Unleash Your Potential - Namagunga Girls Coding Club
Wildfire risk map nrsc2
1. Wild fire risk Mapping in Eastern Mongolia
Using Remote Sensing and Multi Criteria
Analysis. (SMCA).
http://www.icc.mn/
http://geomedeelel.b
logspot.com/
Elbegjargal NASANBAT / chief Technologist of Remote sensing.
n_elbegjargal@yahoo.com
Есдүгээр сарын "Гео-уулзалт"
2. About abstract
The study selected variables for wildfire risk assessment using a combination of data
collection and evaluation decision support (Spatial Multi-Criteria Analysis) methods;
Social Economic, Climate, Geographic Information Systems, Remote Sensing and
statistical yearbook and field validation data.
The data evaluation resulted divided by main three group factors Environmental, Social
Economic factor, Climate factor and Fire information factor into eleven input variables,
which were classified into five categories (risk levels) important criteria and ranks.
3. Contents
Study area
Map of study area with burn area (2000-2011)
Wildfire situation in study area
Causes of Wildfire
Climate Condition
Main objectives
Data to be used
Available data
Pre-processing step
Proposed Methodology
Weighted Straight Ranking
Expected result Outcomes
Validation
4. Study area
Study area: Eastern region of Mongolia, covering 3
provinces (shown with red boundaries)
Central Location: latitude 48.055206 Longitude
113.402393
Area: 289,323 km2
Population: 199,468
Dry season: April to June and August to October
(high possibility of wildfire occurrence )
Burn area are shown with
polygons of different color which
range from 2000 to 2014
5. Wildfire situation in study area
y = 3.6791x + 18.692
0
20
40
60
80
100
120
140
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
Numberoffires
Number of Wildfires occurred in Eastern part of Mongolia,
2000 - 2013
6. Causes of Wildfire
Human behavior
Tourist
Smoke
Cars and Tractor
From ash fire
From object fire
Uncontrolled
children
Natural behavior
Thunder and lightning
coniferous forest
Forest biomass
Dry grass
Underground fire
From burnt area
Less rain
High temperature
High speed wind
Other
Wildfire from
neighboring country
expands to the study
area
Waste
Dew drops
7. 5
8
32
75
1
23
4
21
Percentage of Causes wildfire classification
2010 -2011
Smoke
From ash fire
Tractor and Cars
From neighboring
country
From object fire
From burnt area and
undergraund fire
Thunder and lightning
Other
Unknown
8. Main objectives
The main goal of this study is to identify wildfire risk area, using
weighting tool named Spatial Multi-Criteria Evaluation Analysis
(SMCEA) based on Remote sensing satellite image and thematic
maps from available spatial data.
The specific objectives are as followings:
1. To identify wildfire risk area and non-risk area of wildfire
in the study area
2. To compare the output wildfire risk map with ground truth
information related to burned forest and steppe for validation
9. Data to be used
1. Available Data
Category ID Name
GIS Vector
1 Province boundary
2 Sub-province boundary
3 Sub-province center
4 Road network
GIS Thematic maps
5 River, springs
6 Lakes
7 Well
8 Tourist camps
9 Fire Guard
Climate data
10 Air temperature, 2000-2015
11 Rainfall, 2000-2015
12 Soil moisture, 2000-2015
13 Wind speed and direction, 2000-2015
Social Economic
data
14 Number of Population, 2000-2010
15 Number of Livestock, 2000-2010
Remote Sensing
data
16 Aqua terra/ MODIS, 2000-2015
17 MODIS hotspot, 2000-2015
18 Burned area map, 2000-2015
19 Land cover map, 2010
20
Global Digital Elevation Model (DEM)
30m
10. Geo-Database for wildfire risk map
1. Available Data
The available data were
arranged to Geo-
database in ArcGIS
11.
12. Methodology
In this study, we used Multi-Criteria Analysis (MCA) as a technique for making
complex decision as Mentioned (Alejandro and Jorge, 2003). In order to
identify the high risk or optimal location area for a particular subject, the
variables can be separately divided into four main groups of factors which is
Environmental factors, Climate factors, Fire information factors and social
economic factors. A flow chart can be divided into main groups from wildfire
risk the model.
13. Weighted Straight Ranking methods
Then suitable weights of factors were assigned using the Multi Criteria Analysis (MCA) which is a technique used
to assist for making complex decision as mentioned (Alejandro and Jorge, 2003). The data was weighted and
ranked based on the Malczewski formula (Malczewki, J.C.1999).
Rank sum weights are calculated according to the following equations 1 and 2 as below.
No Variable of Criteria Straight Rank
Numerator
(n-ri+1)
Normalized
Weights
0-100 %
1 Frequency burned area 1 11 0.10 10
2 Fire hotspot density 1 11 0.10 10
3 Land cover class 2 10 0.09 9
4 Agricultural land 2 10 0.09 9
5 Air temperature 3 9 0.08 8
6 Rainfall 3 9 0.08 8
7 Soil moisture 4 8 0.08 8
8 Altitude of topography 3 9 0.08 8
9 Aspect of topography 3 9 0.08 8
10 Slope of topography 2 10 0.09 9
11 Settlements and infrastructure 2 10 0.09 9
Sum 26 106 1 100
14. a) Environmental factors
Criteria Weight Weight %
Land cover 0.1 10
Slope 0.09 9
Aspect 0.09 9
Water bodies 0.1 10
Total 0.38 38
b) Climate factors
Criteria Weight %
Air temperature 0.11 11
Rainfall 0.11 11
Wind speed 0.09 9
Total 0.31 31
c) Fire information factors
Criteria Weight %
Fire frequency 0.1 10
Distance from Settlement 0.1 10
Distance from Road 0.11 11
Total 0.31 31
The purpose of the criteria weighting is to express the importance of each criterion relative to other criteria.
The more important criterion has the greater weight in the overall evaluation. The weight was assigned to
each individual evaluation criteria (factors) making their sum equal to 100%.
Weights and factors
19. y = -0.2659x + 39.799
R= 0.7590
0
5
10
15
20
25
30
35
40
45
0 20 40 60 80 100 120 140
precipitation(mm)
Number of fire
What is correlation between wildfire and climate factors?
y = -0.9684x + 34.753 y = 3.6791x + 18.692
0
5
10
15
20
25
30
35
40
45
0
20
40
60
80
100
120
140
Averageofmay/mm/
Numberofwildfire
The relationship between number of wildfire occurrence and meteorological precipitation
average of may 2000-2013
Хур тунадас
Number of
wildfire
Linear (Хур
тунадас)
Linear (Number
of wildfire)
20. 0
2
4
6
8
10
12
14
16
18
20
0
20
40
60
80
100
120
140
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
Airtemperature,averageofmay(C)
Агаарын температур
Ой хээрийн түймрийн тоо
The relationship between number of wildfire occurrence and meteorological air temperature average
of may 2000-2013
y = -0.029x + 16.701
R= 0.5551
0
2
4
6
8
10
12
14
16
18
20
0 50 100 150
Airtemperature,averageofmay(C)
Number of fire
Data type Intensity
value
r-
value
Mean
value
RMS
Number of fire y = 3.6791x
+ 18.69
46.285 8.219
Precipitation y = -0.9684x
+ 34.75
0.75 27.490 2.879
Air temperature y = 0.03x +
15.13
0.55 15.356 0.430
26. Conclusion
• This map could be help Forest Department officials prevent reduction fire risk
activities within the forest and take proper action when fire breaks out.
• This map could help the decision maker to planning the fire prevention,
suitability place to build fire monitoring prepare the fire fighting team.
27. Thank you for your
attention
n_elbegjargal@yahoo.com
www.icc.mn
Editor's Notes
ACRS oroltsokhoos umnu ene table-iig update khiikh kheregtei