Ecosystem health assessment based on remote sensing a case study of ca river basin, vietnam

Click here to insert pictureECOSYSTEM HEALTH ASSESSMENT BASED ON REMOTE SENSING A CASE STUDY OF CA RIVER BASIN, VIETNAM Bao Quoc Tran MSc Thesis WM-WRM.16-22 Student number: 47074 Ap

Click here to insert picture

ECOSYSTEM HEALTH ASSESSMENT

BASED ON REMOTE SENSING

A CASE STUDY OF CA RIVER BASIN, VIETNAM

Bao Quoc Tran

MSc Thesis WM-WRM.16-22

Student number: 47074

April 2016

ECOSYSTEM HEALTH ASSESSMENT BASED ON

REMOTE SENSING

A CASE STUDY OF CA RIVER BASIN, VIETNAM

Master of Science Thesis

Ir G.J Roerink (WUR-Alterra)

This research is done for the partial fulfilment of requirements for the Master of Science degree at the

UNESCO-IHE Institute for Water Education, Delft, the Netherlands

Delft April 2016

Although the author and UNESCO-IHE Institute for Water Education have made every effort

to ensure that the information in this thesis was correct at press time, the author and IHE do not assume and hereby diSLCaim any liability to any party for any loss, damage, or disruption caused by errors or omissions, whether such errors or omissions result from negligence, accident, or any other cause

UNESCO-© Bao Quoc Tran 2016

This work is licensed under a Creative Commons Attribution-Non Commercial 4.0 International License

i

Abstract

The Ca river basin is the third biggest river basin in Vietnam, in which the upper part belongs to Laos With the World Biosphere Reserve Western Nghe An officially recognized by UNESCO in 2007, the up-to-date information about ecosystem and biodiversity, in particular, flora biodiversity is urgently needed for conservation strategies as well as policy making process in resource management This study aims to evaluate Skidmore et al (2015)’s proposed variables to assess the ecosystem health in the Ca river basin from remote sensing indices, namely leaf area index, soil moisture, net primary production, land use and fire occurrence with the hypothesis that this framework is generic for all ecosystems The remotely sensed imagery was retrieved from Landsat 7 ETM+ SLC-off in March of three years, namely 2005, 2010 and

2015 In addition, a weighted scoring approach has been attempted to assess the vigor aspect of ecosystem health The results showed that the LAI was underestimated, which might imply that the function retrieving LAI from SAVI for all crops was not applicable in this study area In addition, the calculation of soil moisture should be taken into account the weather condition since they were estimated in a day only Accordingly, the fire occurrence map also pointed out some areas where the fire events happened at least twice in three time steps, which might be caused by slash and burn practices of local inhabitants to prepare for the next crop Related to ecosystem health assessment in 2010 and 2015 in comparison with 2005, which is considered

a benchmark, the ecosystem of the study site in 2010 was moderate healthy and getting viable

in 2015 On the other hand, Skidmore et al (2015)’s approach remains with technical and conceptual limitations Likewise, it is believed that an agreement on an optimal resolution, in terms of temporal, spatial and spectral resolutions, should be drawn among research communities to bridge the gaps between remote sensing experts and ecological users Furthermore, the weighted scoring approach should be integrated between professional biologists as well as the statistical data from the field to minimize the bias In the end, local calibration is a must since every ecosystem has its own characteristics and the algorithms might not be applied to all ecosystems

Keywords: Ecosystem Health Assessment, remote sensing, weighted scoring approach

iii

Acknowledgements

I would like to thank Dr Hans van der Kwast for his support and many helpful contributions over the past five months His suggestions on how and where to find support for this study was

a major contributing factor to its completion A special thanks to Tim Hessels who encouraged

me and engaged his time in technical issues Together with intensive assignments and meetings almost every week, their requirements have been keeping me on track with the thesis working

so that I could finish my thesis on time and engaged myself in scientific research related to remote sensing

This paper would not have been possible without Professor Wim G.M Bastiaanssen, who first inspired me to this thesis topic from a lecture given in early last year and guided me with the theoretical concepts on remote sensing and biodiversity

I would also like to acknowledge Adeline, my best friend and my classmate, for assisting with many questions that I had on English writing and discussion We did have a lot of memories when working in DOK, TU Delft library since morning until midnight with nice coffee and food

A special thanks to Louis who is sharing with my all the sorrows we had when struggling with thesis writing We did have several stories to tell about life, about love, about tears, even about the relationship between the duck-canal network in Delft with biodiversity I will miss the time we travelled together to Berlin and worked hard with our thesis on the bus

A special thanks to Jam, Mariel, Clara, Shabana, Saltana and other Water Management classmates who always take care of me and spent crazy time with me in last 18 months

Last but not least, I would like to leave the last paragraph to give all my love to my family, who always encourages and are beside me unconditionally Finally, I would like to give

a big hug to my boyfriend, Quốc Trạng, who did encourage me to get this scholarship in last two years and supports me with love and humor, smile and tears, strengths and efforts to overcome all the obstacles in my life I love you

2.2 Skidmore et al (2015)’s proposed biodiversity variables 10

2.3.1 Normalized Difference Vegetation Index (NDVI) 12

4.3 Retrieving remote sensing indices and biodiversity variables from Landsat 7

5.1 Derivation of biodiversity variables from remote sensing 47

vii

5.2 Ecosystem Health Assessment using Weighted Scoring approach 60

ix

List of Figures

Figure 1-1 Ecosystem services (Sheet 7) in Water Accounting Plus Framework 3

Figure 1-2 The research flowchart 7

Figure 2-1 NDVI and SAVI calculated from a Landsat TM5 image of south-western Idaho 13 Figure 2-2 Spectral response curves of vegetation and burned area 14

Figure 2-3 Flowchart of global vegetation classification logic 19

Figure 2-4 Conceptual representation of a forest standing indicating the relative positions of mean canopy height and scattering phase center height within a single SRTM resolution cell 20

Figure 3-1 Location and topographic map of the case study 25

Figure 3-2 Location of Phu Xai Lai Leng 26

Figure 3-3 Climatogram of study area, data at Vinh station (2013) 27

Figure 3-4 World Biosphere Reserve Western Nghe An 27

Figure 3-5 Representative flora and fauna in study area 28

Figure 3-6 The trend in land use structure from 2000 to 2013 29

Figure 3-7 Land Use map of Ca River Basin, Nghe An, Vietnam (2012) 30

Figure 4-1 The flowchart of retrieving biodiversity variables from remote sensing 34

Figure 4-2 The difference of with and without SLC in processing image 36

Figure 4-3 Inverse Distance Weight Interpolation based on weighted sample point distance (left) and Interpolated IDW surface from elevation vector points (right) 37

Figure 5-1 Leaf Area Index in three time steps (2005, 2010, 2015) 48

Figure 5-2 Mean LAI in three time steps per land use 48

Figure 5-3 Soil moisture in three time steps 51

Figure 5-4 Soil moisture per land use in three time steps 52

Figure 5-5 Burn severity levels of study area in three time steps 55

Figure 5-6 Frequency of fire occurrence from 2005 to 2015 56

Figure 5-7 Net primary production in three years (2005, 2010, 2014) 57

Figure 5-8 Average net primary production per land use in three years 58

Figure 6-1 Spatial and temporal resolution of both ecological processes and remote-sensing observation 69

xi

List of Tables

Table 1.1 Examples of regulation services 2

Table 1.2 Ten proposed biodiversity variables in Skidmore et al (2015)’s framework 5

Table 2.1 Ten proposed biodiversity variables by Skidmore et al (2015) 11

Table 2.2 Typical NDVI values for various cover types 13

Table 4.1 Radiometric range of bands and resolution for the ETM+ sensors 35

Table 4.2 Data preparation for this study 37

Table 4.3 ESUN value for Landsat 7 sensor 39

Table 4.4 List of formulas and methods used in the paper 40

Table 4.5 Ordinal severity levels and example range of dNBR (scaled by 103) 43

Table 4.6 Illustration of approach used in ecosystem health assessment 44

Table 4.7 Illustration of weighted scoring approach for the ecosystem of vigor 45

Table 5.1 Leaf area index per land use in three time steps 50

Table 5.2 Coefficients of the polynomial relationship for Mo between T* and fc 51

Table 5.3 Soil moisture per land use in three time steps 54

Table 5.4 Net primary production per land use in three years 59

Table 5.5 Ecosystem health assessment of LAI in 2010 and 2015 61

Table 5.6 Ecosystem health assessment of soil moisture in 2010 and 2015 62

Table 5.7 Ecosystem health assessment of NPP in 2010 and 2014 63

Table 5.8 Weighted scoring for the vigor of ecosystem health in 2010 and 2015 64

Table 6.1 A combination between indicators of ecosystem health and 71

xiii

Abbreviations

APAR Absorbed Photosynthetically Active Radiation

ASTER Advanced Spaceborne Thermal Emission and Reflection Radiometre

ATM Airborne Topographic Mapper

AVHRR Advanced Very-High Resolution Radiometre

CGIAR Consultative Group on International Agricultural Research

ETM+ Landsat Enhanced Thematic Mapper Plus

EOS Earth Observing System

GPP Gross Primary Productivity

INSAR Interferometric Synthetic Aperture Radar

LAI Leaf Area Index

LIDAR Light Detection and Ranging

LST Land Surface Temperature

MODIS Moderate-resolution Imaging Spectro-radiometre

NASA The U.S National Aeronautics and Space Administration

NBR Normalized Burned Ratio

NPP Net Primary Production

NDVI Normalized Difference Vegetation Index

PAR Photosynthetically Active Radiation

SAR Synthetic Aperture Radar

SAVI Soil Adjusted Vegetation Index

SEBAL Surface Energy Balance Algorithm for Land

SRTM Shuttle Radar Topography Mission

SVIs Spectral Vegetation Indices

SWIR Short-wave Infrared

TIR Thermal Infrared

TM Landsat Thematic Mapper

V/NIR Visible/Near Infrared

WLE Water, Land and Ecosystem

as well as impacts on the hydrological cycle (e.g interception, runoff, ET) Therefore, it is believed that up-to-date information on changes can facilitate the policy makers in water resources management and planning In this sense, the data from earth observation will play an important role in providing the actual and historical information

Accordingly, ecosystem services are defined as the benefits people obtain from ecosystems (Boyd & Banzhaf, 2007) or the combined actions of species in an ecosystem performing functions of value to society (CGIAR, 2014), including provisioning, regulating, cultural and supporting services To emphasize the role of ecosystem services in natural resources management and in economy, Costanza et al (2014) highlighted that the valuation of ecosystem services is not the same as commodification or privatization, they are best considered public good requiring new institution From other perception, Water, Land, and Ecosystems (WLE) focus more on ecosystem as “common pool resources” which emphasizes the impacts

of collective action or large-scales intervention (CGIAR, 2014) WLE also taken into account the horizontal flow of Ecosystem Services and Resilience framework (as the given example in Table 1.1) to investigate the impact of biophysical structure and processes

Introduction 2

Table 1.1 Examples of regulation services

Source: The Consultative Group on International Agricultural Research (2014)

Ecosystem

services

categories

Example of ecosystem services studied by WLE

Plot, farm and small- catchment scale approaches

WLE (Landscape scale approaches)

Regulating services

Regulation of

water flows

Natural drainage irrigation and drought prevention

Infiltration and storage

capacity of cropping systems and field

management practices

Impacts of groundwater regulation, wetland system (e.g Tonle Sap), riparian forest, protected forest areas on flow regulation; impact of landscape-level on water quality, extent of riparian forest and field margin management needed to capture and store excessive nutrient loads and ensure water quality Climate

regulation

Carbon sequestration, influence of vegetation on infiltration and rainfall

Greenhouse gas (GHG) sequestration

of cropping systems

GHG sequestration of alternative land-use compositions and configuration

Accordingly, Bastiaanssen et al (2015)’s Water Accounting Plus (WA+) framework has clarified the ecosystem services through the water consumption and non-consumptive use

of an ecosystem (Fig.1-1) In the latter section, the numbers of flora biodiversity are used as an input to estimate water regulating services

Introduction 3

Figure 1-1 Ecosystem services (Sheet 7) in Water Accounting Plus Framework

Source: www.wateraccounting.org

Introduction 4

Since the 1990s of last century, thanks to the development of space technology, remote sensing and its applications has been acknowledged in several domains, such as agriculture (e.g land use, water accounting in irrigation), transportation (e.g google map) In fact, with the remote sensing analysis, the end-users, namely policy makers, management board, scientists, are able to manage and control the planning at different administrative scales based on the data derived from satellite images The applications of remote sensing are first and foremost for land-use classification, forest fire detection, urbanization Kennedy et al (2009) concluded the vital role of remote sensing in natural resources management as the provision of consistent measurements of landscape condition, which allows detection of both abrupt changes and slow trends over time Besides, Stork & Samways (1995) clarified two major areas where monitoring biodiversity is applicable: the assessment of the effectiveness of biodiversity management which aims to preserve and optimize biodiversity for other goals (e.g plantation) by national

or regional program The latter concentrates more on developing an early-warning system in impending adverse changes before they become too critical

Recently, papers on addressing the biodiversity, ecosystem status and effects of climate variables are noticed: assessing effects of climate change (Bakkenes et al., 2002), forest change detection (Ivits & Koch, 2002; Deslcée et al., 2006), ecosystem services (Peterson, 1997; Groot

et al., 2012), ecosystem health assessment (Rapport et al., 1998; Lu & Li, 2003; Ding et al., 2005) Strand et al (2007) emphasized the ability to detect change in vegetation cover and the associated habitat with this cover by the alteration in the remote sensing signal from one-time

period to another Busby (2003) also questioned the fundamental challenges in translating the

definitions about biodiversity into operational program along with what a high biodiversity value is However, these studies shared the shortcomings in the communication between remote sensing experts and ecologists or biologists relating to an agreement on indicators, which can lead to the insufficient results drawn by remote sensing users Skidmore et al (2015) call on an

agreement on a definitive set of “biodiversity variables between conservation and space

agencies as well as how these will be tracked from space, to address conservation target”

Therefore, in this study, Skidmore et al (2015)’s framework to detect biodiversity will be used This framework included ten proposed variables for satellite monitoring which are grouped into four categories, named as species population, species traits, ecosystem structure and ecosystem functions, as described in Table 1.2

Introduction 5

Table 1.2 Ten proposed biodiversity variables in Skidmore et al (2015)’s framework

Species populations Species occurrence

Species traits (plant traits) Specific leaf area

Leaf nitrogen content

Ecosystem structure Ecosystem distribution

Fragmentation and heterogeneity Land cover

Vegetation height

Ecosystem function Fire occurrence

Vegetation phenology (variability) Primary productivity and leaf area index Inundation

In this research, a part of this framework was applied to evaluate ecosystem health in the Ca river basin, consisting of four proposed variables: leaf area index, net primary production, fire occurrence, land cover In addition, soil moisture was added since it is also a key component in flora biodiversity processes Likewise, the definitions of “ecosystem health” mentioned in this study have been aligned with the concepts of stress ecology, in which ‘health’

is related to system organization, resilience and vigor, as well as the absence of sign of ecosystem distress (Rapport, 1989) In other words, a healthy ecosystem might have the ability

to maintain its structure (organization), function (vigor) and landscapes level over time in face

of external stress (resilience), as highlighted in Ding et al (2005)’s research

To clarify, three sub-objectives have been established:

1 To determine five variables that represent ecosystem structure, and ecosystem function named as (i) leaf area index, (ii) soil moisture, (iii) fire occurrence, (iv) land cover, and (v) net primary production

2 To examine Skidmore et al (2015)’s framework in the study area

3 To assess ecosystem health and flora biodiversity using three indicators related to vigor, organization and resilience of an ecosystem

1.4 Hypothesis

The main working hypothesis is that Skidmore et al (2015)’s framework is applicable for the

Ca river basin (Vietnam) in evaluating ecosystem health and flora biodiversity status, using five biodiversity variables, as described in Fig.1-2

Introduction 7

Figure 1-2 The research flowchart

1.5 Research questions

Based on the addressed problems, the following research questions are proposed:

1 How can the three key environmental variables (LAI, soil moisture, fire occurrence) be derived from remote sensing images?

2 How can ecosystem health be assessed using Skidmore et al (2015)’s approach?

3 What are the limitations of Skidmore et al (2015)’s approach and how can it be improved to be more generic?

1.6 Study site background

The Ca river basin is the third largest river basin of Vietnam This basin merges from the highlands of Lao People Democratic Republic (PDR) and flows into the Tokin Sea This basin encompasses 27,200km2 in which 65% located in Vietnam, covering the province Nghe An, Thanh Hoa, Ha Tinh, Quang Binh The river basin endowed with tropical rainfall patterns, annual average basin-wide rainfall of 1650 mm/year (Bastiaanssen et al., 2015) The majority

BIODIVERSITY VARIABLES

Fire occurrence

Net primary production

Q1

Q2 NBR

Introduction 8

of landscapes consists of natural vegetation, among others forests, bushland, grassland, and herbaceous cover The delta consists essentially of build-up areas, paddy fields and fish ponds Data collection process for flora biodiversity in Ca river basin is mainly based on historical data and ground-based fieldwork, which is time-consuming and requires human resources (e.g biologists, ecologists) Therefore, a strategy to provide up-to-date information and a quick change detection on flora biodiversity should be launched to solve this problem

1.7 Structure of the thesis

In chapter one, the Introduction, the research background is presented in relation to ecosystem health assessment, Skidmore et al (2015)’s framework to evaluate biodiversity as well as the research objectives, research questions and hypothesis of the research Furthermore, in chapter two, the Literature Review will be carried out to evaluate related studies on biodiversity and remote sensing The third chapter will describe the case study site with detailed ecohydrological characteristics Besides, based on reviews mentioned above, chapter Four will continue with the research strategy in which approaches and method used to solve research questions are explained In the next two chapters, results will be depicted with interpretation and discussed

to address the main findings and limitations At the end, conclusions will be drawn to address the significance of the thesis regarding biodiversity evaluation and ecosystem health assessment

as well as recommendation for future research on the same domains

In brief, an overall background of the thesis was introduced in this chapter with key words related to biodiversity evaluation, ecosystem health assessment, remote sensing Research objectives, research questions and the main hypothesis are the backbone for the whole thesis

to be taken in the following chapters

2.1 Ecosystem Health Assessment (EHA)

In general, an healthy ecosystem was defined in view of different disciplines and could be divided into two types: biological – ecological definition and ecological – economic definition (Peng et al., 2007) The former definition suggested by Rapport et al (1998) emphasizes the natural ecological aspects through the ranges of biological physics, ignoring the social economics parts and human health By contrast, the latter definition highlights the natural ecosystem regarding to the human demands and requirements (Liu et al., 2008; Coutts & Hahn, 2015) As above-mentioned, ecosystem health can be assessed through the measurement of vigor, organization, and resilience Vigor emphasizes the measurement of activity, metabolism

or primary productivity; whereas organization may be estimated by the diversity and number

of interactions between system components Besides, resilience or counteractive capacity can

be addressed as a system’s capacity to maintain the structure and function in the presence of stress until reaching an adaptation tipping point, then the system will alter to an alternative state (Rapport et al., 1988)

Additionally, not threatening to the other surrounding ecosystem and meeting the reasonable needs of humankind (Jorgensen et al., 2005), human health effects (Rapport et al., 2009) should be taken into account Human health itself might be evaluated as a ‘synoptic’ measure of ecosystem health In this sense, a healthy ecosystem is characterized by their capacity to sustain healthy human populations In brief, at present, it is acknowledged that the concept of health portrays the vitality of sustainability development as well as the core of integrated ecosystem management and ecosystem services (Peng et al., 2007)

Literature Review 10

To date the cooperation between remote sensing and ecology, the spatial-temporal scale characteristics of ecosystems are taken into considerations since scale issue is one of the key components of recent ecological researches (Lu & Fu, 2001) In this regard, ecological remote sensed-base assessment would be focus on two main upper scales: (i) landscape/region in which the effect of landscape spatial pattern to ecological processes and the dynamic maintenance of ecosystem services functions is engaged; whereas (ii) global scale spotlights the relationship between ecosystem services functions and human demands (Coder et al., 2003; Peng et al., 2007) To put it more specifically, studies at global scale might facilitate the understanding of global ecosystem health trend and public awareness, but the local features will be missed As a consequence, it would be difficult to assess and reflect ecosystem health For this reason, landscape/region scale might be preferable and become the key scale in ecosystem health assessment by connecting macro-(globe) and micro-(ecosystem) scales

Indicant (indicator) species method and indicator system method are widely applied to

evaluate ecosystem health (Kong et al., 2002; Peng et al., 2007) To clarify, the indicant species

method concentrates on the quantity, productivity and structural function (e.g keystone species, area-limited ‘umbrella’ species, resource-limited species or endangered species in a certain ecosystem) (Peterken, 1974; Kremen, 1992; Gerald & McDonald, 2004) Consequently, this approach failed to reflect the ecosystem health as a whole since socio-economic factors and human health are not taken into consideration On the contrary, based on the characteristics of

an ecosystem and its service function, an indicator system in which quantitative evaluation is undertaken will be established Selected indicators of the system mostly consist of ecosystem structure, function, and process along with indicators about socio-economic, landscape pattern, and land use (Peng et al., 2007)

2.2 Skidmore et al (2015)’s proposed biodiversity variables

In order to assess the progress toward the Aichi Biodiversity Target for 2011 – 2020 set by the Convention on Biological Diversity (Secretariat of the Conservation on Biological Diversity,

2010), Skidmore et al (2015) call on an agreement on a definitive set of “biodiversity variables

between conservation and space agencies as well as how these will be tracked from space, to address conservation target” These variables at the first step are retrieved after two workshops

Literature Review 11

in Germany and Italy which aims to bring remote sensing experts and ecology communities to generate the list

Table 2.1 Ten proposed biodiversity variables by Skidmore et al (2015)

Indicators Skidmore et al (2015) Remote sensing

Species

population Species occurrence

Plant traits Leaf area index NDVI (Carlson & Ripley, 1997;

Bastiaanssen, 1998;

Turner et al., 1999;

Boegh et al., 2012) Leaf nitrogen content Red edge (Cho & Skidmore, 2006;

Mutanga & Skidmore, 2007; Clevers & Gitelson, 2013)

Vegetation height SAR images,

( Jaiswal et al., 2002;

Roy et al., 2006;

Hernandez et al., 2006; Key & Benson, 2006 Vegetation phenology NDVI time series

Literature Review 12

Ten variables proposed are divided into four main aspects: (i) species occurrence, (ii) plant traits, (iii) ecosystem structure and (iv) ecosystem function However, species occurrence was taken out because it requires also information about flora and fauna biodiversity relationship which take more time to measure In addition, fragmentation and heterogeneity criterion in ecosystem structure is formulated the expected outcome to evaluate flora biodiversity Moreover, a so-called criterion soil moisture is added to the list owing to its vital roles in water management and quantifying surface temperature and water stress index Table 2.1 addresses the remote sensing indicators which can be derived from remote sensing in order

to identify the biodiversity variables proposed by Skidmore et al (2015)

2.3 Remote sensing indices as required inputs

2.3.1 Normalized Difference Vegetation Index (NDVI)

Normalized Difference Vegetation Index (NDVI), first used by Rouse et al (1973), is mostly used to determine the density of green on patch of land by earth observers Researchers have to differentiate the distinct colors (wavelengths) of visible and near infrared sunlight reflected

by the plants (Weier & Herring, 2000) Calculations of NDVI for a given pixel ranges from minus one (-1) to plus one (+1); however no green leaves gives a value close to zero NDVI is calculated by:

𝑁𝐷𝑉𝐼 = 𝑁𝐼𝑅 − 𝑅𝑒𝑑

𝑁𝐼𝑅 + 𝑅𝑒𝑑 (1)

In this regard, chlorophyll absorbs light in the red channel (0.58 – 0.68 µm) and foliage reflects light in the near infrared channel (0.72 – 1.10 µm) As a result, higher photosynthetic activity will result in low reflectance in the red channel and higher reflectance in the near infrared channel (Holben, 1986) Likewise, typical NDVI values for some cover types are presented in Table 2.2

2.3.2 Soil Adjusted Vegetation Index (SAVI)

In areas where vegetative cover is low (e.g under 40%) and the soil surface is exposed, the reflectance of sunlight in the red and near-infrared spectra can influence vegetation index value The SAVI is structured similar to NDVI but with the addition of a ‘soil brightness correction factor’ L is a correction factor which ranges from 0 (for very high vegetation cover) to 1 (very low vegetation cover) The most typically used value is 0.5 which is for intermediate vegetation cover This value minimizes the influence of background soil for a large variation of leaf area indices (Huete, 1988)

𝑆𝐴𝑉𝐼 = 𝑁𝐼𝑅 − 𝑅𝑒𝑑

𝑁𝐼𝑅 + 𝑅𝑒𝑑 + 𝐿(1 + 𝐿) (2) where NIR is the reflectance value of the near infrared band, RED is the reflectance of the

red band, L is the soil brightness correction factor When L = 0, then SAVI = NDVI

Figure 2-1 NDVI and SAVI calculated from a Landsat TM5 image of south-western Idaho

(Source: http://wiki.landscapetoolbox.org)

Literature Review 14

This image shows a section of South-Fork Owyhee River canyon NDVI image results in a high index value in the rocky river canyon which addresses much more vegetative cover than is actually there On the other hand, the SAVI shows a much better approximation of the amount and cover of vegetation in the canyon as well as in the upland

2.3.3 Normalized Burned Ratio (NBR)

Normalized Burned Ratio was developed by Key & Benson (2006) by integrating band 4 (Near Infrared) and band 7 (Shortwave Infrared) of Landsat TM/ETM+ sensor This index is calculated as:

𝑁𝐵𝑅 = 𝑁𝐼𝑅 − 𝑆𝑊𝐼𝑅

𝑁𝐼𝑅 + 𝑆𝑊𝐼𝑅 (3) Band 4 reflectance naturally reacts positively to leaf area and plant productivity, whereas band 7 reflectance positively responds to drying and some nonvegetated surface characteristics (Fig 2-2) In this sense, band 7 has low reflectance (it is absorbed) over green vegetation and moist surfaces, including wet soil and snow – just the opposite from band 4 NBR measures the difference between band 4 and 7 It is positive when band 4 is greater than band 7, most vegetated areas are productive When it is near zero, it can occur with clouds, non productive vegetation (cured grasses), and drier soils or rock The value is negative, which suggests severe water stress in plants and the nonvegetative traits created within burns

Figure 2-2 Spectral response curves of vegetation and burned area

Source: United States Forest Service (USFS, 2009)

Literature Review 15

In addition, to isolate burned from unburned areas to provide a quantitative measure of change, dNBR or delta dNBR is calculated:

𝑑𝑁𝐵𝑅 = 𝑁𝐵𝑅𝑝𝑟𝑒𝑓𝑖𝑟𝑒− 𝑁𝐵𝑅𝑝𝑜𝑠𝑡𝑓𝑖𝑟𝑒 (4) This measured parameter is hypothesized to be correlative in magnitude to the environmental change caused by fire, or the burn severity as it relates to fire effects on previously existing vegetative communities Theoretically, dNBR (scaled by 1,000) can range between -2,000 and +2,000, but in reality it is rare for valid data to vary much beyond -550 and +1,350 (based on the scope of disturbance factors potentially affecting natural landscapes so far encountered)

Negative values result from post fire NBR being greater than prefire NBR This may be due to clouds in the prefire image, or increased plant productivity in the postfire image Enhanced regrowth is detected in approximately the -500 to -100 range of dNBR A recent burn may exbihit this after one growing season if severity is light and the burn in in mostly herbaceous communities that recover quickly to exceed the productivity existing before fire Also, older burns exhibit this as they recover vegetatively from the first year postfire into subsequent years Pixel below about -550 are likely cloud effects, or noise caused by misregistration or anomalies in orgirinal Landsat data Extreme negative (or positive) values appear where data from one scene overlaps missing data in the other scene, as occurs near scene edges (Key & Benson, 2006)

2.3.4 Land surface temperature

The surface temperature (T0) is the skin temperature of the land surface, i.e., the kinematic temperature of the soil plus the canopy surface (or, in the absence of vegetation, the temperature

of the soil surface) T0 describes the equilibrium between energy supply (radiation balance) and energy consumption (energy balance) The interactive heat mechanisms between land and atmosphere are determined by radiation, conduction and convection of energy transport processes Surface temperature is a key parameter in the surface energy balance (Bastiaanssen, 1998) Moreover, the close relationship of surface energy balance and water status of the surface with land surface temperature was clarified by (Mira et al., 2016) It mainly depends on the amount of radiative energy absorbed by the surface, on the partitioning of heat in sensible and latent heat flux, and on the characteristics of the atmosphere close to the ground (in particular

Literature Review 16

air temperature and turbulence) Land surface temperature was estimated by applying Qin et al (2001)’s mono-window algorithm In this study, land surface temperature will be used as an input for soil moisture estimation

2.4 Using indices to retrieve biodiversity variables

2.4.1 Leaf area index (LAI)

Leaf Area Index or LAI (m2/m2) represents the amount of leaf material in an ecosystem and is geometrically defined as the cumulative area of leaves per unit ground surface area (Gobron, 2000) LAI represents the total biomass and is indicative of crop yield, canopy resistance, and heat fluxes (Bastiaanssen, 1998) Therefore, LAI appears as a key parameter in many models that address vegetation-atmosphere interactions, especially with respect to carbon and water cycles (Liang et al., 2014) Recently, empirical methods and physical methods are mostly applied to retrieve LAI from remotely sensed imagery (Liang et al., 2012) In fact, it seems that the so-called empirical methods are preferable since statistical relationships between LAI and spectral vegetation indices (SVIs) are the key components and distinct vegetation types calibration are modified with field measurement Many attempts estimate LAI from the normalized difference vegetation index (NDVI), simple ratio, reduced simple ratio (Stenberg et al., 2004), soil adjusted vegetation index (SAVI) (Bastiaanssen, 1998) Reduced simple ratio appeared to have more dynamic responses to LAI than simple ratio or NDVI in boreal coniferous forests (Stenberg et al., 2004)

2.4.2 Leaf Nitrogen Content

Nitrogen is one of the most essential macronutrient for plant development (production of protein and chlorophyll), yield, post-grazing regrowth and reproduction as well as for animal nutrition Lamb et al (2002) also highlighted the key role of nitrogen in chlorophyll related to the general reflection of the concentration of chlorophyll in plant leaves For instance, the deficiency of nitrogen can leads to chlorosis (yellowing) of leaves due to a drop of chlorophyll content (Donahue et al., 1977) Related to spectral indices, plant nitrogen levels could be expected to influence canopy reflectance in visible and near-infrared wavelengths, as chlorophyll-related plant pigments (400-700nm) and leaf cell structure (600-900 nm) (Mutanga

& Skidmore, 2007; Zhu et al., 2008) Experiments by Mutanga & Skidmore (2007) also pointed out the relation of red edge shift of the chlorophyll absorption feature and the increase of

Literature Review 17

nitrogen supply in plants Despite the strong correlation between the red edge and foliar nitrogen, its relationship is indirectly defined (through chlorophyll operation), which can be also affected by other elements

Using vegetation index (e.g NDVI) and soil-line vegetation indices (e.g SAVI, OSAVI) to investigate the nitrogen accumulation also can be found in several studies (Kimura

et al., 2004; Zhu et al., 2008; Feng et al., 2008; Bagheri et al., 2012; Hunt et al., 2012; Júnior

& Pinto, n.d.) Bagheri et al (2012) recommended that nitrogen prediction and model crop canopy nitrogen content could retrieve better results by using soil-line vegetation indices; whereas Hunt et al (2012) paid more attention to image analysis concerning about the pixel size which can lead to the inaccuracy in monitoring

2.4.3 Soil moisture

Soil moisture plays a key role in the atmospheric water cycle from a small agricultural scale to modelling land – atmosphere interaction (World Meteorological Organization, 2013) From irrigated perspectives, available moisture at certain root level is much more important to vegetation and crops than rainfall occurrence As a results, in situ soil moisture information is also required in irrigated scheduling and crop yield forecasting Moreover, in the sense of water management, soil moisture information can be used in early warning of drought as well as water allocation It is obvious that only the moisture in the top of few centimetres of soil can be detected by remote sensing (Arnold et al., 1999; Klemas & Pieterse, 2015; Kasim & Usman, 2016;) Therefore, ground-based measurement is essential for validation and calibration

Soil moisture, or soil water content, is the water that is held in the space between soil particles (Arnold et al., 1999) In other words, soil moisture can be defined by an expression of the mass or volume of water in the soil Recently, volumetric water content (m3/m3), the volume

of water in the soil compared to the total volume of dry soil, air and water, is standardized to express soil moisture instead of percentage of the mass of water in the soil versus the mass of dry soil contained (Carvalho-Santos et al., 2013; World Meteorological Organization, 2013) The retrieval of soil moisture from satellite measurements has been witnessed for years Many techniques have been proposed and successfully taken out to derive soil surface wetness from remote sensing such as optical (trapezoid (Moran et al., 1994), triangle method (Carlson, 2007; Yang et al., 2008; Kasim & Usman, 2016;)), microwave (SAR, passive), and thermal infrared techniques

Literature Review 18

In this study, the Triangle Method for estimating soil moisture from satellite imagery proposed by Carlson (2007) was applied This method is based on the relationship between scaled surface temperature (which varies from 0 to 1) and fractional vegetation cover derived from remotely sensed imagery One of the underlying principles is that surface temperature is sensitively dependent upon soil surface wetness Accordingly, spatial and temporal variation in surface wetness are reflected by variations in surface temperature (Kasim & Usman, 2016)

2.4.4 Land cover

Land cover is defined as “the observed (bio)physical cover on the earth’s surface”, whereas

land use is “characterized by the arrangements, activities and input people undertaken in certain land cover type to produce, change or maintain it” (FAO, 2009.) It is clearly that

choosing a classification scheme to have efficient definition legends for the map faces challenges (Horning, 2004) There are a few classification schemes that should be taken into account before carrying out the land cover map: (i) Anderson’s classification (Anderson et al., 1976) , (ii) 14 land cover classes in different spatial scales (Hansen et al., 1998) , (iii) USGS classification developed and modified from Anderson (USGS, 2012) In fact, it remains difficulties in applying these classes in remote sensing analysis

A simple new logic for classifying based on vegetation structure and measurable in the field for validation was presented by Running et al., (1995) This scheme combines three primary attributes of plant canopy structure to build up the classification: (i) permanence of above-ground live biomass, (ii) leaf longevity, and (iii) leaf type, as shown in Fig 2-3

Literature Review 19

Figure 2-3 Flowchart of global vegetation classification logic

Source: Running et al (1995)

2.4.5 Vegetation height

Generally, an optical remote sensing is well suited to the acquisition of the structural information in horizontal dimension (e.g vegetation cover) However, the passive optical systems which depends on the sunlight are limited in addressing through the layers of vegetation That results in the obstacles in the accurate retrieval from vertical dimension, like the canopy height and increase of vegetation density As a consequence, the active remote sensors has been developed to fulfill these requirements Since the Shuttle Radar Topography Mission (SRTM) sensor mapped 80% of the Earth’s land mass with a C-band Interferometric Synthetic Aperture Radar (INSAR) instrument (Fig.2-4.), the majority of incoming electromagnetic energy is reflected by scatters located within the vegetation canopy (Kellndorfer et al., 2004)

2.4.6 Burn severity levels

In terms of fire management, the remotely sensed data can assist at three different stages relative

to fire occurrence (FAO, 2002): (i) before the fire (vegetation biomass, vegetation status, rainfall; monitoring pre-suppression or fire prevention measures), (ii) during the fire (near real-time location of active fires), and (iii) after the fire (assessment of burned areas) Besides, the assessment of burned area could also be applied to test the connectivity of flora biodiversity if there were fire occurrence before FAO (2002) identified two main changes in surface properties following fire, which can be detected by earth observation approach, named as removal of vegetation and the higher heat of burned surface compared to surrounding vegetation (with a maximum contrast in temperature occurring around mid-day) Díaz-Delgado

Literature Review 21

et al (2004) recommended the role of spatial fire occurrence analysis in re-examining the plant communities most affected by fire and in addressing the relationship among fire occurrence, topography, climatic parameters and human activities They also concluded that fire history reconstruction from burned area mapping allows spatial overlay with other environmental layers by means of GIS as well as the link between fire occurrence and vegetation type This can lead to the ability of ecological interpretation for both fire occurrence and fire behavior On the other hand, one of the shortcomings of fire occurrence detection from satellites is the presence of fires after image acquisition (Pleniou et al., 2012) Therefore, fire historical maps and fire risk maps also should be noticed in analysis

Burn severity is defined mainly focusing on the degree of environmental change caused

by fire In this sense, Key (2005) emphasized more on the physical and chemical alterations to the soil, conversion of vegetation and fuels to inorganic carbon and structural or compositional transformations resulting in new microclimates and species assemblages In this regard, the term ‘burn severity’ is chosen to reinforce the notion of an area where fire occurred sometime

in the past More importantly, using two distinct images from pre- and post-fire, or bi-temporal change detection might reduce the spectral confusion between burned areas and spectrally similar terrain features (e.g water, shadow, and dark soil) (Schepers et al., 2014)

2.4.7 Vegetation phenology

Vegetation phenology (or vegetation dynamics) is defined as the biological study of plantal cycle events throughout the year and the seasonal or interannual response by climate variations (Zhang et al., 2003; Didan, 2014).Therefore, the translation of vegetation phenology in remote sensing requires the consistency and length of data records To hit this spot, the combination of multiple satellite data sources in long term studies of vegetation dynamics and phenology should be carried out As a result, time series measures of MODIS enhanced vegetation index have been scrutinized the correlation of vegetation index and phenology information with reference to flux tower photosynthesis in both tropical and temperate ecosystems at seasonal scales (Huete et al., 2008) Zhang et al (2003) underscored the intra-annual vegetation dynamics with series of piecewise logistic functions based on annual time series of MODIS In addition, Reed et al (1994) accentuated 12 seasonal NDVI metrics and their phonological interpretation which can provide indicators for ecosystem dynamics In contrast, Reed et al (1994) asserted some shortcomings in the process, as (i) atmospheric corrections are not

Literature Review 22

included, (ii) it exists a bias toward high view angles at some location, and (iii) not all cloud contamination is removed

2.4.8 Net primary production

Net primary production or net primary productivity is characterized as the difference between how much carbon dioxide is taken in by plants compared to how much is put out by them As plants both consume and emit carbon dioxide through photosynthesis and respiration, the net primary production map is mainly processed through carbon cycle measurement In other point

of views, vegetation primary productivity, in general, is vital to human society with regard to essential materials provision (e.g food, fiber, and wood) and suitable environments for human inhabitation (Zhao et al., 2005) Bastiaanssen & Ali (2003) investigated vegetation primary productivity through biomass by crop yield forecasting model based on satellite measurements This model takes into account the carbon dioxide assimilation (with PAR, APAR parameters), biomass accumulation (Monteith, 1972; Field et al., 1995), SEBAL (Bastiaanssen et al., 1998)

so that it can help to cover above-mentioned net primary production elements Furthermore, MODIS, one of the primary global monitoring sensors with geolocation improvement, atmospheric correction and cloud screening provision, will be in-depth examined in this research thanks to the products supplied by MODIS team

2.4.9 Inundation

Generally, in flood inundation estimation, remote sensing data is used as an input for hydraulic modelling such as MIKE FLOOD (Patro et al., 2009), hydrological process (Zheng et al., 2008) The inputs used in the model are often the precipitation, temperature, as well as the DEM file

In addition, inundation map products are also estimated from MODIS MODIS reflectance from optical bands indicates the presence of water on land surface, previously not covered by water

A global reference database of water bodies is formed – inundation is mapped with respect to the reference water With high resolution and globally consistent, MODIS data can provide coastal inundation mapping due to storm surge or tsunamis However, MODIS provides surface inundation mapping only outside water bodies, information about water depth or water flow is not addressed In addition, surface with the presence of cloud cannot be observed (NASA, 2015)

Literature Review 23

In short, the related researches have been reviewed in this section, along with the elaboration

of the ecosystem health assessment methods and Skidmore et al (2015)’s framework Furthermore, the derivation of the remote sensing inputs as well as the approaches to retrieve biodiversity variables were discussed Accordingly, the detailed methodology will be described in Chapter 4

Case study site description 25

CHAPTER 3

Case study site description

A brief description of the case study site will be presented in this chapter, including the location and the ecohydrological characteristics This chapter aims to provide specific details on physical conditions, the status of flora and fauna as well as the shift in land use which reflects

on the new policy for sustainable development In the last section of this chapter, an overview

on previous research in study area will be introduced

3.1 Location

The study area is situated in the middle and lower Ca river basin, covering most of Nghe An province Located in 300km south of Ha Noi, the capital of Vietnam, Nghe An is considered the heart of North Central Vietnam In addtion, coastal areas can be found in the eastern part of the site and Nghe An also shared a 419 km border with Laos in the west, as can be observed in Fig 3-1 This area plays a vital role in economic development between northern and southern provinces, as well as among remote western highland and coastal areas

Figure 3-1 Location and topographic map of the case study

Case study site description 26

3.2 Ecohydrological characteristics

3.2.1 Topography, geology and soils

The study area is situated in the northeast of the Truong Son mountains The hills account for 83%

of natural landscape with inclined terrain from northwest to southeast By contrast, a small delta can be found in lower Ca river basin The highest peak is Phu Xai Lai Leng (Fig 3-2), with 2720 metres high above sea level, located on the border between Laos and Vietnam (NAG, 2013)

Figure 3-2 Location of Phu Xai Lai Leng

Source: http://www.hist-chron.com/

3.2.2 Climate

The study area is dominated by tropical monsoon climate, in particular, directly affected by the southwesterly wind (dry and hot) from April to August because of the foehn phenominon, which might extend the dry season in the study area As a result, the dry season can starts from December to May of the following year with average rainfall under 100 mm The climatogram

in Fig.3-3 indicates the mean precipitation and temperature at Vinh station, located on in the southeast of the basin The highest precipitation is recorded over 800mm in September and the annual rainfall is approximately from 1,800 to 2,000mm Besides, the mean temperature ranges between 15oC and 35oC Moreover, the mountainous areas dominate in the western part of the basin, which also lead to the alteration in rainfall and temperature For instance, the temperature

at the foothill can reach 43oC (Vietnam Statistic Head Office, 2013)

Case study site description 27

Figure 3-3 Climatogram of study area, data at Vinh station (2013)

3.2.3 Flora and fauna

The World Biosphere Reserve Western Nghe An became the sixth biosphere reserve in Vietnam, which was officially recognized by UNESCO in May 2007 with Pu Mat National Park as the main center This is the largest zone of southeast Asia with a total area of approximately 1,303,285 ha, which links three core areas including Pu Mat National Park, Pu Huong and Pu Hoat Natural Reserve Zone, as described in Fig.3-4 In this regard, this zone helps to reduce habitat fragmentation, creates habitat continuity and maintain biodiversity conservation Accordingly, the study site covers two thirds of the reserved area, consisting of

Pu Mat and Pu Huong with their buffer and transition zones

Figure 3-4 World Biosphere Reserve Western Nghe An

Source: www.kiemlam.org.vn

Case study site description 28

As reported by Nghe An Natural Resource Department (2013), there are 972,910.52 ha

of forestry, including 501,634.85 ha of production forest, 302,067.47 ha for protection, and 169,207.2 ha of special used forest (genes conservation and recreation) With a current total timber volume about 50 million m3, it is a large resouce of raw materials for forest exploitation and forestry-based industrial development

Besides, the diverse ecosytem in the study area witnesses 1,513 species of vascular

plants (Cryptocarya metcalfiana, Neohouzeaua,Castanopsis ferox, Quercus glauca, ) and about 241 fauna species (with species endemic to Vietnam and Laos: Pygathrix

nemaeus, Pseudoryx nghetinhensis, Jabouilleia danjoui, …), as illustrated in Fig.3-5

Quercus glauca Neohouzeaua Castanopsis ferox

Pseudoryx nghetinhensis Pygathrix nemaeus Jabouilleia danjoui

Figure 3-5 Representative flora and fauna in study area

Case study site description 29

3.2.4 Land use and land cover

Year 2013

Figure 3-6 The trend in land use structure from 2000 to 2013

The pie charts illustrate the alteration in proportion of land use recorded in three different years over a 13-year period in the study site The most significant types of land use were forestry and other land, which together accounted for more than 50% of land By constrast, the specific used and homestead land encoutered least proportion in each year In addition, there was a marginal change in both types of land use during 2000 and 2013 On the other hand, a slight increase was observed in the proportion of agricultural production land, from 11.9% in 2000 to 17% in 2013 whereas the forestry land witnessed a steady rising, by 18% in 2013 Finally, the percentages

Case study site description 30

of other land fell dramatically, by 23% over years In conclusion, it is apparent that there was a shift from other land to forestry and agricultural production land in the study area

Figure 3-7 Land Use map of Ca River Basin, Nghe An, Vietnam (2012)

Source: Poortinga (2015) The classification of land cover in the study area is illustrated in Fig.3-7 Forests dominate two thirds of study area, including forest plantation, secondary forest, mangrove, bamboo and broadleaved forests; whereas rice paddies and crops are situated in the northeast

of the basin

3.2.5 Pressures, threats and current outlook

Rising in the mountains of Laos, the Ca river flows through Nghe An and Ha Tinh in Vietnam and empties into the sea More than three-fifths of the catchment area is situated on the territory of Vietnam Accordingly, rural development and biodiversity conservation are the main concerns in the Ca river basin as well as the ongoing process of the land use shifting, forest allocation As a result, the river basin faced a deforestation rate of around 4% per year (Brunner & Nielsen, 1998) It was believed that the unsufficient development programs related

to immigrants and unplanned exploitation were the major causes of forest degradation rather than shifting cultivation In recent years, paralleled with many domestic research on land policy and farming system alteration, economical requirements of the local people (Rambo & Le, 1998; Castella & Dang, 2002; Leisz, 2009), many attempts using remotely sensed imagery are