Deforestation: Change Detection in Forest Cover using ...aulavirtual.ig.conae.gov.ar/moodle/pluginfile.php/513/mod_page... · Deforestation: Change Detection in ... estimated that

Soraya Violini

Seminary – Master in Emergency Early Warning and Response Space Applications.Mario Gulich

Institute, CONAE. Argentina

October, 2013

Deforestation: Change

Detection in Forest Cover

using Remote Sensing

Contents

Abstract……………………………………………………………………………..1

Chapter 1

Introduction….………………………………………………………………..........2

Chapter 2: Forest and Deforestation

2.1 Definitions…..…..……………………………………………………………...3

2.2 Forest and Problem……………...…………………………………………….4

2.3 Lucha contra la Deforestacion……………………………………………......5

Chapter 3: Remote Sensing y Deforestation

3.1 Vegetation Reflectance………….……………………………………..……..11

3.2 Technical of Change Detection….…………………………………………...14

Chapter 4

Conclusions………………………………………………………………………..24

Bibliography………………………………………………………………............25

Deforestation: Change Detection in Forests Cover using Remote Sensing

1

Abstract

The conservation and development of forests are vital to the welfare of

human beings. Forests management is essential to maintain social, economic and ecological

services.

Forrest monitoring allows to track their state of health and productivity, in order to

conduct proper management, according to the state of resources, to enhance their

functionality and promote conservation.

Remote sensors, optical and radar, offer the possibility of locating changes in forest

areas using various analysis techniques, ranging from the purely visual interpretation to the

implementation of a fully automated algorithm.

This report is a review of the literature on the techniques used to observe changes in

forest cover and monitoring through remote sensing.


2

CHAPTER 1

Introduction

Forests provide support for one billion people that live in extreme poverty around

the world, and provide remunerative employment to more than one hundred million. They

contain more than 80% of terrestrial biodiversity (FAO, 2012) and provide essential

environmental services such as soil conservation, watershed management, protection

against floods and landslides, and provide industrial wood (UN).

According to the International Tropical Timber Organization (ITTO) of the UN, it is

estimated that deforestation and forest degradation rise 12.9 million hectares per year and

the current area of degraded forest is 850 million hectares. Most of the changes in forest-

based ecosystems due to: a) conversion of land cover, b) land degradation c) intensification

of land use (Lambin, 1997). These changes have resulted in coverage to a wide variety of

ecological impacts, ranging from local to global scale, including changes in productivity

and forests composition, nutrient dynamics, species diversity, and increased atmospheric

carbon dioxide (Braswell et al 2003).

One way of assessing changes in land use is based on the measurement of changes

in vegetal and no vegetal cover (Bochco, 2001). Technological progress allows a

comprehensive understanding of any region of the earth's surface from satellite images

(Chuvieco et al, 2002). These images of the Earth have been widely used for change

detection, specifically to the mapping and monitoring deforestation. This has favored

international efforts to establish permanent programs (Vargas and Chuvieco, 1991).

This work is intended as a review to the approaches used to monitor deforestation

processes, listing and describing the techniques used to observe changes in forest cover and

monitoring through remote sensing.


3

CHAPTER 2

Forests and Deforestation

2.1. Definitions

Forests

The concepts of Forests and Deforestation were and are defined in different ways

depending on the organization to which we refer. According to FAO (2000) forests and

natural plantation with canopy cover greater than 10% and a surface to 0.5 ha.; determined

by the presence of tree and the absence of other predominant land use, which trees should

be able to reach a minimum height of 5m.

In Argentina, the law 26.331/07 of Minimum Standards for Environmental

Protection of Native Forests, defines forests as natural forest ecosystems composed

predominantly of mature native tree spices, with various species of flora and fauna

associated, along with the surrounding medium - soil, subsoil, atmosphere, climate, water-,

forming an interdependent web with its own characteristics and multiple functions, which

in its natural state, give to the system a dynamic equilibrium condition that provides various

environmental services to society, as well as diverse natural resources with the possibility

of economic use.

Deforestation

The Program Forest Resources Assessment (FRA, 2006) defines deforestation as the

conversion of forests to other land use or cover reduction, less than 10% of their Total (0.5

ha.). Also the United Nations Framework Convention on Climate Change (UNFCCC)

defines deforestation as the direct conversion of human-induced forest land to non-forest

land.


4

2.1. Forests and Problem

The forests of the world cover about 3.4 billion hectares (Figure 1). They are

sources of raw materials and food, and are critical to maintaining agricultural productivity

and ecological balance of the entire planet.

Figure 1: World map of forest distribution.

Forests and woodlands are varied, from the dense jungles of the tropics to the East

African savannah woodlands of mangroves to lush boreal and temperate forests.

In Latin America, tropical forests constitute the largest reserves of this type of

forests worldwide, but there are disappearing at a rate of about 1.3 percent annually.

In Argentina, the richest forested regions in species are the Paranaense Forest and

“Yungas” Forest. By mid 2004 Argentina's native forests spread over approximately

36 million hectares, this means only 15% of the country. Between 1880 and 2003

destroyed approximately 78% of native forest cover throughout Argentina.


5

Figure 2: Phytogeographical regions of Argentina (Cabrera).

The lack of a system for the continuous monitoring of the forest cover, including

forest inventory or geographical information system to periodically update the changes in

land use, prevent obtaining actual figures regarding the disappearance of woodlands. To the

above, are added, controls inefficient and little or no supervision over the damage to the

forest cover by total and selective harvesting, authorized annually; therefore, there are no

updated figures on the extent of commercial forest loss and total deforestation (FAO,

1995).

2.2. Combating Deforestation

Concern about the destruction of the world's forests, has increased considerably in

the past two decades and has led to several initiatives to reverse this trend and develop

strategies and measures for sustainable forest management (ITTO, 2002).

World Forestry Program

In February 1997, the Intergovernmental Panel on Forests (IPF) of the Commission on

Sustainable Development (CSD) defined the global forest program. He reaffirmed that the

conservation and sustainable development of forests are matters of international concern.

La Yunga Bosque de Calden

Selva Misionera

Bosque Chaqueño


6

The Group adopted proposals for action and reach consensus on key issues such as national

forest programs and forest assessments. Here are some of the major issues and

recommended actions were the result of the process of IPF:

National forest programs should be holistic, intersectoral and interactive, and

consistent with the policies and national and local strategies. Should involve all

stakeholders, promote secure land tenure and integrate the conservation and

sustainable use of biological resources.

Indigenous peoples and local communities have traditional rights to be respected.

Working with them is essential to identify, preserve and promote traditional

knowledge related to forests.

The national forestry research capabilities should be improved and create regional

and global networks of research to facilitate information exchange, foster

interdisciplinary research and disseminate the results. Detailed studies are needed of

the underlying causes of deforestation and environmental degradation.

You need to have better evaluation methodologies to obtain reliable estimates of all

forest goods and services, especially those that are not generally traded.

Measures are needed to improve access to markets for forest goods and services,

including the reduction of tariff and non-tariff barriers to trade, in accordance with

existing international obligations and commitments.

It is necessary to adopt innovative methods to make more effective use of existing

financial mechanisms and generate new and additional resources, both nationally

and internationally.

The investment policies and regulations should aim to attract domestic investment

of foreign and local communities for sustainable forest-based industries,

reforestation, conservation and protection of forests. The use of appropriate

economic instruments and incentives and market-based increase income and to

mobilize domestic financial resources.

Cooperation should be encouraged in the transfer of technology related to forests,

through public sector investment and private joint ventures, exchange of

information and a closer relationship between forestry institutions.


7

Improving information systems would improve coordination and data sharing on the

implementation of national forest programs, programming of official development

assistance, the provision of new and additional financial resources, private sector

investment and development and technology transfer.

Should be clarified the roles and mandates of the relevant international

organizations and mechanisms to increase cooperation, eliminate gaps and avoid

duplication (FAO).

International Tropical Timber Agreement

The International Tropical Timber Organization (ITTO) is an intergovernmental

organization that promotes conservation and sustainable management, sustainable use and

trade of tropical forest resources. Its 60 members represent about 80 percent of the world's

tropical forests and accounted for 90 percent of world trade in tropical timber. Documents

develop ITTO internationally agreed policy to promote the conservation and sustainable

management of forests and assists tropical member countries to enable them to adapt such

policies to local circumstances and to implement them through projects.

In addition, ITTO collects, analyzes and disseminates data on the production and

trade of tropical timber and funds a variety of projects and activities for developing

industries at both community and industrial scales. All projects are funded by voluntary

contributions from members, mainly from consumer member. Since it became operational

in 1987, ITTO has funded more than 750 projects, pre-projects and activities with a total

value of over 300 million U.S. dollars. The main donors are the governments of Japan,

Switzerland and the United States of America.

Reducing Emissions from Deforestation and Degradation (REDD)

Deforestation and forest degradation due to agricultural expansion, conversion to

pastureland, infrastructure development, fires, destructive logging, including nearly 20 % of

global emissions of greenhouse gases.

Limiting the impact of climate change within limits that society can tolerate, the

global average temperature must be stabilized within the range of two degrees Celsius


8

above the current temperature. This is almost impossible without reducing emissions from

the forest sector, including mitigation measures.

The initiative for Reducing Emissions from Deforestation and Degradation (REDD,

2008), is an effort to create a financial value for the carbon stored in forests. This provides

incentives to developing countries to reduce emissions from forested lands and invest in

routes of low-carbon sustainable development. “REDD +" goes beyond deforestation and

forest degradation, and includes the role of conservation, sustainable management of forests

and enhancement of forest carbon stocks.

Panama is one of the 9 countries in the world with resources approved by the UN

REDD Program to help design and organize the steps necessary to achieve " readiness" to

allow the country to be ready to implement activities and mechanisms to reduce emissions.

The Joint UN REDD in Panama will contribute to: design a legal framework

validated for the implementation of the national REDD + strategy , develop an operational

framework for the implementation of the REDD + strategy , strengthen national capacities

for the implementation of the REDD + strategy , design system and benefit payments ;

create a national forest inventory and monitoring carbon ; establish a baseline emissions

scenario and design a carbon accounting system and emissions information generation .

The Joint UN- REDD in Panama defines the minimum of preparation, and supports

the country to chart a path towards achieving the implementation of REDD +.

Native Forest Law

In late 2007, Argentinian Congress passed Law 26.331 of Minimum Standards for

Environmental Protection of Native Forests, regulated in February 2009.

The Forest Act provides that the provinces will make the land of its native forests

(OTBN) through a participatory process, categorizes possible uses for forest lands: from

conservation to the possibility of transformation for agriculture, through sustainable forest

use. So forest zoning as follows:

Category I (red): High conservation value areas that must not be removed or used

for timber extraction and forest should remain forever. Include nature reserves and

their surrounding areas, which have outstanding biological values, and / or sites that

protect important watershed (headwaters of rivers and streams).


9

Category II (yellow): areas of high or medium conservation value, which may be

degraded but if it restores can have a high conservation value. These areas cannot be

removed, but may be subject to the following uses: sustainable, tourism, collection

and scientific research.

Category III (green): areas of low conservation value that may become partially or

in full, prior to the completion of environmental impact assessment.

An important aspect that incorporated the standard is the constitution of the

National Fund for the Enrichment and Conservation of Native Forests "in order to

compensate the jurisdictions that preserve native forests for the environmental services they

provide." This compensation mechanism for environmental services, besides being the first

record of its kind in Argentina law, is part of the conception that the land itself becomes

meaningless if not accompanied by active policies to support and promote the use

sustainable forest.

The Forest Act is a tool that must be properly applied to ensure the conservation of

our native forests.


10

CHAPTER 3

Sensors Remotes y Deforestation

3.1. Reflectance de la vegetation

All geographical elements (forests, crops, rivers lakes, buildings, etc.) transformed

differentially receiving electromagnetic radiation from the sun. Each object type represents

a specific type of level in terms of:

Received Radiation = Reflected Radiation + Radiation Absorbed + Radiation Transmitted

The variation of reflectance (reflected radiation) as a function of wavelength

spectral signature called, that is, the function that describes the amount of reflected

radiation with respect to the wavelength of this radiation, the spectral signature is an

objector element (Figure 3).

Figure 3: Spectral signatures and resolution of Landsat multispectral bands

(www.tecnologia.net/teledeteccion/).


11

Vegetation is a resource with many varieties, and different characteristics depending

on the species (leaf, stem, trunk, humidity, etc.), and besides that, for living being, are

subject to change depending on environmental conditions and internal thereof.

Healthy vegetation has a low reflectivity in the visible, but with a peak in the green

color due to chlorophyll and air bags which are generated in the intermediate tissue of

leaves. This reflectivity is very high in the near infrared due to low energy absorption by

the plants in this band. Mid-infrared is a particularly important because decrease in those

wavelengths in which water of the plant absorbs energy.

The ill vegetation has decrease of the reflectivity in the infrared, but an increase in

the red and blue (visible).

An additional factor that affects the vegetation, is the amount of wáter it containes.

When this increases, reflectivity decreases and viceversa (inversely proportional), due to

the behavior of water with respect to radiation. Mid-infrared shows a particularly important

decrease in those wavelengths in which wáter contained in plants absorbs energy (Figure

4).

Figure 4: Reflectivity of vegetation in the spectrum (//tecnologia.net/teledeteccion/).

Vegetation Indexes

Investigators have developed techniques for qualitatively and quantitatively

assessing the vegetation from spectral measurements. Vegetation indexes (VIs) take

advantage of vegetation's reflective contrast (as mentioned before) between the NIR and

visible red (VIS) wavelengths, sometimes with additional channels included; these are one


12

of the most widely used remote sensing measurements. Generally, are analyzed the

reflectance at 660 nm in the red range of the spectrum and the reflectance at 870 nm in the

NIR range. At higher vegetation vigor, more high contrast between reflectance values in

these regions.

Most formulas of vegetation indexes are based on ratios or linear combination and

exploit differences in the reflectance patterns of green vegetation and other objects.

Following a list of the principal Vegetation indexes:

Vegetation Indexes Definition

RVI Ratio Vegetation Índex

nirRVI

r

NDVI Normalized Difference Vegetation Index

nir rNDVI

nir r

PVI Perpendicular Vegetation Index 2

*

1

nir M r QPVI

M

DVI Difference Vegetation Index

DVI nir r

IPVI Infrared Percentage Vegetation Índex

1

2

NDVIIPVI

WDVI Weighted Difference Vegetation

*WDVI nir M r

SAVI Soil Adjusted Vegetation Index

1nir r

SAVI Lnir r L

TSAVI Transformed Soil Adjusted Vegetation Index

2

* *

* * 1

B nir B r QTSAVI

r B nir Q M X M

NIR = value near infrared reflectivity

R = Value of reflectivity in the red

L = soil adjustment factor (L = 0.5; Huete 1998)

M = Slope of the land

Q = y-intercept of the line of the soil


13

3.2. Change detection techniques

Satellite images are a valuable tool for quick access to those areas under ecological

disasters. Their wide coverage area, his ability to quickly view and evaluate the situation of

those places, where the same consequences of the disaster prevent or hinder other types of

approach, are key factors in the management of the recovery actions after the event.

The use of remote sensing in monitoring deforestation processes is performed both

visual and digital analysis. Among the most interesting approaches should consider

employing linear spectral mixing analysis , in order to extract sub - pixel information from

NOAA - AVHRR images ( Braswell et al. , 2003 ; Hlavka and Spanner , 1995 ; Kressler

and Steinnocher , 1999 ) and software CLASlite ( Asner et al. , 2009 ) , the use of radar

images , in sectors with very frequent cloud coverage ( Thapa et al. , 2013 ) , multitemporal

analysis ( Lambien and Ehrlich , 2010 ; Chuvieco , 2002 Chuvieco Vargas , 1991 ).

Visual Analysis

For the detection of changes, such as deforestation and fragmentation, visual

interpretation procedure is appropriate since replacing farmland forest represents changes in

the spectral values of the contrasting images as well as ways that favor their identification

features (Figure 5).

Figure 5: Combinations of color Landsat image Paysandú forestation (//teledet.com).


14

Linear Spectral Mixing Analysis

In the mixing model are estimated the proportions of the components of mixed

pixels, from the spectral response of the components, this model is especially useful in low

spatial resolution images.

The Advanced Very High Resolution Radiometer (AVHRR) is a sensor on board

the satellite platform of the National Oceanic and Atmospheric Administration (NOAA)

and was originally designed for meteorological applications. However, in recent years, is

used to monitor the surface of the earth, particularly the vegetation dynamics regional and

global level (spatial resolution of 1.1 km), using reflective bands (580-680 nm in the visible

red, 725-1100 nm in the near infrared).

The use of this sensor may be combined with other high spatial resolution, such as

Landsat providing more detailed information to calibrate AVHRR data.

In Hlavka and Spanner (1995) the model was applied to the radiance (L) of a given

pixel, considering that DN are a linear function of radiation, which groups coverage that

wanted to distinguish (clearcutting, forest and succession) has spectral signatures different

and that the patches contain all three types of covers.

True

Color

(3, 2, 1)

False Color NIR

(4, 3, 2)

False Color

Corine

(4, 5, 3)

SWIR

(GeoCover)

(7, 4, 2)

Trees and

shrubs

Olive Green Red Brown-Orange Various Green

Crops Green to

light Green

Rose to Red Yellowish Various Green

Wetlands Dark Green

to Black

Dark Red Black Various Green

Water Blue-Green Blue Black Black to Dark

Blue

Urban

Areas

White to

Light Blue

Blue to Grey Gray-green to

Green-Blue

Violet

Soil White to

Light Blue

Blue to Grey Blue-Green to

White

Violet to rose


15

L (p, b) = Ml(b) * Fl(p) + M2(b) * F2(p) + M3(b) * F 3 b ) + e(p, b) (1)

Subject to the constraint

F l (p) + F2 (p) + F3 (p) = 1

Where Ml (b), M2 (b), and M3 (b) are the mean L values for classes 1, 2, and 3 in

band b; and Fl (p), F2 (p), and F3 (p) are the fractions of the three classes. The error term e

(p, b) is a term representing the combined effect of local deviations of L values of the

components from their average values. Braswell et al. (2003), applied this technique using

reflectance data from MISR and MODIS sensors and ISODATA unsupervised

classification of land use from Landsat ETM + (30 m resolution) of which were recognized

5 types of cover (Figure 6).

Figure 6: Observed and predicted values of forest, secondary, and cleared fractional areas for the reference case in Ruropolis (Braswell et al., 2003).

For the evaluation of this methodology results, Steinnocher Kessler (1999) propose

three methods: visual analysis, average squared error calculation and the calculation of the

overflow fraction. The RMS error can be calculated for each pixel or the whole image,

smaller RMS, better the model. In the third test, the fraction of land cover components

should be between 0 and 1, if the model is not well-built fractions values are outside this

range.


16

Other studies estimated the percentage of pixel occupied by forest and no forest in

MODIS images. Hansen et al. (2008) propose an approach that uses a regional coverage

MODIS product to calibrate high-resolution Landsat imagery. The method is an automated

decision tree algorithm that uses an algorism tool to characterize forest cover, evaluate

presence of clouds and shadows, and generate maps identifying as afforestation those pixels

with more than 60 % coverage and no afforestation, those with less this value. Then are

performed Landsat compositions and finally, the maps are elaborated (Figure 7).

Figure 7: Flow diagram of multi-resolution forest cover mapping and change detection

methodology where the following text sub headings from the Methods section are highlighted:

1) Generate regional MODIS 250 m forest non-forest cover map forest non-forest cover map, 2) Georectification/resampling of satellite data, 3) Landsat normalization, 4) Landsat cloud

and shadow flagging, 5) Landsat decision tree forest mapping procedure, 6) Landsat

compositing, 7) Landsat forest change mapping (Hansen et al., 2008).


17

CLASlite Software

CLASlite is a highly automated system for converting satellite imagery from its

original format (unprocessed), through calibration, pre-processing, atmospheric correction

and cloud masking steps, Monte Carlo Spectral Mixture Analysis and expert classification

to obtaining high-resolution images output (Figure 8).

Figure 8: Automated System CLASlite (http://claslite.stanford.edu).

This tool was designed specifically to support forest monitoring for REDD program.

Its outputs include maps of percent cover of living and dead vegetation, bare soil and other

substrates in addition to quantitative measures of the uncertainty in each pixel of the image.

These maps are interpreted in terms of forest cover, deforestation and forest disturbance

using automated decision trees.

The output images CLASlite can enter directly to other screening programs,

geographic information systems (GIS), Google Earth or other display systems.

A detail of system processes were described by Asner et al. (2009) (Figure 9):

http://claslite.stanford.edu/


18

Figure 9: CLASlite version 2.0 processing stream (Asner et al., 2009)

Satellite imagery

Currently, the following images are supported: Landsat 4 and 5 Thematic Mapper

(TM), Landsat 7 Enhanced Thematic Mapper Plus (ETM+), Advanced Spaceborne

Thermal Emission and Reflection Radiometer (ASTER), Earth Observing-1 Advanced

Land Imager (ALI), Satellite pour l'Observation de la Terre 4 and 5 (SPOT), and the

Moderate Resolution Imaging Spectrometer (MODIS). In addition to the image, the

software requires basic information about the sensor and geographical location (the

metadata of the image), for processing.

Calibration and atmospheric correction

Applies gains and offsets set for the sensor, to calibrate exo-atmospherically in each

band, then the radiance data are passed automatically to reflectance values with

atmospheric radioactive transfer model 6S.

Use the latest version of 6S (http://6s.ltdri.org/) supporting Landsat-4, 5 and 7,

ASTER, ALI and SPOT. Do not support MODIS atmospheric correction, because these

images already have processed products.

Corrects the fog for optical sensors in tropical regions.


19

Initial Image Masking

Water bodies are masked automatically to be removed in subsequent analyzes. This

is accomplished by detecting the reflectance properties only of water (reduced from blue to

near infrared) and by clouds leftovers mask by identifying pixels that appear negative

reflectance values .

On Landsat images, it only has the option to hide a portion of the pixels with clouds

using the thermal band, to reduce the processing time for the next step.

Sub - pixel Analysis

The core process is a sub CLASlite - model called AutoMCU (Automated Monte

Carlo unmixing), which provides a quantitative analysis of the fraction or percent cover (0-

100 %) of live and dead vegetation and bare substrate within each pixel. Live vegetation or

photosynthetic vegetation (PV) has unique spectral properties associated with leaf

photosynthetic pigments and water content of the canopy. Senescent vegetation fraction is

called non-photosynthetic vegetation (NPV), is seen in the spectrum as bright surface

material spectral features associated with carbon compounds from the plant.

The AutoMCU is based on a probabilistic algorithm that use three spectral libraries

derived from field measurements and hyperspectral satellite images to decompose each

pixel image using a linear equation. Iteratively selects a PV, NPV and bare substrate

spectrum of each library, and the pixel reflectance unmixing in housing constituent

fractions using the equation. The random selection process is repeated until the solution

converges to a mean value for each fraction of covered surface.

To assess performance, pixel by pixel, standard deviation and RMSE images are

generated, which allows the user to identify areas of concern (Figure 10).


20

Figure 10: Processing stream for the Automated Monte Carlo

Unmixing (AutoMCU) sub-model within CLASlite (Asner et al., 2009)

Secondary Image Masking

After determining PV fractional coverage, VPN and bare substrate within each pixel

of the image, there is a secondary masking and a scale change. The masking is applied with

a threshold value selected by the user based on RMSE derived from the AutoMCU model.

Deforestation and disturbance mapping (multi-image mode)

CLASlite includes a fully automated ability to detect the change from a time series

of images taken from the same geographical area. Analysis of multiple images is the most

accurate method for detecting the loss of forests (deforestation), the gain (secondary

regrowth) or degradation (forest disturbance zones persistent).

Forest Cover Analysis (single-image mode)

Although it is preferable to use at least two consecutive images to detect

deforestation and forest disturbance, it is possible to map the forest canopy coverage using


21

only one image. These images of forest cover often indicate areas of past logging and

disturbance.

Technical limitations

CLASlite outputs, especially maps disturbance usually require further analysis in

order to interpret the specific types of disturbances.

Radar

In areas characterized by constant cloudiness and fog, data from Synthetic

Aperture Radar are the only alternative for the study of vegetation cover.

According Thapa et al. (2013), the radar images ALOS - PALSAR (L-band) HH

-HV polarization and angle of 34.3 °, are highly applicable to the characterization of

several tropical forests.

Image segmentation is used and applied thresholds for discriminating between

types of soil coverage on the basis of backscattered HH and HV bands. Another

segmentation technique, multiresolution is applied in order to minimize heterogeneity and

maximize homogeneity.

The segmentation procedure begins with a pixel of a single image object and

repeatedly merges in pairs in several loops to larger units, provided that an upper threshold

of homogeneity that cannot be exceeded locally. Then the objects in the image are labeled

by establishing standards for different types of land cover (Figure 11).


22

Figure 11: Four land cover maps (A. 2007, B. 2008, C. 2009, and D. 2010) and locations of

concession and conservation areas including deforested landscape observed in 2007e2010 (E).

Multitemporal Analysis

Multitemporal analysis techniques from satellite imagery to track deforestation

provide a useful source for the management of the territories affected by this phenomenon.

The analysis is based on the interpretation and comparison of images from different dates

and different sensors (Landsat and Spot).

In the case of multi-sensor studies, you should keep in mind that the images should

be compatible. For this, it’s necessary to proceed with the normal pre-processing

(radiometric calibration, geometric and atmospheric correction setting) before starting the

multitemporal analysis.

For the study of deforestation is useful to use transformations of images that

reinforce plant component as indexes normalized difference vegetation index (NDVI ) (

critical value for ground cover around 0.1 and 0.5 dense vegetation ) indicating the state of

vitality of vegetation, reducing its value with the loss of vitality and facilitating the

identification of areas of forest cover changes . With a histogram segmentation, forest cover

of each image obtained from NDVI, defined categories of interest (afforestation and

reforestation). Are obtained thus, binary maps of each date and multitemporal crossover is

performed to identify themselves with the changes and / or changes in the vegetation in the

area (Chuvieco, 2002 Chuvieco Vargas, 1991).


23

Yoshikawa and Sanga-Ngoie (2013), performed an analysis of NDVI multitemporal

images NOAA / AVHRR (corrected) using Geographic Information System (GIS) Idrisi32

(Figure 12).

Figure 12: Method of Analysis (Yoshikawa y Sanga-Ngoi, 2013)

Another approach is to detect deforestation from "hot spots" on a large spatial scale.

"Hot spots" can be defined as areas that experience a high rate of deforestation in the

present and / or in the recent past. To define different variables can be used, starting with

those that allow the initial identification of large areas susceptible to deforestation,

excluding those where the probability of deforestation in the short term is low and

continuing with increasingly relevant variables (Figure 13).


24

Figure 13: Sample "funnel" to detect hot spots of deforestation (Lambin y Ehrlich, 1997).

Two variables highly related to deforestation processes can be derived directly from

large-scale maps of land cover such as that are produced by the TREES project (maps with

NOAA-AVHRR images 1.1 km resolution): the proportion of land cover and the spatial

fragmentation of these ones.

Maps are developed of areas that have high prior probability of deforestation,

according to a single variable and overlaps through a GIS. With the aim of increasing the

precision of hot spots, variables are combined (not more than 4 variables) according to

logical relationships (high density presence of fires with high rates of growth of the

population) (Lambin, 1997).


25

CHAPTER 4

Conclusions

The conservation and management of forests are very important for the

development of human life. However, this is only possible if conservation measures

contemplate economic and social situations unfair in different countries. Developing

countries are those with the largest areas of tropical forests but also have countless

economic and social needs, so that, forest conservation is not a priority and in many cases,

are an obstacle development of economic activities such as agriculture, oil extraction,

extraction of minerals and timber.

The forest conservation policies have to offer financial compensation, to involve

communities (especially indigenous peoples), should be established based on multi-

disciplinary and jurisdictional studies and must be applied consistently at all scales, but

mostly it is necessary to adopt innovative techniques in order to make more efficient the

use of resources, since the monitoring of forest tracts to measure deforestation, degradation

and / or evolution of the forests are not easy.

In this sense, the remote sensors offer a quick and low cost way to map thousands of

acres per day to monitoring deforestation, logging and other disturbances, as well as,

following the recovery of forests. The access to images is becoming easier and cheaper and

analysis techniques range from visual to those fully automated and can when applied to

optical (spectral values) and radar sensors (backscatter values ).


26

References

Martin Whittle, Shaum Quegan, Yumiko Uryu, Michel Stuewe and Kokok

Yulianto, 2012. Detection of tropical deforestation using ALOS-PALSAR: A Sumatran

case study. Remote Sensing of Environment 124, 83–98.

Rajesh Bahadur Thapa, Masanobu Shimada, Manabu Watanabe, Takeshi Motohka

and Tomohiro Shiraishi, 2013. The tropical forest in south east Asia: Monitoring and

scenario modeling using synthetic aperture radar data. Applied Geography 41, 168e178

Emilio Chuvieco, Javier Salas y Enrique Meza, 2002. Empleo de la teledetección en

el análisis de la deforestación tropical: el caso de la reserva forestal de Ticoporo

(Venezuela). Serie Geográfica Nº. 10: 55 – 76.

E. F. Lambin y D. Ehrlich, 1997. The identification of tropical deforestation fronts

at broad spatial scales. International Journal of Remote Sensing. 18:17, 3551-3568.

F. Vargas y E. Chuvieco, 1991. Dinámica de la Deforestación en Regiones

Tropicales a partir del Análisis Multitemporal de Imágenes Spot. Teledetección y Medio

Ambiente. IV Reunión Científica Asoiación Española de Teledetección. Pp. 70-76.

Gerardo Bocco, Manuel Mendoza y Omar Masera, 2001. Dinámica del cambio del

uso del suelo en Michoacán. Una porpuesta para el de los procesos de Deforestación. Red

de Revistas Científicas de América Latina, el Caribe, España y Portugal núm. 44, pp. 18-

38.

S. Yoshikawa a & K. Sanga-Ngoie, 2013. Deforestation dynamics in Mato Grosso

in the southern Brazilian Amazon using GIS and NOAA/AVHRR data. International

Journal of Remote Sensing, 37-41.

Florian Kressler and Klaus Stinnocher, 1999. Detecting land cover changes from

NOAA-AVHRR data by using spectral mixture analysis. JAG l Volume 1 - Issue 1.


27

Chistine Hlavka and Michel Spanner, 1995. Unmixing AVHRR Imagery to Assess

Clearcuts and Forest Regrowth in Oregon. IEEE TRANSACTIONS ON GEOSCIENCE

AND REMOTE SENSING, VOL. 33, NO. 3.

N. C. Ferreira, L. G. Ferreira, A. R. Huete and M. E. Ferreira, 2007. An operational

deforestation mapping system using MODIS data and spatial context analysis. International

Journal of Remote Sensing Vol. 28, No. 1, 47–62.

Matthew Hansen, David Roy, Erik Lindquist, Bernard Adusei, Christopher Justice

and Alice Altstatt, 2008. A method for integrating MODIS and Landsat data for systematic

monitoring of forest cover and change in the Congo Basin. Remote Sensing of

Environment 112, 2495–2513

Gregory P. Asner, David E. Knapp, Aravindh Balaji, and Guayana Páez-Acosta,

2009. Automated mapping of tropical deforestation and forest degradation: CLASlite.

Journal of Applied Remote Sensing, Vol. 3.

B.H. Braswell, S.C. Hagen, S.E. Frolking, W.A. Salas, 2003. A multivariable

approach for mapping sub-pixel land cover distributions using MISR and MODIS:

Application in the Brazilian Amazon region.

Claudia Herrán, 2012. PROGRAMA ONU-REDD: LA CONTRIBUCIÓN DE LOS

PAÍSES EN DESARROLLO PARA FRENAR EL CAMBIO CLIMÁTICO, www.fes-

energiayclima.org.

FAO, 1995. Necesidades y recursos. Geográfica de la agricultura y la alimentación.

www.fao.org.

FAO, 2012. Los Bosques del Mundo, www.fao.org.

OIMT, 2002. Directrices de la OIMT para la restauración, ordenación y

rehabilitación de bosques tropicales secundarios y degradados. Serie de políticas forestales

N° 13. www.itto.int.

http://www.fes-energiayclima.org/

http://www.fes-energiayclima.org/

http://www.fao.org/

http://www.fao.org/

http://www.itto.int/


28

OIMT, 2008. Plan de Acción de la OIMT 2008-2011. Serie OIMT de políticas

forestales N° 18. www.itto.int.

Ley de Bosques Nativos N° 26.331/07. LEY DE PRESUPUESTOS MINIMOS DE

PROTECCION AMBIENTAL DE LOS BOSQUES NATIVOS.

http://www.diputados.gov.ar.

http://www.itto.int/

http://www.diputados.gov.ar/