disease.
Similar study could be carried out for many DF endemic
areas looking towards same methodology. Similar studies
would help researcher in choosing different local envi-
ronmental, climatological and socio-economic parameters
as decision making criteria.
For spatial analysis, it is a very crucial task to judge the
result with actual background. As per KMC report in last
year from January to October more than 600 dengue cases
reported in KMC out of which 6 were death. Moreover
2000 malaria cases were reported in different KMC malaria
clinics. So, a field survey was carried out in June and July
of 2018 in different wards of each 16 Borough of KMC.
Looking towards surrounding environment and respon-
dent’s response on structured questions, different GPS
locations were recorded. These locations were further
verified in Google Earth open software and export as point
locations. After sample categorization, many areas in
southern, south-eastern and western part come under very
low prevalence of MBDs, some areas found with no cases
of such diseases, whereas northern and central parts come
under high to moderate susceptible zone to MBDs. Thus,
some areas with very high and high prevalence of such
diseases were plotted to merge with final output of MBDs
(Fig. 14). The result shows that these reported areas are
more uniform with final layer and constantly merged with
very high to high zones as derived from spatial analysis of
multi-criteria based decision factors.
4 Conclusions
The present study highlights the capability of geographic
information system and application of analytic hierarchy
process in identifying vulnerable zones of mosquito-borne
diseases. Geographic information system offers data inte-
gration and AHP provides sound decision towards specific
Fig. 13 Major hotspot of
mosquito-borne diseases in
Kolkata Municipal Corporation
123
Mapping of mosquito-borne diseases in Kolkata Municipal Corporation using GIS and AHP based…
incidence to its users. Recently, the application of GIS in
public health assessment has become popular due to its
spatial analysis, integration and proximate analysis. Multi-
criteria approach in combination with GIS for mapping of
mosquito-borne diseases can offer construct results with
spatial relationship. The consistency measurement through
AHP also verifies the whole process of decision making
and helps in identifying the weighted criteria. The results
of present study explained that mosquito-borne diseases are
greatly influenced by different factors and environmental
conditions. These all factors are interrelated and play a
vigorous role in the occurrence of these mosquito-borne
diseases either directly or indirectly. For instance, surface
temperature, moisture content, water index and vegetation
index have interrelationship among each other and mis-
balance in any one of these factors can harm to mosquito’s
growth. Thus, suitable ranges of all selected factors were
considered and taken as high vulnerable zones for mos-
quito-borne diseases. With the help of satellite data and
integration of various inter-connected factors in GIS sup-
ports in allocating areas which required more surveillance
and managemental steps to prevent spread of such diseases.
The utilization of geospatial data in mapping and analysis
of diseases assessment cannot be ignored because it is low
cost or free of cost, less time consuming and offers accu-
rate result.
In order to reduce the impacts of MBDs, it is essential to
destruct either the means of mosquito parasite or human
host. Suitable breeding sites detection and destruction is
considered as best method in achieving this target. So,
present research efforts on some of the associated and
inter-connected factors that responsible for mosquito’s
growth were considered and GIS based spatial analysis was
used for susceptible zonation of mosquito-borne diseases.
The result shows many MBDs prone areas in KMC. The
results also depicts that proximity to water bodies, presence
of moisture in air, high water index and congested built
areas have high possibility to mosquito-borne diseases.
Like present study with the application of geospatial
technology can be applied in different MBDs assessment
including malaria, dengue, chikungunya, yellow fever,
filariasis, Japanese encephalitis etc. in different geograph-
ical locations of the earth which help in detecting vulner-
able areas and better surveillance.
Acknowledgements We thankfully acknowledge the anonymous
reviewers for their valuable time, productive comments and sugges-
tions for improving the overall quality of the manuscript.
References
1. Ahmad, F., Goparaju, L., & Qayum, A. (2017). Studying malaria
epidemic for vulnerability zones: Multi-criteria approach of
geospatial tools. Journal of Geoscience and Environment Pro-
tection, 5, 30–53. https://doi.org/10.4236/gep.2017.55003.
2. Sharma, S. N., Ghosh, D., Srivastava, P. K., Sonal, G. S., &
Dhariwal, A. C. (2014). Vector borne diseases in Kolkata
Municipal Corporation KMC: Achievements and challenges.
Journal of Communicative Disease, 46(2), 68–76.
3. Wondim, Y. K., Alemayehu, E. B., & Abebe, W. B. (2017).
Malaria hazard and risk mapping using GIS based spatial mul-
ticriteria evaluation technique (SMCET) in Tekeze Basin
Development Corridor, Amhara Region, Ethiopia. Journal of
Environment and Earth Science, 7(5), 76–87.
4. Scott, F. L. (2007). Anti-protozoals. Department of Pharmaceu-
tical Sciences, College of Pharmacy, Oklahoma State University,
Fig. 14 Maps showing the accuracy of risk zonation amajor mosquito-borne diseases affected areas in KMC bGIS and AHP based
identification of MBDs hotspot zones
123
S. A. Ali, A. Ahmad
Weatherford. http://faculty.swosu.edu/scott.long/phcl/antiprot.
htm.2007. Accessed 13 Sept 2018.
5. Morin, C. W., Comrie, A. C., & Ernst, K. (2013). Climate and
dengue transmission: Evidence and implications. Environmental
Health Perspectives, 121, 1264–1272.
6. Hii, Y. L., Zaki, R. A., Aghamohammadi, N., & Rocklov, J.
(2016). Research on climate and dengue in Malaysia: A sys-
tematic review. Current Environmental Health Reports, 3(1),
81–90.
7. Rattanarithikul, R., et al. (1995). Larval habitats of malaria
vectors and other Anopheles mosquitoes around a transmission
focus in Northwestern Thailand. Journal of the American Mos-
quito Control Association, 11, 428–433.
8. Mushinzimana, E., Munga, S., Minakawa, N., Li, L., Feng, C. C.,
Bian, L., et al. (2006). Landscape determinants and remote
sensing of anopheline mosquito larval habitats in the western
Kenya highlands. Malaria Journal, 5, 13. https://doi.org/10.1186/
1475-2875-5-13.
9. Ayele, D. G., Temesgen, T. Z., & Henry, G. M. (2012). Preva-
lence and risk factor of malaria in Ethiopia. Malaria Journal, 11,
195.
10. WHO. (2012). World Malaria Report. Geneva: World Health
Organization. www.who.int/malaria/publications/world_malaria_
report_2012/en/. Accessed 7 Sept 2018.
11. WHO. (2017). World Malaria Report. Geneva: World Health
Organization. www.who.int/malaria/publications/world-malaria-
report-2017/report/en/. Accessed 7 Sept 2018.
12. Ali, S. A., & Ahmad, A. (2018). Using analytic hierarchy process
with GIS for dengue risk mapping in Kolkata Municipal Corpo-
ration, West Bengal, India. Spatial Information Research, 26,
449. https://doi.org/10.1007/s41324-018-0187-x.
13. Nazri, C. D., Ahmad, A. H., Latif, Z. A., Ismail, R., & Pradhan,
B. (2012). Coupling of remote sensing data and environmental
related parameters for dengue transmission risk assessment in
Subang Jaya, Malaysia. Geocarto International, 28(3), 1–15.
https://doi.org/10.1080/10106049.2012.696726.
14. Sarfraz, M. S., Tripathi, N. K., Faruque, F. S., Bajwa, U. I.,
Kitamoto, A., & Souris, M. (2014). Mapping urban and peri-
urban breeding habitats of Aedes mosquitoes using a fuzzy ana-
lytical hierarchical process based on climatic and physical
parameters. Geospatial Health, 8(3), S685–S697.
15. Tu, C., Fang, Y., Huang, Z., & Tan, R. (2014). Application of the
analytic hierarchy process to a risk assessment of emerging
infectious diseases in Shaoxing city in southern China. Japanese
Journal of Infectious Diseases, 67(6), 417–422.
16. Dongus, S., et al. (2009). Urban agriculture and Anopheles
habitats in Dar es Salaam, Tanzania. Geospatial Health, 3,
189–210. https://doi.org/10.4081/gh.2009.220.
17. Delgado, L., Camardiel, A., Aguilar, V., Martinez, N., Codova,
K., & Ramos, S. (2011). Geospatial tools for the identification of
a malaria corridor in Estado Sucre, a Venezuelan North Eastern
State. Geospatial Health, 5, 169–176. https://doi.org/10.4081/gh.
2011.168.
18. Sharma, V. P., et al. (1996). Study on feasibility of delivering
mosquitogenic conditions in and around Delhi using IRS data.
Indian Journal of Malariology, 33, 107–125.
19. Robert, V., Macintyre, K., Keating, J., Trape, J. F., Duchemin, J.
B., & Warren, M. (2003). Malaria transmission in urban sub-
Saharan Africa. The American Journal of Tropical Medicine and
Hygieneg, 68, 169–176.
20. Palaniyandi, M. (2013). GIS mapping of vector breeding habitats.
Geospatial World Weekly, 9, 1–4.
21. Wood, B. L., Beck, L. R., Washino, R. K., Hibbard, K. A., &
Salute, J. S. (1992). Estimating high mosquito-producing rice
fields using spectral and spatial data. International Journal of
Remote Sensing, 13, 2813–2826. https://doi.org/10.1080/
01431169208904083.
22. Qayum, A., Arya, R., Kumar, P., & Lynn, A. M. (2015). Socio-
economic, epidemiological and geographic features based on GIS
integrated mapping to identify malarial hotspots. Malaria Jour-
nal, 14, 192. https://doi.org/10.1186/s12936-015-0685-4.
23. Nakhapakorn, K., & Tripathi, N. K. (2005). An information value
based analysis of physical and climatic factors affecting dengue
fever and dengue haemorrhagic fever incidence. International
Journal of Health Geographics, 4, 1–13.
24. Rochon, G. L., Quansah, J. E., Fall, S., Araya, B., Biehl, L. L.,
Thiam, T., et al. (2010). Remote sensing, public health and dis-
aster mitigation. Geospatial Technologies in Environmental
Management, 3, 187–209.
25. Hongoh, V., Anne, H., Gatewood Hoen, A., Ce
´cile, W., Jean-
Philippe, B., et al. (2011). Spatially explicit multi-criteria deci-
sion analysis for managing vector-borne diseases. International
Journal of Health Geographics, 10, 70. http://www.ij-healthgeo
graphics.com/content/10/1/70. Accessed 17 Sept 2018.
26. Khormi, H. M., & Kumar, L. (2011). Modelling dengue fever risk
based on socioeconomic parameters, nationality and age groups:
GIS and remote sensing based case study. Science of the Total
Environment, 409(22), 4713–4719.
27. Yadav, K., Nath, M. J., Talukdar, P. K., Saikia, P. K., Baruah, I.,
& Singh, L. (2012). Malaria risk areas of Udalguri District of
Assam, India: A GIS-based study. International Journal of
Geographical Information Science, 26, 123–131. https://doi.org/
10.1080/13658816.2011.576678.
28. Walker, M., et al. (2013). Temporal and micro spatial hetero-
geneity in the distribution of Anopheles vectors of malaria along
the Kenyan Coast. Parasites and Vectors, 6, 311. https://doi.org/
10.1186/1756-3305-6-311.
29. Nazri, C. D., Ahmad, A. H., Latif, Z. A., & Ismail, R. (2016).
Application of GIS-based analytical hierarchy process as a tool
for dengue risk assessment. Asian Pacific Journal of Tropical
Disease, 6(12), 930–937.
30. Sheela, A. M., Ghermandi, A., Vineetha, P., Sheeja, R. V., Justus,
J., & Ajayakrishna, K. (2017). Assessment of relation of land use
characteristics with vector-borne diseases in tropical areas. Land
Use Policy, 63, 369–380. https://doi.org/10.1016/j.landusepol.
2017.01.047.
31. Norris, D. E. (2004). Mosquito-borne diseases as a consequence
of land use change. EcoHealth, 1, 19–24. https://doi.org/10.1007/
s10393-004-0008-7.
32. Kleinschmidt, I., Bagayoko, M., Clarke, G. P. Y., Craig, M., &
Sueur, D. L. (2000). A spatial statistical approach to malaria
mapping. International Journal of Epidemiology, 29, 355–361.
33. Brown, H., Diuk-Wasser, M., Andreadis, T., & Fish, D. (2008).
Remotely-sensed vegetation indices identify mosquito clusters of
West Nile Virus vectors in an urban landscape in the North-
eastern United States. Vector-Borne and Zoonotic Diseases, 8,
197–206.
34. Lourenc¸o, P. M., Sousa, C. A., Seixas, J., Lopes, P., Novo, M. T.,
& Almeida, A. P. G. (2011). Anopheles atroparvus density
modeling using MODIS NDVI in a former malarious area in
Portugal. Journal of Vector Ecology, 36, 279–291.
35. Gemperli, A., Sogoba, N., Fondjo, E., Mabaso, M., Bagayoko,
M., Brie
¨t, O. J. T., et al. (2006). Mapping malaria transmission in
West and Central Africa. Tropical Medicine & International
Health, 11(7), 10321046.
36. Rouse, J. W., Haas, R. H., Schell, J. A., & Deering, D. W. (1974).
Monitoring vegetation systems in the Great Plains with ERTS. In
Proceedings of the third earth resources technology satellite1
symposium (pp. 301–317). Greenbelt: NASA SP351.
123
Mapping of mosquito-borne diseases in Kolkata Municipal Corporation using GIS and AHP based…
37. Sandru, M. I. V. (2014). Promoting spatial data synthesis for
vector-borne disease assessment in Romania. Romanian Review
of Regional Studies, X(2), 75–86.
38. Yu, X., Guo, X., & Wu, Z. (2014). Land surface temperature
retrieval from Landsat 8 TIRS—Comparison between radiative
transfer equation-based method, split window algorithm and
single channel method. Remote Sensing, 6, 9829–9852. https://
doi.org/10.3390/rs6109829.
39. Epstein, P. R. (2001). Climate change and emerging infectious
diseases. Microbes and Infection, 3, 747–754.
40. McFeeters, S. K. (1996). The use of the normalized difference
water index (NDWI) in the delineation of open water features.
International Journal of Remote Sensing, 17, 1425–1432.
41. Xu, H. (2006). Modification of normalised difference water index
(NDWI) to enhance open water features in remotely sensed
imagery. International Journal of Remote Sensing, 27,
3025–3033.
42. Atieli, H., et al. (2011). Topography as a modifier of breeding
habitats and concurrent vulnerability to malaria risk in the
western Kenya highlands. Parasites and Vectors, 4, 241. https://
doi.org/10.1186/1756-3305-4-241.
43. Sergo, P. (2007). Dengue fever warming up to human habits.
Retrieved from http://www.scienceline.org. Accessed 22 Sept
2018
44. Ahmed, A. (2014). GIS and remote sensing for malaria risk
mapping, Ethiopia. International Archives of the Photogramme-
try, Remote Sensing and Spatial Information Sciences, XL-8,
155–161.
45. Mu, E., & Rojas, M. P. (2017). Understanding the analytical
hierarchy process. In Springer briefs in operational research.
https://doi.org/10.1007/978-3-319-33861-3_2.
46. Saaty, T. L. (1990). How to make a decision: The analytic
hierarchy process. European Journal of Operational Research,
48, 9–26.
47. Saaty, T. L., & Vargas, L. G. (2001). Models, methods, concepts
and applications of the analytic hierarchy process. Norwell:
Kluwer Academic Publishers. https://doi.org/10.1007/978-1-
4615-1665-1.
48. Saaty, T. L. (2012). Decision making for leaders: The analytic
hierarchy process for decisions in a complex world. Pittsburgh:
RWS Publication.
49. Jeefoo, P., Tripathi, N. K., & Hara, S. (2008). Analytical hier-
archy process modeling for malaria risk zonation in Kan-
chanaburi, Thailand. In International symposium on
geoinformatics for spatial infrastructure development in earth
and allied sciences. Accessed from www.gisws.media.osaka-cu.
ac.jp/gisideas08/viewpaper.php?id=284. Accessed 23 Sept 2018.
50. Bhatt, B., & Joshi, J. P. (2014). Analytical hierarchy process
modeling for malaria risk zones in Vadodara district, Gujarat. The
International Archives of the Photogrammetry, Remote Sensing
and Spatial Information Sciences, XL-8. Accessed from https://
www.int-arch-photogramm-remote-sens-spatial-inf-sci.net/XL-8/
171/2014/isprsarchives-XL-8171201. Accessed 24 Sept 2018.
51. Saaty, T. L., & Vargas, L. G. (2012). Models, methods, concepts
and applications of the analytic hierarchy process. Dordrecht:
Kluwer.
52. Barrera, R., Amador, M., & MacKay, A. J. (2011). Population
dynamics of aedes aegypti and dengue as influenced by weather
and human behaviour in San Juan, Puerto Rico. PLoS Neglected
Tropical Diseases, 5, e1378. https://doi.org/10.1371/journal.pntd.
0001378.
53. Cheong, Y. L., Leitao, P. J., & Lakes, T. (2014). Assessment of
land use factors associated with dengue cases in Malaysia using
boosted regression trees. Spatial and Spatio-temporal Epidemi-
ology, 10, 75–84.
123
S. A. Ali, A. Ahmad
