APPLICATIONS OF GIS, REMOTE SENSING AND

PUBLIC DOMAIN DATA IN MAPPING AND

ANALYZING THE VELDFIRE INCIDENTS FOR
MAKONI AND CHIMANIMANI DISTRICTS FOR
2005 AND 2006 VELDFIRE SEASON

HONOURS DEGREE THESIS SUBMITTED TO THE MIDLANDS STATE

UNIVERSITY IN FULFILLMENT OF THE HONOURS DEGREE
BY
CHARLES GUMISAI CHIGURAH
December 2007

SUPERVISED BY: Mr. JERIE

1
Table of Contents
Dedication…………………………………………………………………. …………… i
Acknowledgements……………………………………………………….………. ……ii
Abstract……………………………………………………………………. ……………iii
List of figures……………………………………………………………………………iv
List of tables……………………………………………………………………………..v

CHAPTER 1............................................................................................... 4

1.1 PROBLEM STATEMENT.................................................................................................. 7

LAND CATEGORY ..................................................................................... 7

1.2 JUSTIFICATION. ............................................................................................................... 8

1.3 AIMS AND RESEARCH OBJECTIVES. .......................................................................... 9
1.4 RESEARCH QUESTIONS ................................................................................................. 9
1.5 STUDY AREA CHARACTERISTICS. ........................................................................... 10
1.6 DEFINITION OF KEY CONCEPTS. ....................................................................................... 14
1.7 LIMITATIONS....................................................................................................................... 14
1.8 SUMMARY. ........................................................................................................................... 14

CHAPTER 2............................................................................................. 15

2.1. INTRODUCTION ................................................................................................................. 15

2.2 .............................................................................. KEY CONCEPTS.

............................................................................................................... 15

2.1.1 REMOTE SENSING. ............................................................................................................. 15

2.1.2 FIRE DETECTION AND MONITORING. ............................................................................. 17
2.1.3 LIMITATIONS .................................................................................................................... 18
2.1.4 WILDFIRE TEMPERATURE RETRIEVAL .......................................................................... 19
2.2 GEOGRAPHICAL INFORMATION SYSTEMS ........................................................................ 19
2.2.1 GIS APPLICATIONS. ......................................................................................................... 20
2.2.2 OTHER SENSORS CURRENTLY BEING USED TO PROVIDE INFORMATION ABOUT
FOREST FIRES............................................................................................................................... 21
2.3 CAUSES OF FIRES ................................................................................................................ 22
2.4 DIFFERENT TYPES OF VELDFIRES ..................................................................................... 23
2.5 FACTORS DETERMINING INCIDENTS AND EFFECTS OF VELDFIRES ................................ 24
2.5.1 VEGETATION TYPE. ......................................................................................................... 24
2.5.2 SOCIO-ECONOMIC AND DEMOGRAPHIC TRENDS .......................................................... 25

2
2.5.3 METEOROLOGY ............................................................................................................... 26
2.5.4 TOPOGRAPHY. .................................................................................................................. 30
2.6 IMPACTS OF VELDFIRES ..................................................................................................... 30
2.7 CONTRIBUTION BY THE COMMUNITY IN FIRE MANAGEMENT AND FIRE INCIDENTS
MAPPING. ...................................................................................................................................... 35
2.8 SUMMARY ............................................................................................................................ 37

CHAPTER 3............................................................................................. 38

RESEARCH DESIGN AND METHODOLOGY ..................................................................... 38

3.1 RESEARCH DESIGN ............................................................................................................. 38
3.2 EXPLORATORY AND DESCRIPTIVE RESEARCH. ............................................................... 39
3.3 RESEARCH INSTRUMENTS. ........................................................................................ 40
3.4 SUMMARY ............................................................................................................................ 45

CHAPTER 4............................................................................................. 46

RESULTS AND DISCUSSIONS. ............................................................................................... 46

4.1 RESULTS ........................................................................................................................... 47
4.1.1 SPATIAL LOCATION. .................................................................................................. 47
4.1.2 NUMBER OF FIRE INCIDENTS FOR THE 2005/2006 FIRE SEASONS AND THE MONTHLY
SPATIAL DISTRIBUTION. .............................................................................................................. 54
4.1.3 THE RELATIONSHIP BETWEEN FIRE INCIDENCES AND CROP USE/CROP USE INTENSITY
(CUI) AND VEGETATION COVER AND TYPE. ............................................................................... 63
4.1.4 HOW EFFECTIVE WAS THE SYSTEM OF DETECTING AND REPORTING FIRES? ............. 66
4.1.5 FIRE INCIDENT RECURRENCE INTERVAL ...................................................................... 68
4.2 OFFICERS’ BACKGROUND KNOWLEDGE ON REMOTE SENSING
INSTRUMENTS AND THEIR APPLICATIONS IN VELDFIRE MONITORING. ............ 70
4.2.2 PREFERRED METHODS USED IN COLLECTING STATISTICS ON FIRE INCIDENTS ........ 72
4.2.3 CONSTRAINTS FACED IN COLLECTING DATA AND STATISTICS ONVELDFIRES. ......... 73
4.3 DISCUSSION. .................................................................................................................... 75
4.4 SUMMARY. ........................................................................................................................... 77

CHAPTER 5............................................................................................. 79

CONCLUSIONS AND RECOMMENDATIONS. ................................................................................. 79

5.1 INTRODUCTION. .................................................................................................................. 79
5.2 CONCLUSIONS. .................................................................................................................... 79
5.3 RECOMMENDATIONS ................................................................................................... 81

BIBLIOGRAPHY ...................................................................................... 83

APPENDICE A: QUESTIONNAIRE DESIGNED FOR THE EMPLOYEES.......................................... 87

3
 Chapter 1

1.0 INTRODUCTION AND BACKGROUND.

Sub-Saharan Africa experiences extensive biomass degradation with

approximately 1.7 million km2 (17 % of the land area) burns annually south of the

equator (Dube 2003). South of the equator is host to tropical savannas that are

characterized by a unique co-existence of tree and herbaceous species. Terrestrial

savanna ecosystems, though shifting mosaics, are perceived to be fire tolerant

because of their inherent ability to regenerate through coppicing and seeds (Zolho

2005). Extensive biomass burning is thus a core component of the ecological

disturbance regimes affecting these savanna ecosystems.

In Southern Africa fire is perceived ambiguously, there is a tension between fire

as a crucial process in certain ecosystems and fire events as a threat to

infrastructure and life. In both cases though, spatio-temporal awareness of fire

likelihood, fire occurrence and fire behaviour is key to appropriate intervention.

Fires occur due to both natural and anthropogenic causes, primarily lightning and

land management.

Systematic monitoring of burned areas is needed by the global change research

community as an input to biogeochemical models and is required to understand

and model how far regimes are likely to change as a function of climate,

population dynamics and land use changes. There is no adequate data on fire

4
incidents, size, distribution or trends in fire statistics or areas burned annually in

Southern Africa. Local data exist for national parks, forests and conservation

areas, but are not representative of the region as a whole because these lands are

largely protected from the influence of people and are subjected to specific fire

management policies.

Most of the fires are started by people both to modify their environment for a

variety of reasons related to land use and to facilitate harvesting of natural

resources. Fires are a source of greenhouse gases that contribute to changes in

atmosphere chemistry and potentially to global climate change, an issue of

increasing international concern. At a regional level, wild fires, together with

other sources of biomass burning for example domestic hearth fires using wood

fuel, contribute to the regional aerosol load, which in turn affect energy balance

and hydrology at sub continental scale.

In view of the centrality of fire as an ecological disturbance agent it is imperative

therefore to spatially investigate, understand and map the fire occurrence and

recurrence, awareness of fire likelihood, and its relationship with land use – this is

what is referred to as veldfire dynamics. The Global Observation of Forest

Cover/Global Observation of Land cover Dynamics project notes that “Global

change researchers, natural resource managers and policy decision-makers require

better information on the causes, location, extent and impacts of fire and the

sources, volumes and impacts of the associated fire emissions.” (Frost 1999). The

5
risk to human livelihood posed by fires and the large economic cost of extreme

fire events makes it a leading disaster-management

The study of veldfire dynamics using remote sensing and public domain data,

which is data or information obtained both from the local community and the

district officers, is therefore central to the determination of the economic

turnaround of the development agenda of Zimbabwe. The Environmental

Management Agency of Zimbabwe, the focal point of all environmental issues in

Zimbabwe, has of late taken a keen interest in fire prevention and management.

Efforts of the Environmental Management Agency of Zimbabwe in 2006 focused

on fire prevention, but because of financial and resource constraints the work of

fire prevention and detection was severely constrained and ineffective.

It is imperative therefore to explore other avenues that will assist in spatial

mapping and detection of bushfires at least cost and this call for the use of public

domain data, remote sensing data such as Moderate-Resolution Imaging

Spectrometer (MODIS) and Geographic Information Systems. The next trophic

level of GIS and Remote sensing application involves the incorporation of

Normalised Differentiation Vegetation Index (NDVI) as a proxy of vegetation

„damage, which this research does not explore because of time constraints.

6
1.1 PROBLEM STATEMENT

In 2005 widespread fire incidents in Chimanimani district caused damage to

property and terrestrial ecosystems due to lack of fire awareness campaigns owing

to limited financial and human resources. The Provincial Fire Strategy notes that,

“in 2005 Manicaland experienced widespread fires that affected most parts of the

district with the most affected being plantations, resettlement and communal

areas. Affected plantation hectarage for 2005 the affected area was at least 6

703h. A summary of the damage caused by fires in other land categories is

tabulated in table 1.1

LAND CATEGORY BURNT AREA (HA) %

Communal area 184 236 20
Resettlement area 87 300 30.8
Commercial farming area (Fast track) 141 412 50
Small Scale Farms 4 153 16.5
Urban 900
Table 1.1 Burnt areas in Manicaland.
The National Fire Protection Strategy launched in Chimanimani is thus a drive to

respond to veldfire problem and proactively prevent the occurrence of fires. It is

in view of the lack of applications of remote sensing instruments such as MODIS

in combination with public data that the researcher conceptualized carrying out an

assessment of the fire season innovatively using public domain data integrated

with Remote Sensing and GIS.

7
1.2 JUSTIFICATION.

The judicious use of Geographical Information Systems, Remote Sensing fire

products and public domain data in veldfire management will inform pre-fire

season and fire season planning, prediction and decision making at reduced costs,

as the technologies allow for better allocation of limited financial and human

resources institutions involved in fire prevention and management in Zimbabwe.

This is an essential pre-requisite for sustainable development. Chimanimani was

selected for this analysis because of its ecological uniqueness and its importance

to the economy of Zimbabwe. Also in 2005 Chimanimani was one of the Districts

in Manicaland with the highest number of fire incidents and damage to property.

It thus stood as an interesting area for this assessment since in 2006 the

Government through the Ministry of Environment and Tourism, the

Environmental Management Agency and the local timber companies teamed up

and launched the National Fire Strategy in Chimanimani. Indirectly, this

assessment can be taken as a barometer of the effectiveness of awareness

campaigns. Makoni was of interest because in Manicaland it is one of the

relatively drier Districts in comparison with the border district and it recorded the

highest fire incidents in the 2006 fire season.

8
1.3 AIMS AND RESEARCH OBJECTIVES.

Aim.
To develop a framework for the integration of Geographical Information Systems,

Remote Sensing data and public domain data in veldfire assessment, monitoring

and analysis in the country.

Objectives

 To spatially map veldfire incidence using GIS ArcView 3.2a software and

MODIS

 To identify imperative methods that will assist in spatial mapping and

detection of bushfires at least cost.

 To assess the effectiveness of remote sensing and public domain data in

monitoring and detection of fire incidents.

 To suggest a suitable data infrastructure to develop and support the use of

GIS and remote sensing for fire management.

1.4 RESEARCH QUESTIONS

 Where did the fires occur?

 How many fires incidents occurred in 2005 and 2006?

 Compared to 2005 did the fires occur in the same vicinity?

 Did the Districts experience a decrease or increase in the number of fire

incidents?

9
  How effective was the system of detecting and reporting fires?

 How effective were education campaigns in reducing fire incidents?

 Under which tenure did most fires occur?

1.5 STUDY AREA CHARACTERISTICS.

Chimanimani District.

It is located in the south-eastern part of Zimbabwe as shown by Fig 1.1 and it

covers an approximate area of 3353km2. It is divided into 23 administrative

wards with an average household size of 4.4 and a population of 115 250 (CSO:

2002)

Zimbabwe District Boundaries

Hurun gwe Guruve

Cen tena ry
Mt. Da rw in Rush inga
MAKONI DISTRICT
Ka rib a UMP Mu dzi
Ma kond e Sh amva
Ma zo we
Zvimb a Mu toko
Gokwe North

Bin ga Harare Mu re hwa Nyang a

Che gu tu
Gokwe So uth Ka dom a Se ke
MA R O N D E RA
Ma koni
Lup ane W E D ZA Mu tasa
Hw ang e Nkayi Kw ekwe
Chiko mba

Chiruma nzu Bu hera Mu tare

DISTRICT BOUNDARIES
Gwe ru Gutu
Tsh olotsho Bu bi 0.04 - 0.313
Sh urugw i
Umg uza CHIMAN IMANI
Bu law ayo Bikita
0.313 - 0.567
Zvisha vane
Umzingw an e
In siza
Ma svin go 0.567 - 0.934
Bu lilimama ngw e Chivi Zaka Chip ing e
0.934 - 1.671
Mb ereng wa 1.671 - 2.552
Ma tobo Chired zi
Gwa nda
Mw en ezi

CHIMANIMANI DISTRICT
Be itb rid ge

N
300 0 300 Kilometers By:
Chigurah. G.C
023896710/0912833531

Fig 1.1: Map showing the districts of Chimanimani and Makoni.

Rainfall Pattern. goe

10
Chimanimani district has a wide range of varying rainfall and temperature patterns.

Rainfall ranges from 600mm to above 1000mm.

Soils

The two major soil types that dominate Chimanimani district are:

 The dolomite and umkondo rocks, which are derived from weathered

metamorphic rocks and limestone. These soils are found largely in the Eastern

parts of the district ranging from Chikukwa in the east running through

Chimanimani, Nyahode, Rusitu Valley, Manyuseni and Gwindingwe in the west.

The soils have very high agricultural productivity, which is a characteristic of

agro-ecological region 1. The soils are also suitable for growing maize and

horticultural products. Their particles are well graded and consolidated making

them less vulnerable to erosion enabling farmers to plough and grow crops on

slopes and hilly places.

 The second group is the moderately leached soils of the ferscallitic group, which

are derived from weathered granite rocks. These soils form the major part of the

Central and Western zones. They range from light to dark grey in colour and

sandy to clay in several other areas. They are susceptible to erosion due to poor

particle grading and weak consolidation and are relatively poor in nutrients.

11
Makoni District.

Location

Makoni District is located in the Manicaland Province in Zimbabwe. The geographical

location of Makoni district is shown on the map of Zimbabwe in Fig 1.1. The total

population for Makoni District is 151 596.

Climate and vegetation

The study area falls within the Sub-region IIb, of farming regions as defined by,

AGRITEX, (1997). The region receives an average of 16-18 rainy pentads per season

and is subject either to rather more severe dry spells during the rainy season or to the

occurrence of relatively short rainy seasons, (AGRITEX, 1997). The rainy season starts

around mid October and tails off around end of March, (Meteorological 5 year rainfall

data), with the 5-year average precipitation shown in Figure 1.2 . The pick rainy period

is December to February, while June to September are the dry months, with very isolated

rainy days in June and August. In either event, crop yields in certain years are affected,

but not sufficiently frequently to change the overall utilization from intensive systems of

farming, (AGRITEX, 1997). Grasslands are predominant and open areas without trees

are quite common.

12
 200.0
Precipitation (mm)

150.0

100.0

50.0

0.0
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Month

Fig 1.2: Average 5-year monthly data for 1998-2002

Geology.

The geology in the Makoni District area largely comprises of granitic rocks with minor

greenstone remnants and dolerites, (Geological Survey, 2005). Gneisses that

predominate in the area result in a generally flat to undulating terrain, with low pavement

outcrops, and they weather into pale-sandy soils. The gneisses grade into younger,

homogenous, coarse-grained porphyritic granite that appears on the western and north-

eastern parts, (Geological Survey, 2005). The area is truncated by numerous faults, of

which two major sets are outstanding, the north/north-east and north-east trending sets.

Most of these faults are part of the Great Dyke related fractures, with a prominent

north/north-east trending buck quartz vein in the middle of Geological Survey, (2005),

the Popoteke fault system, occupying such a fracture.

13
1.6 Definition of Key Concepts.
Remote Sensing

According to Wright (2001), remote sensing is a method of visualizing the

radiative properties of the earth‟s surface using instrumentation mounted in

satellites or aircrafts.

Geographical Information Systems.

It is defined as a system of hardware, software and procedures to facilitate the

management, manipulation, analysis, modelling, representation and display of

georeferenced data to solve complex problems regarding planning and

management of resources (NCGIA: 1990).

Veldfire

It is defined as a fire that burns uncontrollably in a natural setting for example a

forest or a grassland.

1.7 Limitations.
The researcher encountered the following problems when carrying out the

research:

A lack of adequate funds and research material at the university library thereby

compromising the quality of work.

1.8 Summary.
The aim of this chapter was to give general information to the project. It

highlighted the background of the problem, statement of the problem, objectives,

and research questions. It also cited limitations of the study. The research also

defined major terms as they are complied in the study.

14
 Chapter 2

2.1. Introduction

This chapter reviews some of the literature on applications of Remote Sensing and

public domain data on veldfire monitoring. It defines key concepts and terms used

in this study. The chapter also examines issues related to veldfires especially their

causes, characteristics and impact on biodiversity. The chapter will also examine

other Remote Sensing techniques that are used in veldfire monitoring, analyses

and mapping.

2.2 KEY CONCEPTS.

2.1.1 Remote Sensing.

In simple terms remote sensing is the acquisition of data and information about an

object without contacting it physically. Common methods include aerial

photography, radar, and satellite imaging. However, remote sensing is more than

what has been expressed by this simple definition. Wright (2001) noted that

Remote Sensing is a method of visualizing the radiative properties of the Earth's

surface using instrumentation mounted in satellites or aircrafts. Remote sensing

instrumentation measures the radiation reflected and emitted from the earth at

different wavelengths, primarily at those wavelengths not absorbed by the

atmosphere. Remotely sensed data can be converted to maps showing the visible

or thermal properties of an area.

15
 According to FAO (2003), Remote Sensing covers all techniques related to the

analysis and use of data from the environment and earth resources satellites such

as Meteosat, NOAA-AVHRR, Landsat Thematic Mapper (TM), SPOT and ERS-

SAR and from aerial photographs. The main function of Remote Sensing is to

map and monitor the earth‟s resources. According to Green (2001), the quality of

remote sensing data “consists of its spatial, spectral, radiometric and temporal

resolutions. Spatial resolution refers to the size of a pixel that is recorded in a

raster image - typically pixels may correspond to square areas ranging in side

length from 1 to 1000 metres. Spectral resolution refers to the number of different

frequency bands recorded - usually, this is equivalent to the number of sensors

carried by the satellite or plane. Landsat images have seven bands, including

several in the infra-red spectrum.

Fig 2.1: data acquisation through remote sensing.

16
2.1.2 Fire Detection and Monitoring.
The use of satellites in the role of monitoring, mapping and analyzing fire

activities has been amply shown in recent years and has been summerised in

recent literatures (for example Harris et al; 2000, Flynn et al; 2000, Schneider et

al 2000). These include a group of studies that concentrate on high temporal

resolution and low spatial resolution data sets that are refreshed every few

minutes to few days, and hence are ideal for continuous monitoring of fire

activities. According to Frost (2006) the perfect fire detection satellite capable of

identifying very small fires over large areas regularly does not exist yet.

Wooster and Rothery (2000) defined MODIS as a nadir viewing system having a

swath width of 2330km and offering 36 spectral bands of data at one time of 3

spatial resolutions of 250 500 or 1000m per pixel. Of these 36 spectral bands, 10

are useful for detecting thermal thermal radiance from active fire activities.

MODIS was launched into a sun-synchronous orbit as part of the Terra and Aqua

platform of instruments on December 18 1999 and June 2000 respectively

(Morissette, Privette and Justice; 2002).

Justice and Giglio; (2002), stated that the MODIS active fire products fall within

the suite of terrestrial products and provide information about actively burning

fires including their location and timing, instantaneous radiative power and

smoldering ratio, presented at a selection of spatial and temporal scales.

17
 The two MODIS satellites are polar orbiting that is they are moving around the

North and South Poles every 98 minutes while the earth is turning from west to

east. Terra scan the Southern African region between 10h00-11h30 while Aqua

scans in the afternoons between 14h00-15h30. Each satellite also scans the region

at night (Terra at 22h00 and Aqua at 03h00). Polar orbiting satellites have the

advantage of detecting small fires, but with a very low overpass frequency, while

geostationary satellites have the advantage of frequent views over large areas

(every 15 minutes) but with a very low resolution that is can detect big fires only

(Frost et al 2006).

2.1.3 Limitations
Each fire detection represents the centre of a 1 km pixel flagged as containing one

or more actively burning fires within that pixel. In any given scene the minimum

detectable fire size is a function of scan angle, biome, sun position, land surface

temperature, cloud cover, amount of smoke and wind direction. The precise value

will vary slightly with these conditions but generally, in many biomes the

minimum flaming (~800-1000K) fire size typically detectable at 50% probability

with MODIS is in the order of 100m² (NASA, 2004a), because each fire event is

fitted in a 1 km pixel independently of its size, makes the exact location of the fire

incident difficult in the field unless the fires are directly monitored and mapped in

the field. Small hot fires late in the dry season can be detected and been shown in

1 km pixel which makes searching for the exact position in 1 km² grid difficult.

MODIS fire information can be very powerful fire monitor tool where ground

truthing can be carried out immediately after or during the fire incidence.

18
2.1.4 Wildfire Temperature Retrieval
Wildfire temperature retrieval commonly uses measured radiance from a middle

infrared channel and a thermal infrared channel to separate fire-emitted radiance

from the background emitted radiance. Dennison; (2003), suggested that emitted

radiance at shorter wavelengths, including the shortwave infrared, is measurable

for objects above a temperature of 500k.The spectral shape and radiance of

thermal emission within the shortwave infrared can be used to retrieve

temperature.

Fire propagates through the combustion of fuels consisting of live and dead plant

material. Fuel temperature must be high enough to volatilize and ignite these

materials. Once ignition has occurred, the energy released through combustion

raises the temperature of adjacent fuels. Pyne et al., (1996); link stages of

combustion to temperatures at which they typically occur. As fuel temperature

increases above 400k, the volatilization of fuels begins in a process called

pyrolysis. As the temperature of the combusting fuels increases, the energy

radiated by the fire increases and shifts to shorter wavelengths. By measuring

thermal emission within multiple channels, remote sensing can be used to

determine the dominant temperature and location of a fire.

2.2 Geographical Information Systems

Like the field of Geography, the term Geographical Information Systems is hard

to define. It represents the integration of many subject areas. Accordingly there is

no absolutely agreed upon definition of a Geographical Information Systems as

postulated by DeMers, (1997). A broadly accepted definition of Geographical

19
 Information Systems is the one provided by the National Centre of Geographic

Information (NCGIA: 1990), which defined a Geographical Information Systems

as a system of hardware, software and procedures to facilitate the management,

manipulation, analysis, modeling, representation and display of georeferenced

data to solve complex problems regarding planning and management of resources.

According to Dr.Escobar., et al:(2003); geographical information systems have

emerged in the last decade as an essential tool for urban and resource planning

and management. Their capacity to store, retrieve, analyze, model and map large

areas with huge volumes of spatial data has led to an extraordinary proliferation

of applications. Geographical Information Systems is now being used for land use

planning, ecosystems modeling, landscape assessment and veldfire monitoring.

2.2.1 GIS Applications.

According to Berhardsen, (1996), GIS can be used to perform the following tasks:

a) Mapping Locations.

It allows for the creation of maps through automated mapping, data capture and

surveying analysis tools.

b) Mapping Quantities.

People map quantities like where the most and least are, to find places that meet

their criteria and take action, or to see the relationships between places. This gives

an additional level of information beyond simply mapping the locations of

features.

c) Mapping Densities.

20
 While concentrations can be seen by simply mapping the locations of features, in

areas with many features it can be difficult to see which areas have a higher

concentration than others. A density map lets you measure the number of features

using a uniform aerial unit such as acres.

d) Mapping and Monitoring Change.

GIS can be used to map the change in and to anticipate future conditions, decide

on a course of action, or to evaluate the results of an action or policy.

2.2.2 Other Sensors Currently Being used to Provide Information About Forest
Fires
i. Advanced Very High Resolution Radiometer (AVHRR).

Advanced Very High Resolution Radiometer is the most commonly used satellite

sensor for detecting fires over large regions or entire countries. It can provide a

daily snapshot of Canada at a 1-km nominal resolution in fire spectral channels.

How Fires are Detected by the AVHRR Satellite Sensor.

The most important AVHRR channel for fire detection is the mid-infrared channel,

which measures a combination of reflected and thermal energy at a wavelength

around 3.7 micrometers. This channel is highly sensitive to objects that are

emitting thermal energy at high temperatures over 200 degrees Celsius such as

vegetation fires. For this reason, the AVHRR sensor can detect fires that cover

only a fraction less than 0.1% of a 1.2 km2 AVHRR pixel. Since other types of

objects, such as cloud edges and exposed soil also produce a large response in the

mid-infrared channel, information from other AVHRR channel is required to filter

out extraneous data (false alarms); (Li; Khananian and Fraser, 2000).

21
 i. The satellite Pour L‟Observation de la Terre (SPOT) Vegetation (VGT) Sensor.

According to Li; Nadon and Cihlar, (2000); the sensor was launched in 1998 and

it has four channels measure reflected energy from the earth. Like AVHRR, VGT

provides daily coverage at a 1-km resolution, because it lacks thermal channels,

the sensor is not well suited to detecting active fires as AVHRR. However it does

include near infrared and short-wave infrared channels that are highly effective

for mapping the extent of burned forest after a fire has stopped burning. VGT

imager is currently being used for annual mapping of burned forests across

Canada, (Li, et al; 2000).

ii. Landsat‟s Thematic Mapper (TM) Sensor.

The sensor allows areas to be observed at 30-m resolution in seven channels. The

high resolution comes at the expense of observing a given location only once

every 16 days. Thematic Mapper is best suited to providing detailed maps of areas

burned by individual fires or fire complexes. These maps can be used to plan

salvage logging operations and to verify the extent of the burned areas mapped

using coarser resolution VGT imagery; (Fraser et al, 2000).

2.3 Causes of Fires

In tropical savannas of Africa, intentional burning has been practiced for at least

50 000 years and has changed little from those of the ancestors (Rose-Innes 1971,

Chidumayo 1997) People cause fires in the preparation of land for cultivation,

clear bush and undergrowth to improve visibility around settlements and foot

paths, keeping away dangerous animals, clear roadside areas before or after grass

slashing in road maintenance operations, management of grazing land for both

22
 livestock and wildlife, charcoal burning, honey collection and hunting. Fires also

originated accidentally by transients in abandoned kitchens left alight in overnight

resting camps along pathways linking distant settlements. Cultural and tradition

believes of the many inhabitants of miombo region discourage bush burning until

crops have been harvested because of the danger fire pose to crops, while the first

rains usually marks the end of the burning season (Chidumayo, 1997). In the

study area fire are still set for other purposes such as hunting and improvement of

grazing areas and often cause great losses in properties and human life. Man made

fires are by far more frequent than natural fires and natural fires are mostly caused

by lightning particularly at the beginning of the rain season (West, 1972).

2.4 Different Types of Veldfires

Normally they are classified by where they are burning and the temperature of the

fire:

a) Burning Location.

There are two types of wildfires. These are surface fires, which burn along the

ground and crown fires, which burn the upper parts of trees. Ground fires occur in

mixed Mopane woodlands with little grasscover. Crown fires occur in the timber

plantations of the Eastern Highlands. Ground and crown fires occur in typical

savanna woodlands in many parts of the country. Many late season late season hot

fires are a combination of surface and crown fires.

b) Temperature at which they are Burning.

More importantly in the Zimbabwean situation, veldfires are classified as either

cool fires or hot fires. Cool fires usually occur early in the dry season when there

is still moisture in the grass and leaves. Hot fires usually occur late in the dry

23
 season when the grass and trees are very dry. These fires are often made hotter

and more destructive by the windy conditions in August, September and October.

2.5 Factors Determining Incidents and Effects of Veldfires

There are many factors, which determine the occurrence and effects of veldfies

such as weather patterns, socio-economic and demographic trends and vegetation

types.

2.5.1 Vegetation Type.

The type of vegetation and the amount of rain, which falls in any given season,

have a marked influence upon the effects of fires. In CAMPFIRE areas supporting

wildlife, there are three broad environmental regions. In each region fire can be

expected to behave differently:

a) The Zambezi Valley Escarpment and Foothills.

There are four main vegetation types found in the hills region. The incidents and

effect of fire will vary between each vegetation type.

Mopane woodlands are found in the low lying and low rainfall areas. They do not

usually present major fire problems because they contain very little grass.

Jesse thickets are found on the deep sandy soils. They also do not present major

fire problems because they contain very little grass.

Riverine woodlands are found along the major rivers. The incidence effect of fire

will depend upon the amount and conditions of the grass they contain.

Mixed woodlands are found in the higher rainfall hilly escarpment areas of the

Zambezi Valley. These well-grassed and wooded areas are most sensitive to hot

late dry season fires.

24
b) Lowveld and Matebeleland.

The two main vegetation types over much of the south-eastern Lowveld and

Matebeleland are Mopane woodland and the Mixed Combretum Acacia

woodlands. In periods of very low rainfall, veldfires in this region are less likely

as there is little grass to burn. In better rainfall years when there is a lot of grass,

there is a high risk of hot late season fires.

c) Eastern Highlands.

In the Eastern Highlands there are three major vegetation types. These are the

natural forests or woodlands, grasslands and the exotic timber plantations. The

natural forests do not have enough dry vegetation to be a fire risk. The grasslands

and the exotic timber plantations are extremely sensitive to fire. The natural

forests can often be scorched and badly affected by the hot fires on the grasslands.

2.5.2 Socio-Economic and Demographic Trends

Socio-economic and demographic trends that have prevailed in rural areas of

Europe during the last four decades reinforce the susceptibility to fire. Many rural

areas have experienced substantial population decreases during the second half of

the 20th Century, leading to the abandonment of agricultural lands, decrease in the

sizes of herds and a reduction in the consumption of forest fuels by animal

grazing and by human harvesting for firewood (Pansas and Vallejo; 1999). Areas

of marginally productive agriculture were converted to forest plantations or

abandoned to the natural process of ecological succession and thus converted to

shrublands and woodlands (Pyne; 1995 and Mareno; 1999). This has led to an

increase in the connectivity of fire prone vegetation patches, generating

25
 landscapes more vulnerable to the uninterrupted spread of large wildfires (Lepart

and Debussche; 1992).

2.5.3 Meteorology
The meteorological variables most influencing the development of fires are

temperature, wind speed, air relative humidity and stability of the atmosphere. In

the stable and dry summer environments the energy received from the sun

increases temperature and reduces the relative air humidity. Both variables

(temperature and relative humidity of the air) control the hydration state of dead

fuels. The number of days with maximum extreme temperatures will increase in

the summer.

The spread of fire is favoured during the day by the temperature increase and the

decrease in relative air humidity, which can reduce the moisture content of dead

fuels, lowering the threshold for ignition and making an ignition event more likely

to lead to fire. Likewise, night-time temperature increases will be proportionally

greater than the daytime ones (Easterling et al.1997). In other words, temperatures

during the night tend to become comparatively higher, with the consequent

negative effect on fuel moistening. Assuming that the number of ignition sources

and the vegetation do not vary, flammability can therefore be expected to be

greater and the fires more frequent, and that once they have broken out, they will

spread better and get bigger .The tendencies for precipitation during this century

are not consistent among models, although, they all agree the total annual

precipitation will decrease, in particular in spring and summer.

26
Precipitation patterns determine the level of soil moisture reserves, and recharge

periods are critical with regard to providing the soil with greater stability in water

content (Martínez-Fernández and Ceballos 2003). Assuming that total

precipitation does not vary, the concentration of this in winter and the consequent

lack of rainy days in spring and summer will affect live and dead fuels. This,

together with the temperature increase, will cause and increment in potential

evapotranspiration (Pausas 2004). Rainfall during the growth period has a great

influence on the abundance of herbaceous species (Figueroa and Davy 1991).

Rainy springs maintain more surface moisture in the soil, leading to greater

development of fine herbaceous fuels, which will subsequently dry out.

Temperature increases may cause the development period of herbaceous species

to be advanced to early spring or winter, so that, even in a scenario of reduced

springtime rainfall, this vegetation may develop well, thus contributing an

element of hazard relatively early in the year. This is more relevant in the humid

areas which, in time, may become more susceptible to greater summer dryness,

which may also appear earlier in the year. Furthermore, less moisture availability

in the surface layers of the soil will make the dead fuels in the soil dry out sooner.

The lower number of rainy days will keep them dry for a longer time. In pine

forests and ecosystems with well-developed litter, there will be an increase in

flammability and in the period of susceptibility to fire.

27
Fig 2.2: Climate change in Zimbabwe: trends in temperature and rainfall.

Zimbabwe is experiencing more hot days and fewer cold days, and the amount of

precipitation it receives is deviating from the mean more frequently. This resource

includes two graphics. The first shows the number of days with a minimum

temperature of 12 degrees Celsius and the number of days with a minimum

temperature of 30 degrees Celsius from 1950 to 1990. The second graphic shows

the amount of precipitation in millimetres that was a departure from the long-term

mean amount for the time period 1910 to 2000. The trends shown on both graphs

favour the ignition of fires.

Wind is another critical element: the speed of the spread of the fire front is

directly proportional to wind speed. The most dangerous situations are those

28
involving strong, dry winds. Föhn-type winds are particularly critical; these are

winds blowing on the leeward side of mountains as a result of the adiabatic

compression of the air on blowing down the slopes (Millán et al. 1998), and these

are the cause of some of the large fires in Spain (Gómez-Tejedor et al. 2000).

Fire heats the air, which rises, drawing in cool air towards the base of the fire,

which provides oxygen to sustain combustion. When there is wind, this effect is

augmented on the downwind side by the wind-driven airflow. The stability of the

lower levels of the atmosphere determines the degree of intensity of the local

wind caused by the fire. Situations of atmospheric instability favour the vertical

movement of the hot air, facilitating the lateral movement of the air towards the

fire front. To the contrary, in stable conditions, fires are relatively less dangerous.

Thus, with two parameters of atmospheric stability, Díez et al. (2000) calculated,

to a high degree of accuracy, the daily occurrence of fires in Galicia. The synoptic

situations determining the state of the atmosphere are therefore critical for the

occurrence of forest fires (Díez et al. 1994). These determine atmospheric flow,

and, through this, wind, precipitation or lightening discharges, among other

phenomena (Gómez-Tejedor et al. 2000; González-Hidalgo et al. 2001; Goodess

and Jones 2002; García-Herrera et al. 2003; Muñoz-Díaz and Rodrigo 2003;

Tomás et al. 2004). Consequently, many fires occur in determined synoptic

conditions (Bardají et al. 1998). This is similar for the rest of the world (Da

Camara et al. 1998; Johnson and Wowchuk 1993).

29
2.5.4 Topography.
The spread of fire increases with the angle that the terrain offers to the fire front.

An upslope spread is therefore rapid and dangerous. Fires do not occur by chance,

but rather are more frequent in certain topographies (Vázquez and Moreno 2001;

Lloret et al. 2002). Although topography may not change; vegetation can do so,

particularly after a fire. This makes the risk in a given area vary in time as the

vegetation changes, and according to the topographic conditions of the area in

question.

2.6 Impacts of Veldfires

Effects on natural vegetation

Fire is important in the dynamics of tropical vegetation, primarily as a natural

disturbance that has little effect on some plants. The effects depend partly on the

timing and intensity of the fire, and on the intrinsic attributes and physiological

state of the individuals and species concerned. A succession of fires generally has

more effect than any single fire. Frequency is therefore an important variable, as

are the conditions for regeneration during the inter-fire interval (FAO, 1999).

Frequent late dry-season fires transform woodland into open, tall-grass savanna

with only isolated fire-tolerant canopy trees and scattered understorey trees and

shrubs. Trees less than about 2 m tall are more susceptible than bigger

individuals. These fires suppress the regrowth of woody plants, preventing their

recruitment to the canopy and changing the system‟s dynamics. Grasses are

30
generally less affected as they are largely dormant during the dry season (Frost

and Robertson, 1987).

Complete protection from fire, and to a lesser extent early dry season burning,

favours the recruitment and growth of woody plants, which gradually shade out

the grasses. In the savanna-forest boundary zone, for example in West Africa,

protected sites gradually become invaded by more fire-sensitive forest species

(Goldammer et al., 2004).

Effects on inherent soil fertility

The understanding of the effects of fire on soil organic matter and nutrients comes

mainly from small long-term experimental plots burned at specified regular

intervals and times of year. The results may not be representative of large

heterogeneous landscapes subject to a more variable fire regime, but it is unlikely

that these would be more pronounced than those found experimentally (FAO,

2001).

Fire alters the timing, location and magnitude of nutrient cycling from plants and

litter to the soil, as well as nutrient availability in the soil. Depending on the

intensity of the fire and the temperature at which an element is volatilised,

nutrients may be volatilised and enter the atmosphere, or deposited in the ash.

Some ash is transported elsewhere by wind or water, but this is local

redistribution rather than a loss.

31
Nitrogen is particularly susceptible to volatilisation, with reported losses of up to

33 kg N ha in African savanna fires. Losses are likely to be greater from static

fires linked to deforestation Such losses could be expected to reduce long-term

nitrogen availability in these systems, though they are offset to some extent by the

deposition of nitrogen in rainfall and nitrogen fixation by micro-organisms.

Apart from nitrogen, most nutrients released by fire are reincorporated in the soil

with the ash causing the concentrations of these elements in the soil to be higher

for some weeks after burning. Nevertheless, over time, the stock of exchangeable

bases declines on plots subjected annually to hot fires for many years.

Phosphorus is often limiting in tropical soils, and recorded levels of extractable P

in the surface soils of annually burnt sites are higher than those of unburnt sites.

Moreover, plots subjected to more intense, late dry season fires have more

phosphorus than either early-burn or no-burn plots, probably because more

material is burned, which produces more ash, a higher soil pH, and more available

phosphorus (FAO, 2005).

Since fire reduces the amount of litter on the soil surface, repeated burning should

eventually diminish soil organic matter levels. Likewise, high temperatures may

destroy some of the soil organic matter close to the surface, though this depends

on the magnitude and duration of heating. Not surprisingly, therefore, organic

carbon levels are lower on many annually burnt sites and those burned during the

32
late dry season compared with early-burn and unburnt ones (Frost and Robertson,

1987).

Repeated burning and the associated changes to litter and soil organic matter

dynamics should affect the composition and functioning of the soil biota, though

few studies have been undertaken in Africa on this. In general, burning reduces

soil microbial populations, though the effect is transient.

Effects on soil hydrology, runoff and erosion

While adverse effects of fire on runoff and soil erosion are frequently asserted,

good evidence is generally lacking. Most studies have been done on small

experimental plots where the effects are usually small, temporary and not clearly

discernable at a landscape or larger scale.

Burning reduces both vegetation and litter cover. This exposes the soil to

raindrops and solar radiation, and alters rainfall interception, infiltration, and

runoff, depending on soil surface properties and the topography of the landscape,

and on how long it takes for the vegetation to recover. Post-fire activities such as

grazing and ploughing also affect the outcome. Given this complexity, a wide

variation in the impact of fire can be expected (Barbosa et al., 1999).

Fire obviously reduces evapotranspiration but generally only temporarily, lasting

little more than a few months in grasslands, where the plants recover relatively

33
quickly, to one or two years in wooded savannas, where full recovery takes

longer.

The effects may be longer-lasting in forests where the canopy is severely

damaged and recovery is slow because of intrinsic limitations on growth or

subsequent disturbances. Although stormflow volumes may increase temporarily,

there is little evidence for a discernable impact on streamflow and total annual

water yield at a catchment scale. The small increases recorded from mountain

catchments in South Africa are in line with that expected from the reduction in

evapotranspiration and interception, both of which subsequently increased as the

vegetation recovered (Andreae, 1997).

The potential for greatly increased soil erosion following burning is a common

cause for concern. Whether soil loss increases depends on how long the soil

remains exposed after burning, the erodibility of the soil itself, the angle and

length of slope of the burned area, and the extent to which rainfall intensity

exceeds the infiltration capacity of the soil. Infiltration rates may be reduced. Ash,

together with volatilised organic compounds deposited on the surface, can form

partial water-repellent seals, though this effect is also transient.

34
2.7 Contribution by the Community in Fire Management and Fire Incidents
Mapping.

Fire is a disturbance that has played and will continue to play a major role in both fire

sensitive and fire adapted ecosystems throughout the world. In almost all of these

ecosystems, humans have altered the natural fire regimes by changing the frequency

and intensity of fires. In many parts of the world, local communities are often blamed

for what are considered harmful fires. This view often encourages fire and forest

management institutions to perceive local communities as part of the problem, and

certainly not the solution. Evidently because local people usually have most at stake

in the event of a harmful fire, they should clearly be involved in mitigating unwanted

fires. Community Based Fire Management (CBFiM) is an option for blending

participatory community development strategies and fire management to reduce and

report unwanted fires and their impact (Darlong, 2002).

During the last few years, there has been quite a lot of discussion about what is now

referred to as Community Based Fire Management (CBFiM). The term has been used

to describe such a wide variety of different ways in which communities are involved

in fire management, that it is difficult to make any systematic comparisons or

generalisations. The definition proposed in recent work (Ganz et al 2003) is; CBFiM

is a type of land and forest management in which a locally resident community (with

or without collaboration of other stakeholders) has substantial involvement in

deciding the objectives and practices involved in preventing, controlling or utilizing

fires.

35
The term „community based‟ in the context of fire covers a wide spectrum of

situations from potentially forced engagement in an activity (coercion), to free and

willing participation in actions developed by local actors themselves (empowerment).

The emphasis on community based is not only the community involvement but also

where community capacity has been recognised and supported by external agencies

(governments, non-governmental organisations, projects and others). This may

include support to an existing indigenous system through formalising, modifying or

otherwise elaborating on it or instituting new systems (Moore et al 2002). Many of

these systems and approaches are considered more effective in tempering

uncontrolled fires, more beneficial to local ecosystems and more cost efficient over

the long term. More common are instances where CBFiM has resulted from the

formation of community institutions and mechanisms that support more efficient fire

management entities (such as the two cases documented by FAO from Cal and

Berganain Turkey-FAO 2003). In some countries, the driving force behind CBFiM

approaches is indigenous land and /or user rights, including the right to use fire as

management tool. The security of these rights may ultimately help maintain the

beneficial use of managed fires for such objectives as controlling weeds, reducing the

impact of pests and diseases and generating income for non timber forest products

(Arnstein et al 1969).

Several of the CBFiM approaches documented in various sources occur in remote

locations where the government‟s fire control/suppression approaches are severely

hindered by access and response time. In such remote locations, communities are

36
 present and have a significant role to play in the prevention and suppression and

reporting of harmful fires that have a detrimental impact on their lives thereby

contributing to the mapping of fire incidents but the government must not relinquish

all accountability.

2.8 Summary
The chapter has reviewed the available literature concerning MODIS instrument and

other veldfire detection satellite instruments. It has dealt with the factors affecting fire

incidents, causes and impacts of veldfires in depth. Studies by other scholars were also

included in order to provide a wider understanding of applications of remote sensing

in veldfire monitoring. The last part of the chapter provided the brief overview of

community based fire management and how the community can be involved in

mapping fire incidents.

37
 CHAPTER 3

RESEARCH DESIGN AND METHODOLOGY

3.0 Introduction.

Research design is the most important stage in any scientific work. It provides a

framework for the establishment of sample units, data collection, analysis and the

outputs. To make the most of a research work, the design must be a prior

involving careful thinking around what is intended to be achieved. Poor design

may result in inconsistent results and ineffective use of resources. This chapter

presents the tools and methods that were used to carry out the research and the

analytical framework that guides the study. It also contains the explanation of the

theoretical concepts for collecting and analysing data.

3.1 Research Design

Zikmund (1996) defined a research design “as a master plan for collecting and

analysing the needed data.” The research design therefore gives the framework of

the research plan of action. The research design used in this study was both

exploratory and descriptive in nature. Fig 3.1 is a flowchart, which helps to

illustrate the different stages that were taken by the researcher in coming out with

the required results.

38
 MODIS DATA
from the
MODIS website
on the internet

OFFICER‟S
MONTHLY
Buffering, map FIRE REPORTS
making and
DATA
map overlaying
ANALYSIS

GIS ArcView DATA

3.2A INTERPRETATION
Applications

DATA OUTPUT

Maps Graphs
Tables

Fig 3.1: Flowchart of different stages taken in coming out with the results.

3.2 Exploratory and Descriptive Research.

According to Skinner (1990), an exploratory study is basically based on

secondary data that are readily available. Kotler (2000) described descriptive

research as research undertaken when one needs to understand the characteristics

of certain phenomenon to solve a particular problem. Focus was on the

39
 application of remote sensing and public domain data in mapping and analysing

veldfire dynamics. In this study the researcher used exploratory study in his quest

to analyse and map the 2006 fire incidents for Chimanimani and Makoni Districts.

The researcher also investigated the problems associated with the use of remote

sensing and public domain data in both monitoring and mapping fire incidents.

3.3 RESEARCH INSTRUMENTS.

Primary sources

There were interviews using questionnaires and this was done to individual

officers from the two districts. The major advantage of these interviews was that

an opportunity was offered of conveying information while getting an immediate

response. Interpersonal communication was impinged during the interviews

although some of the officers were expecting to receive some kind of help from

the interviewer. The other good thing about primary data is that it is not outdated,

meaning that the information is recent.

Questionnaire

Open ended and closed questions were designed and delivered to 8 officers from

Chimanimani and Makoni district, which was done randomly. These interviews

were convenient in the sense that the feedback was quick and some arising

questions were being asked during the interviews.

Selection of the sample

The rationale in selecting the 8 officers to take part in the study was due to the
fact that the districts are composed of 8 officers, 4 from each district.

40
 Desk Research.

This is the collection of data, which is related to the research under study and is

already in existence. As stated by David and Chava Nachmias (1981), this

involves the researcher going into files and written materials and collecting facts

he/she wants. The fire incidents were collected on the MODIS website on the

Internet and from monthly reports submitted to the Environmental Management

Agency Manicaland Provincial Offices by the District Officers. This made the

research less time consuming and less costly since the data was readily available.

Advantages of MODIS Instrument.

 MODIS data set is refreshed globally almost twice every diurnal cycle that is once

during the day and once during the night.

 MODIS data are collected from a number of receiving stations around the globe and

processed at a central facility, the Goddard Distributed Active Archive Center

making implementation of global scale search algorithms possible.

 MODIS image data are available free of charge through a data distribution network

(http://edcimswww.cr.usgs.gov/pub/imswelcome/plain.html). These distributed data

sets include 500m spatial resolution near-Infra Red data that can be used to

supplement thermal alert observations.

Disadvantages of MODIS

 High probability of cloud cover in the rainy season makes small fires difficult to

detect.

41
 Advantages of Public Data.

 The public is directly involved with the ecosystem hence they have the answers of

where and how fires in their local area may have started.

Disadvantages of Public Data

 Officers‟ reports did not provide geographical co-ordinates of fire incidents.

 As a result of immobility and inaccessibity some fire incidents were not recorded.

Analysis Operations Conducted

The assessment, mapping and analysis of the Chimanimani and Makoni fire incidents for

2006 will be done using Arc View 3.2 Geographical Information Systems Software with

spatial analyst and extensions. A combination of the following was used:

i. Buffering.

A buffer zone is an area of a specified width that is drawn on one or more map

elements. For example to define a forest area were logging is not permitted, a 200m

buffer zone can be drawn around dust roads. Complex analysis may require values to

be calculated for large number of point locations and may involve overlay operations

with multiple data layers. For this research the researcher will use a 200m buffer zone

to assess the fire recurrence interval.

ii. Overlaying.

Standard overlay operators were used, which make use of two or more input data

layers which are geo-referenced in the same system and overlap in the study area.

The principle of the spatial overlay is to compare the characteristics of the same

42
 location on both data layers and produce a new characteristic of each location in the

ultimate levels. Simple operations are performed for example laying a ward

boundary map for Chimanimani over a map of local landuse intensity for the same

area.

iii. Standardisation of Datasets.

Since the data used originated from different sources, the data will be standardised

using the following parameters:

 Projection: UTM 1927

 Spheroid: Clark 1880

 Universal Transverse Mercator Zone: 36

 Central Meridian: 33

 Scale Factor: 0.996

 False Easting: 500 000

 False Northing: 0

 Distance Units: Meters.

Microsoft Excel

The researcher analyzed all quantitative data using Microsoft excel to produce bar

graphs, line graphs, and all other graphs.

43
Data Presentation.

This section entails the overall products to organize, describe and analyze the collected

data.

Data presentation- The researcher chose the data presentation methods such as

tabulation, percentage calculations, maps and graphic presentations such as bar graphs

and line graphs. These methods organized data by categories to facilitate cross

comparison of variables. It enabled summarization of the data to get general trends and to

communicate meaning of data.

Graphic Display – allows more summarized presentation of data were reader could

quickly make sense of the data collected. A very short explanation of the graph would

follow thus helping the reader to easily follow through the process.

Tabulation – data simply arranged in a table form.

Data analysis is the process of systematically applying statistical and or or logical

technique to describe, illustrate, condense, recap and evaluate data. The data is processed

as follows.

Editing-The researcher then used editing techniques to put sense to the data researched

for analysis.

44
3.4 Summary
This chapter gave an overview of the research methodology utilized in this study.

The reasons for applying the descriptive and exploratory research design were

discussed. The justifications of the sample survey method used in this study were

also treated including also the research instruments utilized and the methods

employed to ensure validity and reliability.

45
 Chapter 4

RESULTS AND DISCUSSIONS.

4.0 INTRODUCTION.

This chapter focused on the presentation, interpretation and discussion of the

results. The main purpose of this chapter was to review the results that were

obtained from the research. In this chapter data is presented in various forms,

which include tables, graphs and charts. The data is described and analysed also

then discussed and interpreted as findings are presented and analysed in detail.

The 2006 fire season saw a more proactive stance being taken by the

Environmental Management Agency in a bid to stem the tide of destruction and

incidence of veldfires or Bushfires. The launch of the National Fire Prevention

strategy in Manicaland province was a landmark occasion not only for the whole

nation but also the Province. The National Fire Prevention Strategy ushered in

„new‟ fire prevention and management approach, which in part saw the collection

of fire statistics in the Province by the Environmental Management Agency.

Below is a cumulative fire report for the period 1 July to 31 November 2006.

46
4.1 RESULTS

4.1.1 SPATIAL LOCATION.

Chimanimani District.

Wild fires start early on in the dry season, in about May or June. Most fires

however, occur between August and October and sometimes November. The

exact time of year at which fires start and finish depends on the past rainy season

and the start of the rains in the current season. For Chimanimani 2005 most of the

fires were located in and around forest plantations of the Charter Estate, Martin

Forest and Gwindingwi Forests. Availability of excess burning fuel from dry tree

leaves and underlying grass gives rise to early ignition hence promoting a lot of

fire outbreaks. Land use practices in communal lands do not produce favourable

conditions for fire outbreaks. Practices such as cultivation, animal grazing and

cutting down of trees by the community for firewood reduce the amount of dead

dry fuel on the ground thereby reducing the possibility of major fire outbreaks.

Wards such as Nyanyadzi, Chakohwa, Changazi,Guune and Mhakwe did not

experience any detectable fire outbreaks. Shinja experienced 6 outbreaks; Rupise

2; Mhandarume1; Chikwakwa 2; Chikukwa 1; Ngorima A and B 14; Biriri 3;

Nyahode Resettlement 44 and Bumha (Shinja Resettlement) recorded the highest

outbreaks of 35 incidents due to the availability of burnt fuels because its forests

are intact and there are few human and animal tracks that can act as natural

fireguards in the face of a fire outbreak.

47
 Chimanimani 2005 fires T
$
T
$
T
$

T
$
Month of Occurrence T
$
T
$
T$ $
$ T
T
$ T
$
$ T
T T
$
$$
TT $
T T $
$ T
T
$ T
$
T
$ T $
T T $T
T $
$ $
MHANDARUME T
$
T
$ T
$ T
$ Chima nimani 200 5 Fire Incidenc es
$$
TT Chima nimani Fires M onthly Oc currenc e

N GUUNE T
$
T
$
CHAKOWA T
$
T
$ T
$

T
$
T$
$ T T
$
T
$
T
$
T $
T $
$ T
T
$
T
$ T
$
RUPISE BUMHA (SHINJA RS)
T
$
T$
$ T $ $
T T $
T
T
$ T
$ T
$
T$
$ T
T
$
T
$ T $
T
CHIKUKW
$ A
T$
$ T T$
$T$T T
$ T
$
$ $
T T
T$
$ T$
T T
$ $
T T$
T
T
$ $ T
T T
$ T
$
T $
$ T T$
$
$ $
T T
$
$ $
T T $T $ $$
TT
T
T$
$T
T
$ T
$
T
$ T
$ T
$ T$
$ T $
T $ T
T
$ T$
T$
T
T$
$ T
$
TT
$
T
$
T
$
T$
$ TT$
$ T$
$
T
$ T T
T$
$ T $
$ T T
$
SHINJA$ T $
$ T $T$
T $$
TT$
T$T T T
$ T$
$ TT$
$ T$TT
$
T
$ T $
T $$
T T$
T$T $
T T$
$ T
$
T
$T
$ T
$
T T
$ $$T $ $
TT
$T$
T
$T T
$ T
$
T
$
T
$ T $
$ T
T
$
T
$
T$
$ T
$T
$ T
$
NYANYADZI T$
$ T T$ T T
$ T $
$ T
T $ T
T
$ T
$ T $
$ T$
T$
T
$ T $
$ T$
T T
$ T$
$ T
$
BIRIRI T $
$ T
T
$ T$
$T$
TT $T$T$
$
T T$
T T$
$TT $T
T
$ T $
$ T $
$ TT$
$ T $
T$ T
$ T
$
T
$ T$
$ T T
$
TT $T
$ T
$ T
$ T
$
T
$ T
$ T T
$ T
$
KUDZANGA T$
$T T$
$ T
T
$
T
$
T
$ T$
$ T
$ T
$T $
$ T T
$
T
$
T$
$ T$
T $
$ T T $
$ T $ T
$
T$
$ T $
T $
$ T$
T$
T T
$ T$
$ T$
TT
$T
$T $
$ T$
T
T
MHAKWE T T
$ T T$
$ T$
T$ T$
T$
T$
$ T$ T$
$T$T$T$
T T$ T
T $
$ T$
T$
$ T $ T
$
T T$ T
T
$ T$
$ T$T
$
TT
$
T
$
T$
$
T
T$
$ T
$
T $
T
$
$ T T$
$TT
$ T $T$
TT
$T $
$
T T$
T $ T $T T$
$ T T
$
CHIKWAKWA
T $
$ T T $T T$
T$
T$
T T
T $
$ T
$ T$
$
T
$
T
$ T
$
T $
$ T $
$ T T
$T$
$
T
$T$
$ T
$T
$
T
$ T $T
T $
$ T T
$ T$
TT
$ T
$ T
$
T$
$T $T T$
T$
$
T$
TT
$ T
$
TT$
$ T
$
T
$
T$
$T$T$
T T $
$ T
$
T T $
$ T
T
$ T$
$ T$
TT $
$ T
$
T $
$ T $T T $T $ T
T
$ T
$
CHANGAZI T
$T
$
T
$
T
$ $ T $T T
$ T
$
T $
$ T $ T
$ T $
$ T
T T $
$ T
T$
$ T$
T$T$ T
$T
$ T
$
T
$ T $
$ T$T T
$ T
$ T
$ T
$
T
$ T
$ T
$
TT
$ T$
$ T
$
T$
$ T NYAHONDE (NAGODI RS) T$ T$
T$
$T$ T$
$ T T
$
T$
$ T$TT
T
$
T
$ T $
$ T
$ T
T
$
T$
$TT $
$ TT $
$ T T $
$ T
T
$
T
$ T $
$ T T $
$ T T
$
T$
$TT
$
T$
$
T$
$T T
T
$
T$
$T
T
$
T
$ T $
$ T $ T$ T $
$ T
T T$ T
$ T$
$ T
$ T $
$ T
$ T
$ T$TT$
$ T
$
TT$
$ T T $
$ T T
T $
$ T$
T $ TT T$
$
T
$
T
$
T$
$T
T
$ T $
$ T T
$ T$
T$T $
T T $
$ T
T $
$ TT
$ T
$ T
$ T
$ T
$T
$T
$
T
$T
$ T
$T
$ T
$
T $
$ T T$
$ T$
T
T $
$ T$ TT
T$
T$ T $
$ T T
$ T
$T$
$ T$ T
$
T
$ T$
$ T T $
$ T $
TT$
$ T T $
$ T
$
T
$
T
$
T
$T$
$TT$
$ T$
T
$ T T $
$ T
T $TT$
$T T $
$ T
$
T
$ T
T
$T$
$TT$
$T$ T
$ T $
$ TT$
T $ T
T $
$ T $
$ T$
$ T
$
TT$
$TT$
$T $
$
T T T
$
T
$ TT$
$
T$
$ T
$ T $
$ TT
$ T$
T$
$ TT
$
T
$T
$
T
$
TT$
$ T
T $
$ T $
T
$
TT
$ T
$
T T
$ T
$ T
TT
$ T
$T
$ T
$ T
$ NGORIMA B
T
$ T $
$ T$TT$
$ T
$
T$
$ T
$ T$
$ T
T T $
$ T
0 3 6 9 12 Kilometers
T
$
NGORIMA A
T
$
T
$
T
$
CHABIKA

Fig 4.1: The fire incidences for 2005 Fire Season for Chimanimani District

A total of 112 incidents were detected in the communal and resettlement areas accounting

for only 16.16% of the total incidents detected in 2005. Nyahode and Bumha

Resettlement areas recorded a total of 89 major fire outbreaks constituting 12.8% of the

total detected fires for 2005.

The most affected areas were Charter Forest Estate, Tilbury Forest, Gwindingwi Forest,

Martin Forest, Tarka Forest and Cashel were a total of 581 of the 693 fires were detected.

This constituted 83.84% of the detected fires throughout the district for 2005.

48
 Chimanimani 2006 Fire Incidents
32°19'58" 32°29'57" 32°39'56" 32°49'55" 32°59'54" 33°9'53"

# ³
³ #
CAS HE L

#³
³# ##
³³
19°30'09" ##
³ #³
³ # 19°30'09"
MH A N DA R U M E ³
#
#³
³
C H IR A M BA
G UU N E TIL BU R Y
#³
³ #
#³
³
##³
³ #
#³
³#
CH A K O W A
#
³ #³
³ #
#
³
#
#³
³
CH A Y A M ITI
#
³
19°40'08" #
³ B UM H A (S H IN JA R S ) #
³ #³
³ #
# ³
#
19°40'08"
#
³ ³#
³ #
#
RU P IS E CH IK U K W³
³
A
#
³
#³
³#
#
³
S HIN JA M AR T IN E ST A T E
#
#³
³ #³
#³
³ # #
³
#³
³
##
NY A N Y A D ZI
#³
³ #³
##
³
³
#³
³# #³
K UD ZA NG A MH A K W E B IRIR I
# ³##³
³#
19°50'07" ³ # 19°50'07"
# #³
³³#
#³
³
#³
³ # # ³
³ #
CH IK W A K W A #³
# GW IN DIN GW I ³ Chimanim ani 2006 Incidents
#³
³# #
³ Chimanim ani W ard Boundaries
CH A N G A ZI
#CH A R TE R
#³
³ E S T A TE #³
³ #³
#
#³
³ #
#³ #
³
NY A H O N D E (N A G O DI R S )
# #³###³
#
³
³ #³
##
³
#
#³
##³
³ #
#³
#³
³ ## ³ #³#
³³
#
³
#
³#³
³# ³
³ #
#
³
³³
#³ #
³#
³#
³
#³
³
#
³³# #
³
#³³# #
³#³ # #³
# ³
#³
³##³
# ³
³ #³
# ³
#³
³
#
20°00'06" #
³
#³
# ³# ³# #
³ 20°00'06"
#³
³ CH A B IK A NG O RIM A B
NG O RIM A A
#
³
N
#T IN
MA R³

32°19'58" 32°29'57" 32°39'56" 32°49'55" 32°59'54" 33°9'53"

10 0 10 20 Kilometers By
Gumisai Chigurah.
023 896710/0912833531

Fig 4.2:Chimanimani District Map showing Locations Of 2006 Fire Incidents.

In 2006, Shinja, Chayamiti, Bumha, Manyuseni, Chabika, Ngorima B, Nyahode and

Chikukwa recorded 2; 1; 3; 1; 13; 20 and 6 fire outbreaks respectively for the whole of

2006. Rupise, Kudyanda, Nyanyadzi, Chakohwa, Chiramba, Guune, Changazi,

Chikwakwa, Mhakwe, Biriri, Ngorima A and Mhandarume did not experience any

detectable fires in 2006 as shown in Fig 4.2. The communal and resettlement areas had a

total of 37 fire outbreaks with Nyahode Resettlement area recorded the highest detection

of 20 incidents followed by Chikukwa ward, which had 6 incidents. The incidents that

were detected in the communal and resettlement areas constituted 24.83% while the

remaining 75.168% were detected in exotic forest plantations.

49
Makoni District.

Makoni 2005 Fire Incidents bb

b Makoni 2006 Fire Incidents
b
TS AGUR
bb A TS AGUR A

b b b
b b
bb b b b bbbbb bb
bb
b b b bbb bb CHITANGA ZU VA
bb b b b b CHITANGA ZU VA b
b b b b bb
b b bb b
b bbb b b b
bb b b bb b b
b bbb bb
b b
bb b bb BATAN AI b
bb BATAN AI GW EZA bb
GWbEZA b
b b bb b b b
bb bbb bb b bbbb
b bb b
bb b b b b bbbbb b b
b
b RONGW E bbb
b bb RONGW bE b
b b bbbbbbbbbbb b b bbbb bb
bbb b bbb bb bb b bbbb
bbb bbbb bbb b bb bb bb bbbbb b bb bb b b b
b
b bbbb bbb bbb b b bb bbb bb bb bb bb bbb b
b bb b b bbb bb bbbbb b
b
b bb b b bbb bbb
bbb bb bb b bb b b bb b b b bb b b
b b bbbbb b bb
b
bbb b bbbb b b
bb b
bbb b bb b bb b bb b bbb b
bbb b b bbbbbbb b b bbb b bb bb
bb
b bbbbbbbb bbbbbbbbbbb b FARMS b b bb b bb bb b bb
b bbbbbb bb b bb b
b b
bb b b bb b
bbbbbb bb bb b b b bb b b b
b b b b bbbbb b b b b b bbbbbbbbb b b bb
bb b
bb b b b bb bb b b bb b b bbb bbb bb bb bbb
bb bb
b b b b
b b b bbbb bbbbb bb b
b b b b
bb b
bbbb bb b b b
bb FARMS b
b bbb bbbb b
bb
b bb bb b b bb bb b b bb bbb bbbbbbb bbbb b bbb
b b b b bbb
bb bb b bb b b bbbb bb bbbbb b b bbb b b b b bb bb b b bb b b
bb b bbbb b b bbb b b bb b bbb b bbb
b bbbb bb b b b
bbbbb b bb
b bb bb b
b bbbbbbb bb bb b bb bb b b bbbb bbb DUMBA MW E
b bbb b bb bb bb b bb bbb bb b b DUMBA b MW E b bb b b bb bb b b b
b b bb b
b b b bbbb bbb b b bb DOWb Ab b b b b b bb b b b
b bbb bbbb bbb bb b bbbb b b b b bb b b
b DOW b bbbAbbb bbb bbb b b bb b b b
b
bb bb bbb MA TOTW E
b
bbbb bb b bbb b bbbbb b bb bb b b bb
bbb b b
b bbb bb b MA TOTW E b MHE ZI b
b b b
MHE ZI bbb bbb b bbb bbb NYAMA GURA
b b bb
bb b b b b BEMB ERO
BEMB ERO b b bb bb bbb bb b b NYAMA GURA bbbbbbb b
b b bb b
b b bbb bbbbb b bbbb b bb b b b b b bb ZUR URA
b TIK Wb bIRI bbbb bb
b bb
b b
bb b ZUR URA
b
bbbb b bbbb bbb b
bb b b b RUOMBW E bbb b b b
b
NGOW E b
b bbbbb bb bbb b b bb bbb bbbbbb
RUOMBW b bb bE b bbb bb CHINY AMAHU MB A
bbb bb CHIDUK U bb
bbb
CHIDUK U b
bbbb b DENZV A
DENZV A b
bbb bb NYAH ANGA RE
NYAH ANGA RE bb
b
NYAMIDZI
b Mako ni_2005_fi res in ci dents b
Mako ni Ward s

b
b
b b 60 0 60 Kilom eters

Fig 4.3: Map of Makoni District showing location of 2005 and 2006 fire incidents.

Fig 4.3 above shows the location of different fire incidents that were detected in the

Makoni District in 2005 and 2006. In 2005 the district recorded 743 detections of major

fire outbreaks and in 2006 it recorded 341 detections of fire outbreaks. In 2005, 47

incidents were detected in Gweza, Rongwe, Tsagura, Dowa, Ngome, Nyamagura,

Matotwe, Dumbamamwe, Batanai and Chitangazuva and the remaining 696 were

detected in the surrounding farms. Below is a table showing the fire incidents that were

detected in the communal areas and the surrounding farms of Makoni District in 2005.

50
WARD NUMBER OF INCIDENTS DETECTED
Batanai 4
Chitangazura 6
Dowa 15
Dumbamwe 3
Gweza 4
Matotwe 2
Nyamagura 1
Ngome 3
Rongwe 4
Tsagura 8
Surrounding Farms 696
TOTAL 743
Table 4.1: Number of Fire Incidents Detected per ward.

Batanai recorded 4 outbreaks; Chitangazura 6; Dumbamwe 3; Gweza 4; Matotwe 2;

Nyamagura 1; Ngome 3; Rongwe 4; Dowa 15 and Tsagura 8 as indicated in Table 4.1.

The incidents that were detected in the communal areas accounted for 6.33% of the total

incidents detected in Makoni District in 2005, the remaining 93.67% were detected in the

surrounding farms. Chinyamahumba, Zurura, Nyamidzi, Denzva, Nyahangare, Chiduku

and Ruombwe did not experience any major fire outbreaks in 2005.

51
Fig 4.4: Fire incidents per ward for Makoni District 2005 veldfire season.

In 2006, 341 incidents were detected compared to 743 incidents that were detected in

2005. There is a 54.01% disparity from last year‟s outbreaks. A total of 330 were

detected in farms while the remaining 11 incidents were detected in Rongwe, Dowa and

Ngome as shown in the Table 4.2 and Fig 4.6

WARD NUMBER OF FIRE INCIDENTS

Dowa 9
Ngome 1
Rongwe 1
Farms 330
TOTAL 341
Table 4.2 Fire incidents per ward for Makoni District 2006.
Dowa ward had the highest fire incidents that were detected in communal areas. It

recorded a total of 9 detections of major fires followed by Ngome and Rongwe, which

had 1 each. The communal areas accounted for 3.23% of the total incidents detected,

52
Dowa alone had 2.64%. The farms had the prevalent detected incidents. Farms

constituted 96.77% of the total detected incidents.

Fig 4.5: Makoni Fire Incidents per Ward for 2006.

53
4.1.2 Number of fire incidents for the 2005/2006 fire seasons and The Monthly
Spatial Distribution.

Fig 4.5 illustrates the trends of fire incidents that occurred in Chimanimani and

Makoni for 2005 and 2007.

Fig 4.6: Chimanimani and Makoni Fire Incidents for 2005/2006.

In 2005 Chimanimani District recorded a total of 693 fire incidents through out the

district with the months of September and October recording the highest number of

incidents with a total of 221 incidents respectively. In 2006 Chimanimani district

recorded a 78.5% drop in fire incidents from a total of 693 in 2005 to a total of 149 in

2006. As shown in Fig 4.6 September and October constituted the highest fire outbreaks

of 32% followed by October with a total 25%, July 6%, January and June constituted 2

apiece and February had 1%.

54
 Table 4.3: Monthly Spatial Distribution for Chimanimani and Makoni Districts for 2005-
2006

CHIMANIMANI CHIMANIMANI MAKONI MAKONI

2005 2006 2005 2006
Month Fire % of Month Fire % of Month Fire % of Month Fire % of
Incident Total Incident Total Incident Total Incident Total
s Inciden s Incident s Incident s Incident
ts s s s
January 14 2.02 January 0 0 January 1 0.13 January 0 0
February 5 0.72 February 0 0 February 0 0 February 0 0
March 1 0.14 March 0 0 March 0 0 March 0 0
April 1 0.14 April 0 0 April 6 0.81 April 0 0
May 0 0 May 0 0 May 8 1.08 May 2 0.59
June 17 2.45 June 0 0 June 43 5.79 June 2 0.59
July 41 5.92 July 20 13.42 July 131 17.63 July 38 11.14
August 221 31.89 August 41 27.52 August 290 39.03 August 106 31.09
September 221 31.89 Septembe 46 30.87 Septembe 140 18.84 Septembe 165 48.39
r r r
October 172 24.82 October 42 28.10 October 124 16.69 October 28 8.20
November 0 0 Novembe 0 0 Novembe 0 0 Novembe 0 0
r r r
December 0 0 Decembe 0 0 Decembe 0 0 Decembe 0 0
r r r

TOTAL 693 99.99 149 99.91 743 100 341 100

Chimanimani 2005 Fire Incidents Percentages

1% 0%
0% 2% 0% 2%
6%
25%

32%

32%

Jan Feb Mar Apr May Jun July Aug Sep Oct Nov Dec

Fig 4.7 Chimanimani monthly fire incidents by percentages for 2005.

In 2005 fires started as early as January where there was a total of 14 outbreaks,

which were all detected in Charter Forest Plantation as shown in Fig 4.8. May,

55
 November and December recorded nil fire incidents while the highest outbreaks

were detected in August and September where a total of 221 incidents were

detected apiece. March and April had one outbreak detected each. In August and

September the majority of the fires erupted in the exotic tree plantations of

Charter, Gwindingwi, Martin and Tilbury Forests.

Chimanimani 2005 Spatial Distribution

32°30' 32°45' 33°00' 33°15'
Chim anim ai 2 005 Monthl y Fire Distribution
T January
$
ââ b Mar ch - Apr il
â
# May - J ul y
³
d August
â âd
ddd d â Septem ber - O ctobe r
d â âd â d d Chim anim ani W ar ds
dâd d
d â
â CAS d L VA LL E Y
d HE
19°30' â
MH AN DAR UME â ââ â
d
d
d d 19°30'
d â
â
â
GUU N E TIL BUR Y
d d
CH AKO W A â
d d
d
dd d
d
CH AYAMITI #³
³
#â â
³ â³
##
# BU MH A (SH INJA
³ # dRS)
RU PISE â³
â ââ
d â ³# # dCH IKU
³ N
#³ # d
d ddKW A
d ³
d dd d
ââ dd d
â d ââ âââ d âd â dd âââ
d d d
d d
ââ ³# ³# d â
â d d d ³ #³ # # d
d
â ââ T
$
T
$ âââ ³
SH INJA # ââ ââ â
ââââ
³#³ # â
â d
19°45' âââââ ³ #³ âââ
19°45'
#³ âââ
â â
â
â ââ
ââ â ³##³ #³ â
ââ
â â
ââ
â
âT
$ â â
NYANYADZ I â â â
â
MARd d âd âd$
TTIN
ââ d d
dd d
#³#
b âd # #³
d ââ d d ³
â BIR IR I d d d ââââââ #³
³ #
d³
#
³ â
ââââ ââ d âd d d d d â
KU D ZANGA d â ââ
ââââââ d dd d d
ddd d d ââ ââ âââ ââ d â
MH AKW E d dd d d â â ââ³#
â
â
ââ ââ
ââ
d
dd d â â ââ³ #ââââââ
â â â ddâ
d
â â
dddddâ d d #
³
AKW A ââ â d dd d dd d âd âââ d dd d d
CH IKW
â â ââ
CHA dd RTE
dddd d dd
d d dR d E S TATE ³
dâ
# GW IN d DIN GW I EST ATE
ââ #³
³#³ ³
##â ddd d ddddd ddddddddd d dddd â âd #³
³ #
d d
CH ANG AZI â â âd d ³ #d d d d d âd d ³#
ââ ââ
d d d #³
³ # ââ â
â ddââ â ââ # AG OD I R S)
ââ E ³
dââ
ââââ âââ â â â â
ââ ââ â â
ââ ââââ
âââd
â â NYAHO â âââ ND
â (N
ââ
â ââââ â â
ââ dâ d dddââ
â â
dd â ââ dd d â
â d
dd âd
d â â â d
â ââ ââ â â â d d d d d
âââ â
ââd bb b â ââ â â ââââ ââââ
dâ ââ ââ
ââ â
T$
$ T$T â â â â
ââ â â
â
â
â ââ
³#³
³
#T
$
#
³ T³
#$ #
T³
$ # d³ #âââ â â â â âââ
â ââ
# ââââ ââb
d â ââ³
b â â
ââ
â³# ³# T
$ â
â
â â ââ â
â
â
ââ
â
ââ #
³
â
â â
# b$
T â ââ
20°00' ³
20°00'
T$
$ Tâââ
â ââ â ââ
â NGO RIMA B
ââ
NGO RIMA A
ââ â
â CH ABIKA
MAN YUSEN I

32°30' 32°45' 33°00' 0 33°15' 30 Kilometers

Fig4.8: Monthly spatial distribution by location for Chimanimani 2005.

In August there were 221 fires detected throughout the district, September and October

had a total of 393 fire outbreaks. November and December did not experience any major

outbreaks detected.

56
 Chimanimani 2006 Monthly Fire Outbreaks

32°15' 32°30' 32°45' 33°00'

ÿ Ú

19°30' $$
19°30'
CASHEL$
$
$Ú $$
M H AN D A R U M E
³#³#
GU U N E
TIL BU R Y
C H AK OW A
ÿÿÿÿÿ ÿÿ
$ ³#³³##
ÿÿ
C H AYA M ITI $
R U PISE ÿ BU M H A ( SH IN JA R S ) $ $ $ $$$
ÿ ÿÿ
C H IK U KW A
ÿ
19°45' ÿ
$$ 19°45'
SH IN JA
ÿÿ $
³#³# ³# MARTIN
N YAN Y AD ZI ÚÚ
Ú Ú Ú Ú
BIR IR I Ú
$$ ÚÚÚÚÚ
KU D ZA N GA
ÿ #
ÚÚ³
Chimanimani Monthly fire Outbreaks
³#³#
M H AK W E
³# July
$$ $ August
C H IK W AK W A Ú $$
$$ $ ÿ September
C H AN G AZI
³#³#
CHARTER ESTATE ÿ $$
Ú October
$ÚÚ N YAH O N D E ( N AG OD I R S) Ú
Ú ÿ$$ ÿ Chimanimani Ward Boundaries
³#³#ÿ ÿ Ú Ú$ÿÿ$ÿGW
ÚÿÚÚÚÚÿIN DIN
ÿ GW I
$ ÿ ÿÚ Ú ÚÚ ÿ ÿ
20°00' ³#
$ÚÚ$Ú ÚÚ
Ú ÿÿÿ ÿ ÿÿ ÿ
Ú ÚÚÚ 20°00'
$ C$H AB IK A ÿ
³# ³#³#
$
N GO R IM A A
N
M AR TIN ÿ

32°15' 32°30' 32°45' 33°00' By:

Chigurah. G.C
10 0 10 20 Kilometers
023896710/0912833531

1.

Fig 4.9: Monthly fire incidents for Chimanimani Districts by location in 2006.
In 2006 the fire season started late with the first outbreaks of a major fires

detected in July were 20 outbreaks were detected. In 2006 major outbreaks were

only detected in July with 20, August 41, September 46 and October 42, with the

highest outbreaks occurring in September were 46 incidents were detected. There

were no major fire outbreaks from January – June and from November –

December as shown in Fig 4.9.

57
 Chim anim ani 2006 Fire Incidents Percentages

0%
13%

28%

28%

31%

Jan Feb M ar Apr M ay Jun July Aug Sep Oct Nov Dec

Fig 4.10: Monthly spatial distribution for Chimanimani 2006 by percentages.

58
 31°30' 32°00' 32°30' 33°00'

<
<
Makoni 2005 Monthly Spatial Distribution b< <
TSAGURA
#<
þ

< #
þ
Z
$ <
bb #þ
# << b b
b #
þ
þ#
þ
b b $
Z <
b Z
$ <
< $ <<
CHITANGAZUVA
bb b b Z b Z
$
b
Z
$ Z
$
< ÚÚ
b <
< < << b b
<
<< <
b Z<
$
$
Z
18°00' #
#þ
þ
<
BATANAI 18°00'
GWþ#EZA Z
$
< b bb $
Z b Z
$
bb bb
#
þ b
< < Z$
$ Zb b <
Z$
$
Z
$ Z Zb b
$
<
Z
$ b b
Z b
$ b bb bb$
Zb
b RONGW E <
b
Z$
$ Z
$Z
$ b$
Z$ Z
Z
$ Z Z
$
# # #
# Ú < <
#
þ# þ
þ #þZþ
$ þ
#þþ
bb #
þ
<
Z
$ <
b$ Z$ b bb < b bb< Z bb b $ Z
<Z< b b$Z #
þ
< << $ Z$ Z$
$ Zb $ Z
$
b $Z
b
b < < b b Z$
$ Z b < # $
þ ZZ #
þ
bbb #
þ Zb b
$
Ú
ÚÚ < < <
< þ #þ # Z$
$ Z
< <$ #
þ þ# ##
þþ# Z$
$ Z
$
b bb ZZ þ Z
b $ Z$ b ÚÚ << <<b b b # <<
bb Z$
$Z$Z b$
$ Z <b b << þ
b # <
#þ
#þ
þ #
þ
þ#
#
þ
#
þ
þ
#
þ
#þ
#þ
# þ
#þ
#$
#$
Z
#$
Zb
Z
þ#þ# Zb $
Z $
$ #
þ
Z
Zb b
b
< b
b
<
<
<
<
Z$
$ Z$Z
FARMS<b bb < < << <
<<<
þ
<<
þ#þ Z
$
#
þ Ú
# <
< < bbb b b bb
b
b b þ# $ þ#þ# þ ##
þ
þ#
Ú
b Z
$ <
þ
b b b b b b #
þ#þ
#
þ
#
#þ #þ#
# b b$ bþ
þ #þ þ ÚÚ Ú ³b# bb < Z
$ ZZ
b$b <<
<#³< <ÚÚÚ # ³bb < # ³
³b bb < < b bbb b bþ
#
b #
#þ
þ#þ
# þ
#
#Ú
ÚÚ
b
bb b $ Z þ
þ
#
þ# #
þ
þ# b bbb b bb bbbb bbb
< b
b b b $ Z$Z b b b b bb b
b < b< þ# #Ì b Ú Ú b bbbb b # b
b bb b< $ þ#Z
þ#
þ# þ b bb bb b $ Z þ
b bb
Z$Zbb bb$ Z
$ Ú
bÚ bb b # <
<
b <
< < #
þ <b
<#
b bb b
b b#
Z
$
Z
$ #
# þ bþ
b #þ ##
þ
þ
< DUMBAMW E b
#
þ ³
b b b þ #þ#þþ # $
þ#þ
b þ # # $
þ b
Z$Z <
b Ú
bÚ Z b b<bb b b # Z Z b<< #
18°30' DOW A
b$
18°30'
þ Z
$ þ b
ZZ b þ
$ # Ú$ < bb bþ # # Z þ
$ # þ# Z
$
þ#
Ú < $Z$ b b bþ #$Zb Úþ #þ bb
Z bb
$
ZZ
$
$$ b bÚÚ b Úþ
b bb$
Z # < #$
Z Ú b
b b$
Úþ #
#þ
þ#þ#
þÌ Z
Z$
$ Z
Z
$ Z
$ b b b Z
$ Z
$ MATOTW E
MHEZI Z
b < bb b <
bb bb bbb# $
Ì ZZ
ÚÚ
BEMBERO bÌ b b þþ # $ NYAMAGURA
Z
$ Ì$ZÌ b b Ú
< b $ Z$ Z< Z
$
Z
$ #þ
þ
#
# $
#$ ZbÚ $ Z< $ Z
$< $Z
#þ Ìþ Z
$Z Z
Z
$ TIKW IRI þ b b ÚÚ
b $ Z$Z
# b
bb$ Z
b ZURURA Makoni_2005_fires incidents
b b þ b
NGOW E < bb Zb b<<
$ bb $ January
<< Ì < CHINYAMAHUMBA #
RUOMBW E þ# þ
Ú
#
b
Z þ
$
Z
$
#Ú Z
$ ³ February - April
Z$
$ Z
bb b b
b Ì May
b June
CHIDUKU Ú Ú
Z$
$Z #
þ July
$$
ZZ b
DENZVA b b August
b
Z
$ Septem ber
Z< #
$
NYAHANGARE#þ
þ #
#þ
< October
þ
Makoni Wards
NYAMIDZI
19°00' 19°00'
Ú

<

b Z
$

31°30' 32°00' 32°30' 33°00' N

0 70 Kilometers

Fig 4.11: Makoni monthly spatial distribution for 2005.

Makoni district recorded the highest number of wild fire outbreaks in 2005 with a total of

743 outbreaks being detected with the highest of 290 recorded in August. The fire season

in Makoni district for 2005 started earlier than in 2006. In 2005 the veldfire season

started in January although no incidents were recorded between February and March.

Only 1 incident was recorded in January, 6 in April, 8 in May, 43 in June, 131 in July,

290 in August, 140 in September and 124 in October as illustrated in Fig 4.11.

59
 Makoni 2005 Fire Incidents Percentages

1% 1%
6%
17%

18%

19%

38%

Jan Feb Mar Apr May Jun July Aug Sep Oct Nov Dec

Fig 4.12 Monthly spatial distribution by percentages.

In Fig 4.12, the month of August 2005 accounted for 38% of the total fire

incidents detected in the whole district. August had the highest detections of fire

incidents followed by September, which had 19%, July with 18%, October with

17%, June 6%, April and May with 1% apiece. February, March, November and

December did not experience any major detectable fire outbreaks hence they all

recorded a nil percentage. From fig…the lowest outbreaks were erupted in April

and May, which shared 1% apiece.

In 2006 there was a remarkable improvement in fire outbreaks for Makoni

district. The total number of outbreaks dropped from 743 outbreaks in 2005 to

341 outbreaks in 2006 due to the late start of the fire season in 2006.

60
 Makoni 2006 Monthly Spatial Distribution
31°45' 32°00' 32°15' 32°30'

TSAGURA

17°45' d 17°45'
$
Z
Z
$ CHITANGAZUVA
â â ââ
â â Z
â
â $
<<
$$
ZZ
â Ú
ââ BATANAI
18°00' GWEZA $$
ZZ 18°00'
â ââ$
Z$
Z Ú
ÚÚ Ú
â
â
Ú
RONGWE $$
ZZ
Z
$
â ââ <
Ú <Ú
Ú â
ââ
â ÚÚ
ÚÚ
Z$
$ZÚÚ
<ÚÚÚ < ÚÚ âââ ÚÚ
Ú
< <
ÚÚ <<< Ú
â ÚÚ
Ú <<<
â ÚÚÚ â d
â< Ú Ú
Ú ÚÚ
18°15' ââ
18°15'
â
ââ â Ú
â Ú $
< Z
ââââ ââ â âÚâ
ââ
< â
ÚÚÚ Ú ââ â Z
$Z
$ â ÚÚ Ú
â ââ <<
Ú
âââ < ÚÚÚÚÚ ÚÚ ââ
Úâ â
#â ÚÚ
ÚÚ
Ú

â
Ú
âZ ââ<< Ú
< <
Ú
â < <<< â
³
â
ââ
FARMSÚÚâ
ÚÚ

â$
â ÚÚ â â
ââÚÚ ââ Ú
â â $
â Z
Z
$ ÚÚ Ú
Ú
<<
Ú Z$
$ Z
ââ <Ú
<â
ÚâÚ
â â<ÚÚ <
DUMBAMWE
â
â âââ
â Ú
ââ Ú
â â
â â âââ
â
DOWA ââ ZÚââ
$ â
18°30'
â
ÚÚ #
³
â
â â
MATOTWE 18°30'
MHEZI ââ â â â
ÚÚÚ âÚ NYAMAGURA
â
BEMBERO ÚÚ
Ú Z
$
<<<<Ú
ÚÚÚ ââ ZURURA
âÚ ÚÚ
ââ
ââ
â â Zâ âââ â â
Ú â â
$
Ú
RUOMBWE âââ ÚÚ
ââ â ââââââ â$
ââZ
18°45'CHIDUKU ââ 18°45'
Makoni 2006 fire incidents
DENZVA #
³ May
d June
NYAHANGARE
< July
â
Ú August
â September
19°00' 19°00'
â
Z
$ October
Makoni Wards

31°45' 32°00' 32°15' 32°30' N

0 30 Kilometers

Fig 4.13: Monthly spatial distribution for Makoni 2006.

The year 2006 registered a 54.10% drop in fire outbreaks from 2005. As is indicated in

Fig 4.13 and Table 4.3, no fire outbreaks were recorded from January to April of 2006 in

Makoni, this attributed to a drop in fire incidents for 2006. The highest detection was

recorded in September with a high of 165 outbreaks. July, August and October recording

a total of 38, 106 and 28 respectively. May and June recorded a maximum of 2 apiece.

61
 Makoni 2006 Fire Incidents Percentages

8% 1% 1%
11%

31%

48%

Jan Feb Mar Apr May Jun July Aug Sep Oct Nov Dec

Fig 4.14: Monthly spatial distributions by percentages Makoni 2006.

The month of September recorded 48% almost half of the total fire incidents that erupted

in the district, and they were all detected in farms except for one incident that was

detected in Rongwe district. August had the second highest percentage of 31% of the

total incidents followed by September, which had 11% ,October with 8% and finally May

and June with 1% each as illustrated in Fig 4.14.

62
4.1.3 The relationship between fire incidences and Crop use/Crop Use Intensity
(CUI) and vegetation cover and type.

 Agriculture constitutes the major economic activity in both Chimanimani and Makoni

District. It is practiced under different technologies, which include: dry-land farming,

irrigated agriculture, livestock farming and horticulture. These types of farming are

also determined by the agro-ecological regions (I - V) and the land type in which they

fall under. The Eastern part of Chimanimani district, which lies in agro-ecological

region I and II, is characterized by both indigenous and commercial forest plantations.

The percentage of crop use intensity usually determines the likelihood of fir

occurrence as fire illuminates and spreads rapidly where there is dense vegetation

cover during the fire season.

Relationship between CUI & Fire

Incidences (2006) T $
$ T

CHIMANIMANI DISTRICT T$
$ $
T
T T
$
$
T T$
$ T
T
$
$$
TT

$$
T T$
T$T$
T$T$
$
T T
$
T $$
TT
T
$
T$
$T

T
$
T
$ T$
$ T T
$
T
$ T
$
T
$
T
$ T$
$T
$
T
$$
TT
T
$
$
T
$$
TT T $
$$
T T
$$
T
T$
$ T$
T TT
$
T
$
T
$Chimanimani 2006.shp $$
TT $$
TT$
T$
T
R/ship between CUI & Fires T
$
HB4 T$
$T$
T
T$
$ T
HB5 $$
TT
U3 T
$ $$
TT
$$
TT T
$
U4 T$
$ T T T
$ $$
T
U5
T$
$ T$
T T
$
T$T $$
T T$T$
T
$ T$
TT
T$
$
T$
T
T T$
T $ T
$ T$
$ T $$
TT$T$ $
T
$T$
$T
$
T
$ T$
$ T T$
$T
$$
T
T
$TT
$ T$
$ T
T$
$T$ T
$ T$
T$
$ T
N
T $T$
T$T$
T T$
$T T
$
T
$ T $
$ T T
$
TT
$ T$
$
$ T

T
$
0 20 Kilometers
T
$

Fig 4.15: The relationship between CUI and Fire Incidences.

63
The legend on crop use intensity is explained as follows:

 HB3: Hilly, broken (30-49% crop use intensity).

 HB4: Hilly, broken (5-29% crop use intensity).

 HB5: Hilly, broken (0-4 % crop use intensity).

 DI 1: Developed, irrigated (70-100% crop use intensity).

 DI 3: Developed, irrigated (30-49% crop use intensity).

 U3 : Uplands (30-49% crop use intensity).

 U4 : Uplands (5-29% crop use intensity).

 U5 : Uplands (0-4% crop use intensity).

 W : Open water.

Besides illustrating crop use intensity for specific areas the data can also be used to

describe the physical relief of an area. Most of the fires erupted were detected in uplands

(U4) where crop use intensity is between 5-29%. This is because the area is mostly

covered by exotic forest plantations due to its terrain, which is not suitable for crop

growth. The least detected fire incidents were detected in Uplands (U5) where crop use

intensity is between 0-4%. This area occupies the western part of the district and

comprise of communal areas except for Tilbury Forest Plantation to the North-East,

which comprise of Pine and Wattle plantations.

64
 Relationship between CUI & Fire
Incidences (2006).
b
MAKONI DISTRICT.
bb
b b bb
b bbb b bb
bb
b bb b
bb
b bbbb b bbb b
b b bbb
bb bbbbb bb b
b
bb b bb b bbbb
bbbbb b bb bb b b b
b b
b bbbbb bbb bbb
b b
b bbb
bb bbbbbb bb b bbb bbb
bb bbb bb b b b
bbbb
b bbb bb bbbbbbbbbbb bbbb b bbb bbbbb b bb b
bbbb bbb
b b b bbb
bb b b
bb b
bbbb
bbbbb bb bbbb bb b b bbbb bbb b
b b bb b bbb b b
bbb b bb bbbbbbbbb b
b bbbbb b b bb b bb b bbb b b bbb
b
b
bb bb b
b bb b bb
bbb bbbb
N bbbbbbb b
b
b bbbbbbb bb
bbbbbb bbbbb bbbb
b bb b
b b bb bbb bbbbbb
bb

b Makoni 2006 Fire Incidences b

R/ship btwn CUI & Fires
DI1 b
DI3
HB3
HB4
HB5
U3
U4
10 0 10 Kilometers
U5
W

Fig 4.16: The relationship between CUI and Fire Incidences in Makoni District

The majority of the fires in Makoni District were detected in areas in areas were crop use

intensity is between 30-49% (DI 3), this area comprise of well developed irrigation

structures that enable the growing of crops throughout the year and in uplands were crop

use intensity between 30-49%. The least fire incidents were detected in hilly-broken areas

(HB 5) where crop use intensity is between 0-4% meaning that crop production is not

intensive hence ground cover is very low.

65
4.1.4 How effective was the system of detecting and reporting fires?

The year 2006 saw a more passive posture being taken by the Environmental

Management Agency in a bid to curb the tide of destruction resulting from incidences of

veldfires or bushfires. The National Fire Prevention Strategy, which was launched in

Chimanimani, steered in a „new‟ holistic fire prevention and management approach,

which in part saw the collection of fire statistics in the Province from all the seven

districts by the Environmental Management Agency staff based in the districts.

Fig 4.17 clearly illustrate that fire incidents‟ reports from the two districts were vastly

inconspicuous compared to MODIS satellite fire detector. During the fire season all fire

incidents in both districts did not correspond with the data that was detected by MODIS.

For example in the month of September, Makoni and Chimanimani EMA district staff

reported 11 and 2 fire incidents respectively whilst MODIS fire detector captured 165

and 46 fire incidents for both districts respectively. The total incidents recorded by

Chimanimani Officers for 2006 was 12 incidents compared to 149incidents that were

detected by MODIS. The percentage difference is 92%. For Makoni district MODIS

detected 341 incidents while officers recorded a total of 25 incidents for the whole

veldfire season with a percentage difference of 92.7%. This scenario therefore

undoubtedly promotes the use of satellite detectors in fire incidents‟ reporting and

monitoring as it confers accurate information.

66
Fig 4.17: Comparison of Officer’s Fire Reports and MODIS fire detector.

67
4.1.5 Fire Incident Recurrence Interval

#
#
# b # b b#
# #
# # #
#
# # #
b
# ##
# # # # #
# # #
# # #
# # # #
# #
# # #
# # #
#
#
# b b
# #
b # ### # #
##
# # # #
#
# #
CHIMANIMANI 2005\2006 #
##
#
#
#
#
#
#
#
#
# #
## # ##
# #

#
###
2Km RING BUFFERING #
# #
# # #
#
#
# ## # #
#
# b
#

#
#

b b b #
#

# #
#
# # ##
#
#
#
# b
#
# b b b
#
#
# ##
#
b b #
# #
##
#
#
b## # # #
#
#
#
#
# ## #
# # # # # #
# #
# # #
# # #### # # #
b Chimanimani 2 006 Fire Inci de nces # b # #
# ##
#
#
#
Chimanimani w a rd boundarie s ## b # #
# # #
# #
# #
# # #
# b b bb b
# # ##
# ## # # # ## # #
# Chimanimani 2 005 fire Incide nc es #
#
# #
b #
# #
#
#
Buffer 1 of Chimani mani_200 5_fires.shp # # # ## # # b
1km #
#
# #
#
#
#
# #
#
#
b b
# #
# ## # #
2km #
#
#
#
#
#
#
b b
# # # # # # # # # # # #
3km ##
#
#
##
# #
## ## #
4km # # ##
# # #
5km # # b #b #
## #
# # #
#
#
#

# # # #
# #
####
# b #
# #
# # # #
# # # #
# # # #
# # ## b b # #
# # # #
b #
b
# #
##
#
b #
#
## # #
#
#
#
b b # #
# b # #
#
#

#
#
#
#
#
b #b #
# # ##
# ## # # # # #
## # ## # # #
# # ## # # # # ## #
## # # #
#
# # # ## #
b# ## ## #
#
# # #
# # # # #
# # ### b# # # #
## #

N #
#
#

#
#
#
#
#
#
#
## #
## #
b bb b b b bb
b
#
bb
#
# #

b bb #b # # # b# b bb
# # #

b b bb b
# #
# #
#
# ### b ## b
##
#
#
# b b#b # # bb
# # #
#
# # # # ## ##
b b ## #
b b # #
b b
#
# #
# # # #
# # # ##
bb b # #
###
## ## # # b b# b #
## # # # # # #
# ## #
# #b #
# # ## # #
# # #
b# b # ## # #
# b b b b
#
#
## #
#
# #
##
b #
# # # #
# # ###
b # # #
# ## ## #
b # b
# # # b # # #
b bb b # # # ##
# #
#

#
# b
#

0 10 Kilometers

Fig 4.18: Chimanimani fire recurrence interval.

The location of the 2006 incidents is not very different from 2005 fire detections.

However there was a slight change in the recurrence interval as 93.96% of the 2006

detected fire incidents were detected within the range 1-2km buffer zone particularly in

the south-eastern part of the district. 6.04% incidents occurred within the 3-4km buffer

zone in the North-Eastern part of the District as shown in Fig 4.18.

68
 Recurrence Interval Makoni 2006

'
' ' ' ' ' '
' '
' '
' # '
' ' '
' ' # # # ' ' ' '
' # # # # # # '
# ''
' ' '
' # # '
' ' '

' ' '' # ' '

'
#
'
# '
# '
#
'

# #
'
# #
# '
'
#
# #
'
#' ' #''
' ' ' '
' ' ' '

## ##
' '
' ' ' ' '
# # '#
'
#
#'# #'
' '
# # # # ' '
' # # # # ' '' '' ' ' #
'
# # # #
' ' ' ' '

'
# ' ' ' ' '

'
' ' '
'
' '
' '
'
## ' ' # #
' ' '
' '
## # ' #' #
' '
' ' # ' ' ' '
' #
# #
' '
# # # # '
# '
# '
# # #
# '#
' '
' '
' ' ' # ## '
' ' ' '
' '

' ' ' ' ' '

' # ' '
' ' #
# # '
# '
# ' '' ' '' '
'
'
' ' '
#
'
# '
##
' '
'
# # #
# ' ##
' ' ' ' '
'
##
' ' '
'
#
'
'
'
# '
' ' '
' '
# ## ' '
' ' '
' # '
'
# # # # #' ' ' # #
' ' '
' ' ' '
' ' ' ' '
'
' '
'
' '
'
''
' #' # '
'
' '
' ' # # ' '
'
'
# ' ' '
# # # ' #'
# #'
# #
' '
'
'
'' '
' '
#' # # #
'
# ' '
'
'
'
'
'
'
# ' ##
' ' ' ' '
'
'
'
# ' ' ' '
'
' ' '
'
' '
'
' # ' ' '
# #
' ' ##
# ' ' '
' ' # #
'
#
# ' ' ' '
# ' '
''
# # ' '
# #
' ' '
' '
'
' '
' '
# # ' '

' '
' ' ' '
'
' '
' '
' ' '
'
' ' ' ' '
'
' # '
'
'
#'##'# #
' ' ' '
'

Mak o ni_ 2 0 05 fi re inc id e nc es

'

# Mak o ni 20 0 6 fire s in ci de nc e s
Mak o ni Wa rd Bo un da rie s N
Bu ffe r 1 of Ma ko ni_ 2 0 05 _fire s in ci de nce s
1k m
2k m
3k m
4k m
5k m
10 0 10 20 Kilometers

Fig 4.19: Makoni Recurrence Interval

The recurrence interval for Makoni district as shown in Fig 4.19 is not very

different from the Chimanimani District because the majority of the detected fire

incidents erupted within the 1-2km buffer zone.The 3-4km buffer zone from the

previous year‟s incident had 9.4% of the 2006 detected fire incidents whilst the

69
 remaining 90.6% were within the 1-2km buffer zone from the previous year‟s

outbreak.In Makoni most outbreaks were recorded in resettlement farms as

indicated in Fig 3.1 with only a few outbreaks in communal lands.

4.2 OFFICERS’ BACKGROUND KNOWLEDGE ON REMOTE SENSING

INSTRUMENTS AND THEIR APPLICATIONS IN VELDFIRE
MONITORING.
The responses obtained from officers‟ of Chimanimani and Makoni District

during a questionnaire is tabulated in Table 4.4.

Officers with background Officers without background Total

knowledge on Remote Sensing knowledge on Remote Sensing
6 2 8
Table 4.4: Officers’ response on background knowledge on Remote Sensing.

70
 Officers' Background Knowledge on Remote Sensing

4
Officers with background knowledge on Remote
Sensing
Number of Officers 3 Officers without background knowledge on Remote
Sensing

0
1
Officers' Response

Fig 4.20: Responses by officers on background knowledge on remote sensing.

Of the 8 officers in the two districts 6 of them have some background knowledge

on Remote Sensing instruments used in veldfire detection and 2 officers‟ had no

background knowledge on Remote Sensing as illustrated by the graph in Fig 4.20.

. 4.2.1 Officers with knowledge on the Applications of Remote Sensing in Veldfire

Monitoring

Officers‟ with knowledge Officers‟ without TOTAL

on Remote Sensing knowledge on
Applications Applications of Remote
Sensing
5 3 8
Table 4.5: Officers’ knowledge on the applications of remote sensing in veldfires.

A total of 38% of the 8 officers knew how to apply and integrate Remote Sensing

techniques with GIS in and the remaining 62% expressed ignorance on its

71
 applications as shown in Fig 4.22. The 38% said they learnt how to use the GIS

software and remote sensing techniques whilst there were still at college.

Officers' Responses on the Applications of Remote Sensing in Veldfire Monitoring

38%

62%

Officers’ with knowledge on Remote Sensing Applications Officers’ without knowledge on Applications of Remote Sensing

Fig 4.22: Percentages of Officers’ response on the applicability of remote sensing

in veldfire.

4.2.2 Preferred Methods used in Collecting Statistics on Fire Incidents

Remote Ground Public Both TOTAL

Sensing Truthing Domain
2 1 1 4 8
Table 4.6: Methods preferred by officers in veldfire data collection and analysis.

Given that the officers have the necessary technology, appropriate GIS software,

25% preferred to use remote sensing given the vehicle and fuel crisis that is being

experienced in the two districts. Remote Sensing would reduce and ease the

transport and fuel costs and increase efficiency in data collection, evaluation and

monitoring. 12.5% preferred to use data from the public and ground truthing and

72
 the remaining 50% preferred to use both methods to increase the validity and

accuracy of the statistics collected. Remote Sensing is very useful in detecting

fires in remote areas that are inaccessible by vehicles.

Preferred Methods In Veldfire Data Collection

25%

49%

13%

13%

Remote Sensing Ground Truthing Public Domain Both

Fig 4.23: Percentages of officers’ response in relation to methods they prefer to

use in veldfire data collection.

4.2.3 Constraints Faced in Collecting Data and Statistics on Veldfires.

All the officers echoed the same sentiments on the problems of transport

availability in carrying out their routine environmental inspections. The province

was allocated only three vehicles, which are based at the provincial office, but

two of the three circulate within the seven districts if fuel is available.

Another problem that was mentioned is the lack of appropriate technology and

GIS software, 75% of the officers expressed concern on the unavailability of

computers to district officers. Public ignorance in reporting veldfire incidents is

73
another problem that was mentioned that is hindering the efficiency monitoring of

veldfires and natural resources as a whole.

Another problem mentioned by all the officers of the two districts is the lack of

refresher courses on new technology, methods and literature on veldfire

monitoring and collection of veldfire statistics. They pointed out that refresher

courses were necessary to keep them up to date with recent developments in

veldfire monitoring and the management of natural resources.

74
4.3 DISCUSSION.
Chimanimani and Makoni Districts experienced high detections of incidents in

August and September of both 2005 and 2006 because during this time the districts

would be experiencing warm dry season which stretches from mid-August to

November. This is a warm dry season because Zimbabwe experiences a continental

type of climate, which is divided into four seasons namely:

 The rainy season (November to mid/late March)

 The post rainy season (mid/late March to mid May)

 The cool dry season (mid May to mid August)

 The warm dry season (mid August to November).

During the warm dry season temperatures in the Chimanimani and Makoni districts

will be averaging 18 oC . Maximum temperatures in August will be around 18oC

and in September maximum temperatures will be around 21oC. Temperature

influences fire behaviour by the drying and preheating effect it has on fuels which

are described by whether they are dead or alive and by their position under ground,

on the surface or in the air. Surface fuels in the forest plantations that are found in

the Chimanimani District raise crown fuels to ignition temperatures through

convective heat transfer. Examples of dead surface fuels include dead grass, shrubs

and woody debris. Dead aerial fuels consist primarily of dead trees and dead

branches in live trees and examples of live surface fuels are grass and shrubs.

The winter rainfall in the exotic forest plantations of the Chimanimani District and

resettled farms of the Makoni District stimulated the growth of non-native annual

grasses that filled interspaces between trees with continuous fine fuels and this

75
facilitated an increase in the spread of veldfires in both districts in August and

September. In Chimanimani District 2005, August and September constituted

31.89% apiece and in 2006 it was 27.52% and 30.87% respectively while for

Makoni District the same months had 39.03% and 18.84% respectively for 2005 and

for 2006, they constituted 31.09% and 48.39% respectively.

The topography of the majority of forest plantations in the Chimanimani District is

extremely rugged, with ranges of jagged peaks and deep ravines. The main plateau

is at an altitude of 1,500–1,800 m, with peaks reaching 2,400 m and dropping to

320 m in deep gorges and river valleys. Topography influences the movement of air

directing a fire‟s course. Fires burn faster uphill than downhill because of the

preheating of the uphill fuels and the influence of daytime upslope winds. This is

the reason why 83.84% and 75.17% of the total fire incidents detected in 2005 and

2006 respectively were detected in forest plantations of the Chimanimani District

with a crop use intensity classification of (U4) meaning that they are uplands with

5-29% crop use intensity.

Communal areas experienced low fire outbreaks due to the availability of natural

fireguards created by animal and human tracks, settlement pattern and cultivation.

The communal areas are fragmented by settlements, cultivation and well developed

human and animal tracks making less fuel being available for a fire and preventing

the spread of a fire from one point to another. These reduce the chances for large hot

76
 late season fires. The communal areas are also heavily grazed by domestic and wild

animals, thereby reducing the amount of fuel available for a fire.

The recurrence interval for both districts were within 2-4km vicinity and the overall

reduction of fire incidents in 2006 was mainly a result of the vast outbreaks being

experienced in the previous fire season of 2005 which consumed huge amounts of

fuels thereby reducing the amount of fuels available for the 2006 veldfire season.

The discrepancy in reporting of fire incidents by EMA staff was a result of:

 Inaccessibility. Some areas were fire outbreaks were captured by MODIS are

inaccessible to people which makes it difficult to recognize and record such

incidents.

 Immobility. EMA staff could not swiftly identify fire outbreaks in all wards of the

district because of immobility resulting from erratic supply of fuel and availability

of motor vehicles. This made it intricate to carry regular inspections in all the

wards.

 Disinclination by communities to report any fire outbreak to EMA offices made it

virtually impossible to keep update records of fire incidences in all the districts.

4.4 Summary.
This chapter has generally presented and analyzed data obtained from research.

With such an analysis this chapter has addressed the objectives of the study

making sure that the research questions are tackled. In brief it has shown that the

77
public and domain data alone is not very reliable in monitoring fire incidents.

There is need to integrate public domain data, remote sensing and geographical

information systems in order to obtain reliable results in veldfire monitoring.

78
 Chapter 5

Conclusions and Recommendations.

5.1 Introduction.
This chapter gives the conclusions and recommendations taking into

consideration the analysis that was done in chapter 4. It deals with proposed

solutions in the form of recommendations in as far as the implementation of

remote sensing in mapping veldfire incidents is concerned

5.2 Conclusions.
A total of 1 436 fire incidents were detected in Chimanimani and Makoni districts

in the 2005 veldfire season, of the 1 436 incidents 159 were detected in communal

areas whilst the remaining were either detected in farms or exotic forest

plantations. In 2006 there was a marked difference in the number of incidents that

that were detected. The two districts recorded a total of 490 incidents with a

percentage difference of 65.9%. Of the 490 incidents detected 48 incidents were

detected in communal areas.

There is a huge difference between 2005 and 2006 veldfire season due to the fact

that 2005 had the highest fire outbreaks, which consumed most of the burning

fuel, resulting in the decrease of fire incidents in 2006. Communal areas

experienced very few outbreaks due to their settlement set up and the presence of

domestic animal, which feed on dead and live grass, shrubs and tree leaves

thereby reducing the fuel capacity that cause veldfires.

79
The majority of the fires were detected in August and September with both districts

recording 511 and 361 incidents in August and September 2005 respectively. In 2006

there was decrease in the same months in the occurrence of fire incidents. In 2006,

there were 146 and 205 incidents for the same districts in August and September

respectively. January to June did experience less to no outbreaks except for

Chimanimani 2005, which experienced 14 outbreaks in January and 5 in February.

Fire outbreaks are rampant during August and September because the districts will be

experiencing high temperatures and no rainfall, providing suitable conditions for fire

outbreaks because there will be less to no humidity in the burning fuels increasing the

risk of ignition. Few incidents were detected in January to May because of the

wetness that will be experienced during the rainy season and the low temperatures,

which do not favour fire outbreaks. However if rains are good, this will increase the

amount of fuels that can burn during the dry season.

In Chimanimani 2006, MODIS detected 149 outbreaks while the district officers

reported only 12 incidents. The same applies for Makoni District, where MODIS

detected 341 incidents and district officers recorded 25 outbreaks in the same year.

The discrepancy in officers' reporting was due to the unavailability of transport to

carry out their routine environmental inspections, some areas are inaccessible both by

foot and vehicle and disinclination by communities to report fire outbreaks to nearest

district officers.

80
 All the eight officers in the two districts filled the questionnaires and half them

preferred to use remote sensing together with ground truthing, public domain

information when collecting statistics on veldfire occurrences and natural resources

management. A quarter of the officers preferred to use remote sensing and another

12.5% preferred to use public domain information and ground truthing apiece. The

50% that preferred to use both methods appreciated the benefits of remote sensing in

terms of less financial expenses and high efficiency when compiling veldfire

statistics.

All the officers mentioned transport problems, three-quarters expressed concern on

the lack of necessary technology, and one quarter mentioned the people's ignorance in

reporting veldfires. Another problem mentioned by all the officers was lack of

refresher courses in new developments in technology and methods used in acquiring

environmental data and veldfire monitoring.

5.3 RECOMMENDATIONS

 The research demonstrates the potential use of and superiority of

Geographical Information Systems and Remote Sensing data in combination

with public domain data in veld management in view of the discrepancy of

collected veldfire statistics and the current resource constraints being faced

by the Environmental Management Agency. In view of this it is therefore

imperative to develop a framework for the integration of Geographical

Information Systems, Remote Sensing and Public Domain data in veldfire

81
 assessment. Therefore there is need to utilize earth observation systems and

Geographical Information Systems combined with data from districts and the

public domain data to spatially investigate, understand and map the fire

occurrence and recurrence interval, awareness of fire likelihood, impact of

veldfires on vegetation and identifies focal areas in the years to come.

 Increase awareness campaigns both in communal and urban areas with

emphasis in areas near plantations, conservatives and resettlement areas. The

campaigns should cater for all individuals including the disabled such as the

deaf, blind and mentally unstable.

 Practice early burning. This is usually done between May and July when

maximum temperatures are at their lowest.

 Each district must be allocated at least one vehicle and a motorcycle to

increase mobility.

 Districts should be equipped with computers connected to the internet and

appropriate GIS software to perform spatial analysis on data collected

concerning the environmental management.

 Relations between district officers and the community must be enhanced to

allow exchange of veldfire information.

 Refresher courses must be open to everyone regardless of their position and

their station. This will help in capacitating employees at all levels making it

easier to perform regular duties.

82
 BIBLIOGRAPHY

Andreae, M.O. 1997. Emissions of trace gases and aerosols from southern African
savanna fires. Global Fire Monitoring Center (GFMC), Freiburg, Germany.

Barbosa P.M., Stroppiana, D., Grégoire, J-M., and Pereira, J.M.C. 1999. An assessment
community involvement in fire management. FAO & Project FireFight

Bernhardsen, T. 1996. Geographic Information Systems . Halsted Press.

Chidumayo, E.N. 1997. Miombo Ecology and Management: An introduction.
Stockholm Environmental Institute.

Darlong, V. (2002). Traditional community-based fire management among the Mizo

Shifting cultivators.

DeMers, M.N 1999. Fundamentals of Geographic Information Systems. John Wiley

and Sons

Dennison, P.E., Roberts, D.A., Thorgusen, S.R., Regelbrugge, J.C., Weise, D., and Lee,
C. 2003. Modelling seasonal changes in Live fuel moisture and equivalent water
thickness using accumulative water balance index. Remote sensing of the
Environment, 88:442-452.

Escobar, F., Hunter, G., Bishop, I and Zerger, A. 2003. Introduction to Geographic
Information Systems. Department of Geomatics. The University of Melbourne.

FAO (2003) Case Studies of Community Based Fire Management. FAO, Rome.

FAO 2002. Report on Legal Frameworks for Forest Fire Management:

International Agreements

83
FAO, 1999. FAO Meeting on Public Policies Affecting Forest Fires. Forestry Paper
138. ForestryDepartment, FAO, Rome.

FAO, 2001. Global Forest Fire Assessment 1990-2000. Forest Resources

Assessment Programme, Working Paper 55, Forestry Department, FAO, Rome.

FAO, 2004. Report of the Fourteenth Session of the African Forestry and Wildlife
Commission,

FAO, 2005c. Forest Fire Management Working Papers. Global Forest Resource
Assessment 2005 Thematic report on forest fires in the Sub-Sahara Africa (SSA)
Region. Working Paper FFM/9/E, Forestry Department, FAO, Rome. Fire Conference,
October 6-8, Sydney, Australia.

Flynn, L.P., Harris, A.J.L; Rothery, C; Oppenheimer, C. 2000. High- Spatial resolution
remote sensing of active volcanic features using Landsat and hyperspectral data.
AGU Monograph 116: Remote Sensing of Active Volcanism: 161-177.

Fraser, R.H; Cihlar, J. 2000. Hotspot and NDVI differencing synergy (HANDS): A
new technique for burned area mapping over boreal forest. International Journal of
Remote Sensing 74 (3): 362-376.

Frost, P. 1999. Fire in Southern African Woodlands: origins, impacts , effects and
control. In proceedings of an FAO meeting on Public Policies affecting Forest fires.
FAO forestry paper 138: 181-205.

Frost, P. and Vasloo, H. 2006. Satellite-based early warnings of fires:Reducing fire

flasheovers on transmission lines. Meraka Institute.

Frost, P.G.H. and Robertson, F. 1987. The ecological effects of fire in savannas. In:
B.H. Walker

84
Ganz, D., Fisher, R.J. & Moore, P.F. (2003). Further Defining Community-Based Fire
Management: Critical elements and rapid appraisal tools. 3rd International Wildland
Fire Conference, October 6-8 Sydney Australia.

Goldammer, J.G. and de Ronde, C. 2004. Wildland Fire Management Handbook for
Sub-Sahara

Gregoire, J.-M. 2003. An estimate of the area burned in southern Africa during the
2000 dry season. Institute of Planners, Vol 35: 216-224.

Harris, A.J.L., Flynn, L.P; Dean, K; Pilger E; Wooster, M; Okubo, C; Rothery, D. 2000.
Real-time satellite monitoring of volcanic hotspots. AGU Monograph 116: Remote
sensing of Active Volcanism: 139-159

Justice, C.O and Giglio, L. 2002. The MODIS fire products. Remote Sensing of the
Environment83, 244-262.

Li, Z; Khananian, A; Fraser, R.A . 2000a. Detecting smoke from boreal forest using
neural network and threshold approaches applied to AVHRR imagery. IEEE
Transactions in Geoscience and Remote Sensing 39: 362-376.

Liz, Z; Nadon, S; Cilhar, J; Stocks, B. 2000. Satellite based mapping of Canadian

boreal forest fires: Evaluation and comparison of algorithms. International journal of
Remote Sensing 21 (16):3071-3082.

Moriesette, J.T., Privette, J.L and Justice, C.O .2002. A framework for the validation of
MODIS land products. Remote Sensing of the Environment , 83:77-96.

National Centre of Geographic Information systems and Analysis 1990.

85
Pyne, S.J., Andrews, P.L., and Laven, R.D. 1996. Introduction to wildland fire. New
York: Wiley.

Rose-Innes, R. 1971. Fire in West African Vegetation. Proceedings of the Tall Timbers
Fire Ecology Conference 11:147-173.

Schnieder, D.J., Dean, K.G; Dehn, J; Miller, T.P; Kirianov, V.Y. 2000. Monitoring and
analysis of volcanic activity using remote sensing data at the Alaska Volcanic
Observatory. A case study of Kanchatka, Russia;December 1997. AGU Monograph
116: Remote sensing of active volcanism 65-85.

West, O. 1972. Fire, man and wildlife as interacting factors limiting the development
of climax vegetation in Rhodesia. Proceedings of the Tall timbers fire ecology
coference 11:121-145.

Wooster, M.J., and Rothery D.A. 2000. A review of volcano surveillance applications
using the ATSR instrument series. Advances in Environmental Modelling and
Monitoring in review.

Wright, R., Flynn, L.H; Harris,A .and Pilger, E. 2000. Automated volcanic eruption
detection using MODIS, Remote sensing of Environment in review.
WWF-2001. Fire management manual. Southern Africa Regional Programme Office
(SARPO). Harare, Zimbabwe.

86
Appendix A: Questionnaire Designed for the Employees.
Gumisai Charles Chigurah is a student from the Midlands State University doing an
undergraduate honours degree in Geography and Environmental Sciences. I am carrying
out a research on the applications of Remote sensing and public domain data in mapping
and analysing veldfire incidents in Chimanimani and Makoni. Please kindly respond to
the following questions.
1) Sex. M F
2) Educational Qualifications: Degree
Diploma
Certificate
Advanced Level
Ordinary level
3) State any remote sensing instruments you know that are used in veldfire detection.
________________________________________________________________________
________________________________________________________________________
________________________________________________________________________
4) Do you know their applications in veldfire monitoring: Yes No
5)How do you collect your monthly veldfire statistics.
________________________________________________________________________
________________________________________________________________________
________________________________________________________________________
6)What method would you prefer to use in collecting statistics on veldfire incidents:
Remote Sensing Public Domain data

Ground truthing Both

7) In brief what are the constraints being faced in collecting data on veldfire incidents
using the methods you have mentioned in 6.
________________________________________________________________________
________________________________________________________________________
_______________________________________________________________________

87