Remote sensing of ecosystem services: A systematic review

R

R

Ca

Ub

a

ARRA

KRESBE

C

h1

Ecological Indicators 52 (2015) 430–443

Contents lists available at ScienceDirect

Ecological Indicators

journa l homepage: www.e lsev ier .com/ locate /eco l ind

eview

emote sensing of ecosystem services: A systematic review

aio C. de Araujo Barbosaa,∗, Peter M. Atkinsona, John A. Dearingb

Global Environmental Change and Earth Observation Research Group, Geography and Environment, University of Southampton, Southampton SO17 1BJ,nited KingdomPalaeoecological Laboratory, Geography and Environment, University of Southampton, Southampton SO17 1BJ, United Kingdom

r t i c l e i n f o

rticle history:eceived 2 July 2014eceived in revised form 4 January 2015ccepted 5 January 2015

eywords:emote sensingcosystem servicesystematic reviewiophysical proxiesarth observation

a b s t r a c t

Appropriate integration of remote sensing technologies into ecosystem services concepts and practicesleads to potential practical benefits for the protection of biodiversity and the promotion of sustainable useof Earth’s natural assets. The last decade has seen the rapid development of research efforts on the topicof ecosystem services, which has led to a significant increase in the number of scientific publications.This systematic review aims to identify, evaluate and synthesise the evidence provided in published peerreviewed studies framing their work in the context of spatially explicit remote sensing assessment andvaluation of ecosystem services. Initially, a search through indexed scientific databases found 5920 papersmaking direct and/or indirect reference to the topic of “ecosystem services” between the years of 1960 and2013. Among these papers, 211 make direct reference to the use of remote sensing. During the search weaimed at selecting papers that were peer-reviewed publications available through indexed bibliographicdatabases. For this reason, our literature search did not include books, grey literature, extended abstractsand presentations. We quantitatively present the growth of remote sensing applications in ecosystemservices’ research, reviewing the literature to produce a summary of the state of available and feasibleremote sensing variables used in the assessment and valuation of ecosystem services. The results providevaluable information on how remotely sensed Earth observation data are used currently to produce
spatially-explicit assessments and valuation of ecosystem services. Using examples from the literaturewe produce a concise summary of what has been done, what can be done and what can be improvedupon in the future to integrate remote sensing into ecosystem services research. The reason for doingso is to motivate discussion about methodological challenges, solutions and to encourage an uptake ofremote sensing technology and data where it has potential practical applications.
© 2015 Elsevier Ltd. All rights reserved.

ontents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4312. Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 431

2.1. Search and selection strategy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4313. Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 432

3.1. Systematic review: what has been done. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4323.1.1. Numbers and spatial distribution of research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4333.1.2. Remote sensing proxy data used in ecosystem services research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 434

4. Potential of remote sensing: what can be done . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4365. Discussion: suggestions for improvements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 438

5.1. Correcting data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5.2. Land cover . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5.3. Vegetation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

∗ Corresponding author. Tel.: +44 023 80592258.E-mail addresses: [email protected],[email protected] (C

ttp://dx.doi.org/10.1016/j.ecolind.2015.01.007470-160X/© 2015 Elsevier Ltd. All rights reserved.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 438

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 438

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 439

.C. de Araujo Barbosa).

dx.doi.org/10.1016/j.ecolind.2015.01.007

http://www.sciencedirect.com/science/journal/1470160X

http://www.elsevier.com/locate/ecolind

http://crossmark.crossref.org/dialog/?doi=10.1016/j.ecolind.2015.01.007&domain=pdf

mailto:[email protected]

mailto:[email protected]

dx.doi.org/10.1016/j.ecolind.2015.01.007

C.C. de Araujo Barbosa et al. / Ecological Indicators 52 (2015) 430–443 431

5.4. Spatial scale . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4395.5. Temporal extent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4395.6. Cultural ecosystem services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 439

6. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 440Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 440

. . . . . .

1

tem2eMioC

its22wavKS

eaecU2sataa2utgM2

veiheNcfcd(2

uo

(

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. Introduction

The last decade has seen the rapid development of research onhe topic of ecosystem services and, increasing awareness of theconomic value of ecosystem goods and services among decision-akers and the general public (Morgan et al., 2008; Fisher et al.,

009; de Groot et al., 2010; Fish, 2011; de Groot et al., 2012; Maest al., 2012; Bagstad et al., 2013). Following the release of theillennium Ecosystem Assessment in 2005, a significant increase

n the number of scientific publications on the subject has beenbserved (Egoh et al., 2007; Skourtos et al., 2010; White et al., 2010;ostanza and Kubiszewski, 2012).

Preliminary work has made considerable advances in highlight-ng the links between social and environmental change influencinghe capacity of ecosystems to maintain provisioning, regulating,upporting and cultural services (Costanza et al., 1997; Imhoff et al.,004; de Groot, 2006; Reid et al., 2006; Deegan et al., 2012; Porras,012). Since then, spatially-explicit approaches have been usedidely to map a multitude of ecosystems services, merging theory

nd practice to advance more effective sustainability and conser-ation actions (Sutton and Costanza, 2002; Costanza et al., 2008;ienast et al., 2009; Liu et al., 2010; Murray et al., 2012; Alcaraz-egura et al., 2013; Xie and Ng, 2013; Alamgir et al., 2014).

Remote sensing plays an important role in the study of complexnvironmental interactions between natural and social systems,nd has been used widely to quantify and map ecosystem prop-rties and functions and infer ecosystem processes through aombination of existing instruments and data (Chopra et al., 2001;stin et al., 2004; Chambers et al., 2007; Muraoka and Koizumi,009; Palacios-Orueta et al., 2012). A key advantage of remoteensing is the capability to perform synoptic, spatially continuousnd frequent observations resulting in large data volumes and mul-iple datasets at varying spatial and temporal resolutions (Estreguilnd Lambin, 1996; Stolle et al., 2004; Knaeps et al., 2010; Baraldind Boschetti, 2012; Atzberger and Rembold, 2013; Lewis et al.,013). With the emergence of new and more sophisticated prod-cts, Earth observation data will continue to contribute extensivelyo research on modelling, mapping and valuation of ecosystemoods and services (Nemani et al., 2009; Verburg et al., 2009;arghany and Hashim, 2010; Zinnert et al., 2011; Cabello et al.,

012).There have been numerous efforts to classify the services pro-

ided by ecosystems (Costanza et al., 1997; Daily, 1997; de Groott al., 2002; MA, 2005), and more recent efforts on understand-ng the various contexts in which the ecosystem services conceptas been used have the potential to move a step closer to thestablishment of a meaningful and clear classification system.evertheless, there has not yet been agreement on a single classifi-ation scheme that provides a meaningful and consistent definitionor ecosystem services, with different countries adopting differentlassification schemes, making assessment and valuation studiesependent on realities that are in many occasions site specific.Faber, 2006; Boyd and Banzhaf, 2007; Wallace, 2007; Schmitt et al.,
013).
Remote sensing measures reflected radiation, and by makingse of such measurements and a forward model of the interactionf light with the Earth’s surface, it is possible to inversely estimate

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 440

properties of the Earth’s surface such as land cover type, biomass,and leaf area index for each image pixel. Such properties repre-sent spatially varying states at particular points in time. However,other properties of interest (e.g., land use; the function of the land)can only be inversely and indirectly inferred and mapped throughcontextual relations across space.

Some ecosystem services, like primary production have a moredirect relation with the reflected signal (e.g. satellite-derivedchlorophyll measurements), while others, like climate regulationare only indirectly related to the reflected signal (e.g. satellite-derived land surface temperature). Many economically importantservices, such as production of raw materials and food provisioncannot be valued and assessed sufficiently by remote sensing aloneand it may, therefore, be necessary to couple spectral informationwith other available datasets (Fig. 1).

This paper provides a systematic review of the scientific litera-ture related to the use of remotely sensed data within the explicitcontext of ecosystem services valuation assessment. We presentexamples from the literature that summarise (i) what has beendone so far using remote sensing, (ii) what can be done in thefuture, and (iii) what can be further improved. For this purposethe selected literature was reviewed systematically, searching forspecific information regarding the remote sensing platform used,remote sensing product used (e.g. land cover, biophysical indices,etc.) temporal coverage, and other relevant more technical aspectsof remote sensing.

2. Methodology

2.1. Search and selection strategy

This review focused on papers that used ecosystem services con-cepts, and made use of Earth observation data to estimate valueflows of past and current availability of provisioning, regulating,habitat/supporting and cultural services.

The approach taken to querying the literature consisted of aselective keyword search in specific scientific libraries (such as Sco-pus and Web of Knowledge). At each query, terms and keywordssuch as ‘Ecosystem Services’, ‘Ecological Services’, ‘EnvironmentalService’ ‘Remote Sensing’ and ‘Earth Observation were used indi-vidually to produce an extensive list of articles. Using Scopus andWeb of Knowledge the body of literature was searched based on afixed set of inclusion criteria:

(i) The literature should address ecosystem services as eithermain or secondary subject.

(ii) The predefined keywords should exist as a whole in at leastone of the fields: title, keywords or abstract.

iii) The paper should be published in a scientific peer-reviewedjournal.

(iv) The paper should be written in the English language.

Fig. 2 represents the literature search and selection process.
The initial steps of the search process in Fig. 2 returned 5920
published articles on the topic of ecosystem services alone. Thesearch was subsequently refined by querying through the first set ofresults. Searching for predefined keywords and terms related to the

432 C.C. de Araujo Barbosa et al. / Ecological Indicators 52 (2015) 430–443

data i

t2stssrbnt

ts2irhs

Fig. 1. The transform of remotely sensed

opic of remote sensing, the total number of entries was reduced to11 scientific publications. During the data extraction process, theelected publications were scanned to identify and extract informa-ion in the form of descriptive text and identify labels for automatedorting using a custom bibliographic matrix. During our literatureearch we aimed at finding studies meeting the criteria of a peer-eviewed publication, and it should be available through indexedibliographic databases. For this reason, our literature search didot include books, grey literature, extended abstracts and presen-ations.

Throughout the exercise, ecosystem services found withinhe publications were categorised according to the classificationcheme proposed by the Millennium Ecosystem Assessment (MA,005). The MA classification scheme has intrinsic challenges, when

t comes to examining conditions and changes in ecosystems beingeferenced to more than one boundary definition. While thereave been numerous attempts to create a definitive classificationcheme for ecosystem services, there has not been yet an agreement

Fig. 2. Simplified literature search

nto ecosystem services values and flows.

upon which scheme offers a more meaningful and consistent defi-nition (de Groot et al., 2002; Boyd and Banzhaf, 2007; Fisher et al.,2009). The option for the MA classification scheme comes from thisbeing a widely recognised classification scheme that provided acomprehensive assessment of the state of the global environment(MA, 2005).

3. Results

3.1. Systematic review: what has been done.

Total ecosystem services value depends not only on the nat-ural phenomenon and location, but also on context and moresubjective values related to societal wellbeing and cultural percep-
tions. In this context, current mapping methods based on remotesensing data are not sufficient alone for quantifying the full rangeof ecosystem services. At the same time, integration of ecosystemservices valuation methods and remote sensing observation seems
and selection flow process.

ologica

tutsflTers

ehttsoA2

c(FIgcf22

raenadeMB

vBotr(miiw

babtgtbma

cameew

C.C. de Araujo Barbosa et al. / Ec

o be the only practical and realistic approaches to regionalise ournderstanding of complex interdependent socio-ecological sys-ems. Where estimation is possible, translation from remoteensing data into estimates of certain ecosystem services values andows may require data other than the spectral information itself.herefore, if remotely sensed data are to be used to predict multiplecosystem services it is required first to identify the nature of theelationship between the spectral information and the ecosystemervice.

Integration of remote sensing data in the explicit context ofcosystem services assessment has enabled numerous studies onow different social and ecological systems are interacting overime. For example, using land-use/land-cover classifications fromime-series of Landsat images many studies were able to mea-ure deforestation rates and, thus, estimate changes in provisioningf raw materials and genetic materials, for example, in China,rgentina and the United Sates (Wang et al., 2006; Volante et al.,012; Klepeis et al., 2013).

Similarly, remote sensing data classified into land cover classes,ombined with biophysical variables such as vegetation indicesVIs), are often used to quantify and map food production (e.g.egraus et al., 2012; Johnson et al., 2012; Castella et al., 2013).n this context, vegetation indices are used as a proxy for croprowth during the growing season, and then coupled with landover and rainfall estimates to predict crop yield and enableorecasts of food production (e.g. Fuller, 1998; Konarska et al.,002; Gong et al., 2010; Ivits et al., 2012; Palacios-Orueta et al.,012).

Muukkonen and Heiskanen (2005), in their study, usedeflectance data from the Advanced Spaceborne Thermal Emissionnd Reflection Radiometer (ASTER) together with a standwise for-st inventory to estimate the biomass of boreal forest stands, usingon-linear regression analysis and neural networks. Merging of VIsnd land cover data may also be useful in creating models for pre-iction of the spatial distribution of species and biodiversity tostimate the availability of genetic materials (Dalezios et al., 2001;oriondo et al., 2007; Saatchi et al., 2007; Nemani et al., 2009;

arman et al., 2010).Using land cover classifications produced by applying a support

ector machine classification to time-series of Landsat imagery,randt et al. (2012) investigated how changes in forest type, lossf old-growth forest and ecotourism development were affectinghe availability of provisioning services (raw and genetic mate-ial), habitat services (biological refugia) and regulating servicespollination and biological control). They found that changes in

anagement practices and politics had a great influence in increas-ng and decreasing logging rates of high-diversity old-growth forestn the biodiversity hotspot of northwest Yunnan Province in south-

est China.Krishnaswamy et al. (2009) investigated the relationship

etween the spectral distances and tree species diversity usingquantile–quantile plot method as an alternative to regression-

ased approaches. Spectral distances were used in modelling forestype variability assuming that the larger the spectral distance thereater the species diversity within sites. A scale-dependent rela-ionship was found between spectral and the floristic distancesetween areas. This approach can be applied to quantify and mapultiple ecosystems services related to habitat, biodiversity, water

nd carbon storage.Combination of landscape metrics and time-series of land

over classifications from the Landsat Thematic Mapper (TM)nd the Moderate Resolution Imaging Spectroradiometer (MODIS)
atched with the proxy biomes from Costanza et al. (1997), has
nabled numerous studies (e.g. Sutton and Costanza, 2002; Zhaot al., 2004; Chen et al., 2007; Zhang et al., 2007; Niu et al., 2012),ith ecosystem service values (ESV) being associated to land cover

l Indicators 52 (2015) 430–443 433

classes assuming a linear relationship between land cover and mar-ket prices per unit area.

Other studies, aiming at quantifying and mapping the spatio-temporal variation in ecosystem services values, based theirmethodology on time-series of land cover data from satellite sen-sor imagery coupled with landscape fragmentation and urbansprawl analysis (Kreuter et al., 2001; Su et al., 2012; Estoque andMurayama, 2013), econometric calculations (Niu et al., 2012), his-torical data (Cai et al., 2013) and meta-analysis of socio-economicvariables (Camacho-Valdez et al., 2013).

Integration of suspended particulate matter concentration andaerosol optical thickness data (from MODIS-Aqua and MODIS-Terra), land cover (from Landsat and MODIS-Terra), nutrient cycling(from SeaWifs) and NDVI (From Landsat and MODIS-Terra) hasbeen used to measure spatial and seasonal patterns of water pro-vision and flood regulation services (Muthuwatta et al., 2010;Ghobadi et al., 2012), erosion regulation, climate regulation andmaintenance of habitat services (Chopra et al., 2001; Gong et al.,2010; Ke et al., 2011) in India, China, Uganda, Iran, Mexico and theUnited States.

Furthermore, Nedkov and Burkhard (2012) have used waterretention functions (of the vegetation and soil cover), digital ele-vation models and land use information to assess the capacities ofdifferent ecosystems to regulate floods. The resulting maps repre-sent estimates of regional supply-demand balances converted intothe ecosystems’ flood regulating service capacity for the Malki Iskarriver basin in northern Bulgaria.

Multi-date imagery from the Sea-Viewing Wide Field-of-View Sensor (SeaWiFS), MODIS and Landsat sensors has beendemonstrated to be effective in measuring and monitoring waterpurification, erosion regulation and natural hazard regulationservices though estimates on the cumulative effects of shiftsin size and spatial configuration of coastal habitats, vegetationcover, surface temperatures, coastal water salinity, chlorophyll,biomass and total suspended matter of marine ecosystems(Doron et al., 2011; Cloern and Jassby, 2012; Paerl and Paul,2012).

3.1.1. Numbers and spatial distribution of researchAfter the literature search, we found that the number of pub-

lished research papers referring directly to the field of ecosystemservices grew from 115 in 2005 to 5920 in 2013, denoting the rapiddevelopment of research within the field. The exponential growthof scientific publications in the field of ecosystem services researchwas followed similarly by an exponential increase in the number ofpublished papers integrating Earth observation data in ecosystemservices research (Fig. 3).

The number of published research papers referring directly toecosystem services and remote sensing has also increased signifi-cantly. The percentage increase in remote sensing and ecosystemservices integrated studies relative to the total scientific productionin the field of ecosystem services is presented in Fig. 4.

To identify biophysical and socio-economic drivers of changein ecosystem service supply and demand, as well as long- andshort-term trends, it is essential that the temporal coverage be ofsufficient length and resolution. In particular, to monitor effectivelythe effects of time-sensitive socio-ecological interactions, it is nec-essary to use a remote sensing platform with a sufficiently frequentrevisit period. Nevertheless, this is not an easy task since the avail-ability of remote sensing data for long-term monitoring purposesis constrained by sensor characteristics (e.g. revisit time) and envi-
ronmental factors (e.g. cloud cover). From Fig. 5 it is clear that thegreat majority of published material has focused on a small tempo-ral extent. Several articles analysed more than 20 years temporalextent, but the number still falls short of what is required to perform


2013201220112010200920082007200620052004200320022001

0

5

10

15

20

25

30

35

40

45

50

Pu

blis

he

d p

ap

ers

Fig. 3. Number of papers published annually between 2001 and 2013. This plot represents only those research articles which integrate Earth observation data into the explicitcontext of ecosystem services research.

2013201220112010200920082007200620052004200320022001

0

1

2

3

4

5

6

Ye

arly p

erc

en

tag

e

Actual data

Trend Line

F ces reT

ic

aisge

gsivE

ig. 4. Percentage of publications integrating remote sensing and ecosystem servihe fitted regression line clearly indicates a positive trend.

ntegrated analyses, aiming at continuous long-term monitoring ofhanges in ecosystem services supply and demand.

The ecosystem services (and respective categories of servicess defined by the Millennium Ecosystem Assessment (MA, 2005))dentified within the selected literature are presented in Fig. 6 inuch a way that it is possible to visualise which services had areater presence in research integrating Earth observation data intocosystem services research.

From the map in Fig. 7 it is evident that considerable gaps ineographical focus still exist around the world especially, rather
urprisingly, in the African continent. Other areas lacking studiesntegrating remote sensing into ecosystem services assessment andaluation exist in Eastern Europe, Central America and the Middleast.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

0

2

4

6

8

10

12

14

16

18

20

22

24

26

Pe

rce

nta

ge

of p

ub

lica

tio

ns

Temporal exte

Fig. 5. Temporal extent of studies identi

lative to the total scientific production in the field of ecosystem services annually.

3.1.2. Remote sensing proxy data used in ecosystem servicesresearch

Figs. 8–11 show how published works have used remotelysensed biophysical properties to map and quantify provisioning(Fig. 8), regulating (Fig. 9), habitat/supporting (Fig. 10) and cul-tural (Fig. 11) ecosystem services. The figures indicate that thebiophysical properties used generally differ between each categoryof ecosystem service.

For provisioning services (Fig. 8), the categories most well cov-ered are water provision and food provision where seven remote
sensing proxies were used; the least well represented categories aregenetic resources and energy where only two proxies were used,respectively. The dominant ecosystem services proxy variables arerespectively land cover (x̄ ∼= 58%) and NDVI (x̄ ∼= 35%).
21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40

nt (years)

fied within the literature selected.


RecreationCultural Heritage

Soil formationPrimary production (Ocean/River)

Primary productionNutrient cycling

HabitatBiological RefugiaWater regulation

Water purification and waste treatmentPollination

Natural hazard regulationGas regulation

Flood regulationErosion Regulation

Disturbance regulationClimate regulation

Biological controlWater Provision

Raw MaterialsGenetic Resources

FoodEnergy

1614121086420Percentage of published papers

CulturalHabitat/SupportRegulationProvision

vision

cwurela

(t

Fig. 6. Percentage of publications using remote sensing to assess pro

For regulating services (Fig. 9), climate regulation, water purifi-ation and waste treatment are the most well covered categorieshere, respectively, eight and five remote sensing proxies weresed. Important services such as biological control, disturbanceegulation and gas regulation are the least well represented cat-gories where only three remote sensing proxies were used. Again,and cover (x̄ ∼= 47%) is appears as the dominant proxy variable used
mong all the categories, followed by NDVI (x̄ ∼= 20%).
For habitat/supporting services (Fig. 10), primary productionfor marine and riverine environments) and biological refugia arehe most well covered categories where seven remote sensing

Fig. 7. Percentage of total published papers per study site integrating sa

ing, regulating, habitat/supporting and cultural ecosystem services.

proxies were used. Soil formation, primary production (land) andnutrient cycling are the least well represented categories with onlythree remote sensing proxies used. Again, land cover (x̄ ∼= 32%)is the dominant proxy variable used among all the categories ofecosystem services, although LAI (x̄ ∼= 13%) and NDVI (x̄ ∼= 7%) bio-physical data have a significant presence.

Estimates of cultural services are often used among stakehold-
ers for planning, conservation, and environmental management.Nevertheless, there seems to be a lack of remote sensing proxydata in use (Fig. 11), with recreation services being covered byonly three proxy variables, and recreation services being covered
tellite remote sensing data within ecosystem services assessment.


NDVI

Land cover

Sea Surface Temperatures

Land Surface Temperatures

LAI

GPP

Chl-a concentration

Land cover

Sea Surface Temperatures


LAI

GPP

Chl-a concentration

Energy Food Genetic Resources

Raw materials Water provision

34%

66%

24%

45%

12%

6%

3%

7%

3%

15%

70%

15%

56%

4%

4%

9%

4%

55%

4%

4%

4%

4%

9%

ected

balimp

NDVI 23% 19%

Fig. 8. Remote sensing variables used within the sel

y only two remote sensing proxies. In the graph, land coverppears as dominant among all ecosystem services categories, fol-
owed by NDVI. In the case of cultural ecosystem services, it isn many ways challenging to produce satellite measurements of
ore subjective values related to societal wellbeing and culturalerceptions.

CDOM indexLAI

Land coverWater Layer

Suspended solidsSea Surface Temp.

NDVILand Surface Temp.

Elevation dataChl-a concentration

CDOM indexLAI


Suspended solidsSea Surface Temp.

NDVILand Surface Temp.

Elevation dataChl-a concentration

Biological Control Climate Regu

Erosion Regulation Flood Regula

Natural hazard regulation Pollinatio

Water Regulation

17% 33%

50%

2% 3%

36%

2% 2%

42% 11%

2%

60%

10%

20%

10%

33%

17%

17%

33%

40% 20%

20%

20%

33.33% 33.33%

33.33%

52% 16% 16%

16%

Fig. 9. Remote sensing variables used within the selected

literature to assess ecosystem provisioning services.

4. Potential of remote sensing: what can be done

Climate regulating services (e.g. carbon storage and sequestra-tion) can be estimated through quantification of net ecosystemexchange (NEE) of CO2 flux, as NEE can be applied to deter-mine the amount of atmospheric carbon stored in an ecosystem

CDOM indexLAILand coverWater LayerSuspended solidsSea Surface Temp.NDVILand Surface Temp.Elevation dataChl-a concentration

CDOM indexLAILand coverWater LayerSuspended solidsSea Surface Temp.NDVILand Surface Temp.Elevation dataChl-a concentration

lation Disturbance Regulation

tion Gas Regulation

n Water purif. & waste treatment

60%

30% 10%

70%

20% 10%

35% 16%

16% 16% 17%

literature to assess ecosystem regulating services.


Thermal infraredLand Surface Temp.GPPBathymetryNDVILAILand coverWater LayerSea Surface Temp.Elevation dataChl-a concentration

Thermal infraredLand Surface Temp.

GPPBathymetry

NDVILAI


Sea Surface Temp.Elevation data

Chl-a concentration

Biological Refugia Habitat Nutrient Cycling

Primary production (Sea/River) Primary production (Land) Soil formation 8% 8.4% 8.6% 8.8%

16.6% 16.6%

33%

9.2%

9%

9.5%

54%

9%

9.3%

11% 33%

5% 12% 11%

28%

26%

12%12%

26% 12%

12%

34%

34% 16%

16%

15% 57%

14%

14%

ed lite

(Aftfl2

ictitvSe

sgafw(eaaotS

l

Fe

Fig. 10. Remote sensing variables used within the select

Van Tuyl et al., 2005; Soojeong et al., 2006; Potter et al., 2008).dditionally, vegetation indices such as the NDVI, emissivity dif-

erence vegetation index (EDVI) and water band index, as well ashe leaf area index (LAI), can also be used as indicators of net CO2ux (La Puma et al., 2007; Dagg and Lafleur, 2010; Olivas et al.,010; Hao et al., 2012; Kross et al., 2013).

Air quality regulation as the capacity of atmosphere to cleansetself, can be estimated by using trace gases and aerosols as indi-ators, since aerosol data, coupled with simulation models, enablehe source of polluting air to be identified. Additionally, a changen the amount of dust build-up in the air will change the spec-ral reflectance of objects. Thus, pollutants can be detected fromariation in the spectral reflectance (Chu et al., 2003; Emeis andchaefer, 2006; Yong et al., 2010; Roots et al., 2011; Mozumdert al., 2013).

Maintenance of soil fertility can be estimated from remotelyensed indices such as the non-photosynthetic vegetation (NPV),reen vegetation (GV), NDVI and soil spectral mixture analysis, asn increase in NPV is suggested to be related to a decline in soilertility (Numata et al., 2003). Moreover, reflectance data (withinavelengths 0.425–0.695 �m) can be related to soil organic matter

SOM) and used as a proxy for soil fertility (Numata et al., 2003; Jit al., 2012; Liu et al., 2013a). Other mineral features present in soilsnd detectable by remote sensing (e.g. salinity, iron-oxide content,nd heavy metal contamination) change the reflectance propertiesf soils on bare lands and the surrounding vegetation indicating
he past and current status of soils (Taylor et al., 2001; Kemper andommer, 2002; Wu et al., 2005; Melendez-Pastor et al., 2010).
Erosion control regulation can be measured with medium reso-ution optical imagery (e.g. Landsat Thematic Mapper and Landsat

NDVI


Land cover

Cultural Heritage Recreation

40%

60%

17%

33%

50%

ig. 11. Remote sensing variables used within the selected literature to assesscosystem cultural services.

rature to assess ecosystem habitat/supporting services.

Enhanced Thematic Mapper) by identifying the reflectance proper-ties of the constituents of sediments (e.g. lithological composition,grain size and moisture content) potentially enabling the detectionof eroded land and material deposition (Small et al., 2009). Veg-etation indices and elevation data are suitable for predicting soilerosion risk, as mapping variability in vegetation cover and plantresidue can help in revealing areas more or less prone to erosion.Remote sensing-based soil erosion models integrate NDVI, vegeta-tion fraction cover, slope gradient and land use (from SRTM data) toestimate annual soil erosion rates (Hochschild et al., 2003; Rahmanet al., 2009; Wu et al., 2011; Wang et al., 2013b).

The diffuse attenuation coefficient (KPAR) for photosynthet-ically active radiation (PAR) can be utilised to assess waterpurification and waste treatment in ecosystems; since the avail-ability of light is fundamental for phytoplankton and sea grassproduction, light attenuation can be used as a measure of turbidityor water clarity (Chen et al., 2009; Liu et al., 2013b). Thus, cor-relating satellite-derived reflectance data with in situ measuredlight attenuation enables assessment of water clarity and, hence,the water purification capacity of ecosystems (Behrenfeld and Boss,2006; Zhu et al., 2009; Zhao et al., 2011).

Earth observation data can be used to quantify the productioncapacity of forests and agro-ecosystems using biomass as an indi-cator. Narrowband and broadband vegetation indices such as NDVI,EVI, fPAR and LAI can be used as indicators of productivity in a cropgrowing season, as they are able to characterise variation in phen-ology and photosynthetic potential of crops and help identify thecropping cycle and growth (Kastens et al., 2005; Muukkonen andHeiskanen, 2005; Brown and de Beurs, 2008; Wall et al., 2008a;Wardlow and Egbert, 2008; Ling et al., 2009; Prabakaran et al.,2013).

The capacity of ecosystems to deliver services such as the pro-vision of energy through solar power can be identified using dailysurface insolation data and atmospheric transmittance, extractedfrom optical and thermal infrared bands, in addition to modelsof aerosol optical depth and columnar ozone (Perez et al., 1994;Roujean, 1998; Morelli et al., 2012; Bhattacharya et al., 2013). Incoastal regions, the capacity to provide wind energy can be esti-
mated from radar imagery such as synthetic aperture radar (SAR),as it provides spatial patterns of wind speed, useful in mappingoffshore wind resources (backscatter coefficients are physicallyrelated to wind speed by observing the roughness of the open sea)

4 ologica

(2

avnSnetdtssaR

ssSscpetsufa

tncae2lutvsmsB

alt2aacpsaB

5

ipcsi

38 C.C. de Araujo Barbosa et al. / Ec

Johannessen et al., 1999; Hasager et al., 2002; Kozai and Ohsawa,007; Xiulin et al., 2013).

Potential mass flow and rain-induced shallow landslide prob-bility can be predicted using remotely sensed indices, such asegetation indices, the soil brightness index, principal compo-ents transformation and terrain properties (Hyun-Joo et al., 2012;huwen et al., 2012; Yang et al., 2013) and applied in assessingatural hazard regulating services. Furthermore, radar interferom-try (using SAR images) can be applied with image correlationechniques to generate displacement maps (e.g. the Line of Sightirection and the Azimuth direction), resulting in informationhat can be combined routinely to undertake landslide groundurface deformation monitoring by measuring the three dimen-ional surface displacement field at different epochs (Colesantind Wasowski, 2006; Noferini et al., 2007; Righini et al., 2012;aucoules et al., 2013; Tantianuparp et al., 2013).

Remote sensing techniques applied to differentiate the compo-ition of landscape/seascape can enable quantification supportingervices such as nutrient and water cycling (Asner et al., 2004;pillman et al., 2007). Water quality parameters such as dis-olved oxygen, chlorophyll-a and turbidity are closely linked withhanges in absorbance of natural radiation and can be detected byassive remote sensors(Xie et al., 2007; Wang et al., 2013a; Lint al., 2014). Phosphorous concentration levels can also be used torack spatio-temporal water and nutrient cycling patterns acrosseascapes(Aighewi et al., 2013). Other remote sensing variablessed to estimate nutrient cycling are related to measures of forestragmentation and homogeneity, soil moisture and surface temper-tures (Karl, 2002; Chang and Xuan, 2011).

Monitoring and predicting wind storms is of particular impor-ance in assessing the capacity of ecosystems to mitigate theegative effects of storms on coastal areas, croplands and asso-iated infrastructure. Using satellite sensor data it is possible toccomplish this task; for example, using the MODIS reflective andmissive bands and artificial neural networks (Brakenridge et al.,013; El-ossta et al., 2013; Notaro et al., 2013). Moreover, satel-

ite sensor observations of damage caused by storms can also besed to assess an ecosystem’s ability to regulate wind storms, sincehe damage extent on infrastructure and vegetation will most likelyary with the wind speed and strength. Valuable information abouttorm events (e.g. the trends in moderate-duration storms andean sea level) can also be derived from TOPEX/Poseidon and Jason

atellite altimetry data (Sadhuram et al., 2006; Han et al., 2012;haskaran et al., 2013)

Combinations of remote sensing data, stochastic and multi-gent models have clear advantages in modelling and simulatingand cover change, soil erosion, food production, and vegeta-ion dynamics (Evans and Kelley, 2008; Aitkenhead and Aalders,011; Linard et al., 2013). Agent-based and stochastic modellingre potentially suitable approaches for expanding our knowledgebout landscape functioning, climatological variability, biophysi-al constraints, economic and social dynamics. Therefore, these areotentially capable of providing insights into how highly complexystems behave and interact, and how they impact on the avail-bility of ecosystem services (Jepsen et al., 2006; Wada et al., 2007;rown et al., 2008; Evans and Kelley, 2008; Frank et al., 2012).

. Discussion: suggestions for improvements

The use of remote sensing within ecosystem services researchs still limited, and up until 2009 this topic had a relatively small
resence in the literature. Nevertheless, the general trend indi-ates that the number of scientific publications integrating remoteensing into ecosystem services studies is increasing rapidly. Thiss no surprise given the possibilities for integrating spatial data
l Indicators 52 (2015) 430–443

into ecosystem studies and establishing relationships betweenremote sensing data and socio-ecological phenomena. The increas-ing exponential rate of integrated scientific studies is expandingour knowledge and pointing to areas for future development. Inthis section, we consider some of the weaknesses of the reviewedresearch studies and suggest improvements.

5.1. Correcting data

A few points of connection were found relevant within theselected publications. For example, in 85% of the publications therewas no clear indication of whether the effects of the atmosphereon the reflectance values was taken into account prior to the anal-ysis of the remotely sensed data for ecosystem services assessmentand valuation (e.g. land cover classification process). 15% of sci-entific papers claimed to have performed atmospheric correction,although no reference was made to the algorithm applied. Manyindividual studies focused on mapping changes in the delivery ofecosystem services through time. However, to maximise the ben-efits of doing so, change detection should be done automatically inan operational environment, and the resulting information shouldbe sufficiently reliable that it can be related effectively to what isobserved on site (Beresford, 1990; Petit and Lambin, 2001; Songet al., 2001; Coppin et al., 2004; Sesnie et al., 2008; Zeng et al.,2012). This implies an operational atmospheric correction proce-dure to correct for the effects of aerosols, sun glint, thin clouds,cloud shadows and adjacency effects (Lambin, 1996; Cairns et al.,2002; Elmahboub et al., 2006; Vanonckelen et al., 2013).

Where correcting for atmospheric effects is mandatory, even asimple atmospheric correction algorithm such as the dark objectsubtraction can increase the accuracy of land cover classificationand change detection results (Moran et al., 1992; Song et al., 2001).A total of 28% of selected articles used single date imagery to assessecosystem services; in these cases the application of an atmo-spheric correction algorithm may not be as essential as if imageclassification were to be performed for multiple dates (Kawata et al.,1990).

5.2. Land cover

Remote sensing for the assessment and valuation of ecosys-tem services is still highly dependent on land cover data usedas proxies for the actual ecosystem service. Land cover hasbeen used as the main source of remotely sensed data in themajority of articles, serving as proxy data to spatial value transfer-based modelling approaches, which have been demonstrated tobe highly scale-dependent, producing inaccurate estimates whencompared to primary data (Eigenbrod et al., 2010). By assum-ing a ESV = Area * Market price relationship, the valuation exercisewill produce ecosystem services values that rely largely on thearea measurements extracted from the satellite imagery. There-fore, ecosystem services valuation efforts in the context of remotesensing will be highly dependent on spatial resolution and imageclassification accuracy (Cai et al., 2013).

Linking ecosystem functions and benefits to human well-beingrequires the knowledge of where the benefits of ecosystem ser-vices are manifest. Mapping the linkages between areas whereecosystem services are generated and where they are consumed isimportant for management and policy development (Cowling et al.,2008; Reyers et al., 2009; von Haaren and Albert, 2011; Bagstadet al., 2014; Sturck et al., 2014). In this context, through a combi-nation of land cover/land use data, social and economic data the
remote sensing based approach is critical in defining these links,especially when it comes to providing the information we needto define spatially and historically where the ES are produced andwhere these services are consumed by the beneficiaries.

ologica

sptslhrp‘tnrzris

ceuttacttpfiataegdtD

5

bsrncNtsiaDioamBW

5

ttcs


Understanding the linkages between areas where ecosystemervices are generated and where they are received is important forolicy development, because, decisions by ‘upstream’ stakeholderso meet local requirements, may lead to positive or negative con-equences from the perspective of ‘downstream’ stakeholders, atarger scales. Once receptor areas have been identified, then stake-olders who benefit (‘beneficiaries’) and those who either do noteceive services (‘neutrals’) or who see a decrease in service sup-ly (‘losers’) can be identified (node 3B). This approach facilitates a

needs analysis’ that can in turn inform strategic decision-making,hough this is often complicated because stakeholders may be win-ers within the context of some ecosystem services and losers inelation to others, especially when evaluated over longer time hori-ons. The identification of beneficiaries allows spatially explicitepresentation of ecosystem service values. This in turn allowsdentification of strategies for equitable modifications to ecosystemervice provision (node 3D).

Appropriate measures to address the uncertainty of land coverlassification are a prerequisite to any mapping exercise (Olofssont al., 2013; Rocchini et al., 2013; Sexton et al., 2013). Through eval-ation of the selected literature we identified the assessment ofhematic accuracy of land cover classification to be another impor-ant issue, only 17% of the studies within the selected literatureppeared to have performed accuracy assessment on the land coverlassification results. A variety of measures were used to describehe accuracy of land cover classification. For example, to summarisehe results of an accuracy assessment, authors used the overallroportion of area and/or pixels correctly classified, the Kappa coef-cient of agreement, and the user’s and producer’s accuracies. Thepplicability of land cover data relies on its quality, and in this con-ext it is critical to assure the validity of the classification resultsnd their suitability for any particular purpose. In addition, therror budget included in the classification will most likely propa-ate through the analyses linking the final land cover map to otheratasets used for the ecosystem services valuation at any given spa-ial scale (Congalton and Green, 1993; Gahegan and Ehlers, 2000;e Clercq et al., 2009).

.3. Vegetation

Vegetation indices derived from satellite sensor images haveecome a major data source for monitoring ecosystem services,uch as climate regulation, primary production and natural hazardsegulation that are linked to vegetation condition. Nevertheless,one of the studies reviewed accounted for the influence of sensorharacteristics in the calculation of their NDVI values. For example,DVI values derived from MODIS products were applied in cap-

uring vegetation phenology, although the relatively coarse 500 mpatial resolution and smoothing forced by the 16 day BRDF modelnversion reduces their ability to capture fine scale temporal vari-tion in phenology (Jing et al., 2004; Hadjimitsis et al., 2010).ifficulties may, thus, emerge in detecting inter-annual variation

n phenology, and estimating phenological parameters such as thenset of greenness (OG), length of season and end of senescence,ffecting directly the robustness of relationships involved in esti-ating and forecasting crop cycles and crop yield (Fuller, 1998;

asnyat et al., 2004; Brown and de Beurs, 2008; Drolet et al., 2008;all et al., 2008b; Mkhabela et al., 2011).

.4. Spatial scale

The spatial scale at which ecosystem services are assessed
hrough remote sensing will greatly influence the resulting ecosys-em services monetary values, as changes in area within the landover types will impact the amount of change in total ecosystemervice values (Konarska et al., 2002; Nelson et al., 2009). It is
l Indicators 52 (2015) 430–443 439

important to understand how ecosystem assessments and mon-etary valuations will vary with changes in spatial scale, especiallywith most ecosystem services assessment and valuation studiesbeing supported by multi-sensor land cover data. For example, ifusing data from multiple sensors (assuming that they have differ-ent spatial resolutions), the resulting land cover classes will differin area (Boumans et al., 2002), and consequently affect ecosystemservice valuations. Depending on the spatial resolution, data fromdifferent sensors will most likely produce different estimates forthe same area (Konarska et al., 2002).

Another issue will emerge if the image scene contains subdom-inant and rare classes. These exist largely at the sub-pixel scaleand, thus, are difficult to resolve with thematic coarse resolutionsatellite sensor imagery (Serra et al., 2003; Aguirre-Gutierrez et al.,2012). Furthermore, due to the coarse spatial resolution (both spa-tial and spectral) of some satellite sensors, land cover classes thattend to appear in small patches are often omitted from the final clas-sification product, and consequently replaced by other dominantclasses surrounding them (e.g., using the maximum likelihood clas-sification method). Depending on the classification method used,the chosen classifier might be dominated by the prevalent classand fail to account for the rare examples. As a result, small and rareclasses may be underestimated and the large classes overestimated(Congalton, 1991; Hammond and Verbyla, 1996; Doan and Foody,2007).

5.5. Temporal extent

Observation of patterns in human activities over time, and thecapacities of different ecosystems to provide services under dif-ferent scenarios of land use change is a key element in most socialand ecological studies. In the literature selected 44% of studies haveused Earth observation data corresponding to a temporal extentgreater than ten years, and within this value 20% of studies con-centrate on extending observations to more than twenty years.

Ecosystem services studies would benefit greatly from exten-sive repeat surveys, as they are extremely useful in detecting andanalysing the temporal and spatial dynamics of ecosystem servicesand their likely beneficiaries. For example, if an area is convertedfrom forests to agricultural land this inevitably causes a loss of crit-ical habitat, and any ecosystem services study that considers only ashort period of time will fail to accurately assess the costs and ben-efits of that land use/cover conversion and the differences in thetiming of subsequent impacts. Consequently, research efforts inte-grating Earth observation data into ecosystem services studies willneed to ensure that the temporal extent of the analysis is consis-tent with the temporal extent of the projected impacts, such as instudies focused on long term monitoring of shifts on socioeconomicconditions and climate change.

5.6. Cultural ecosystem services

Earth observation data can be used to help define high priorityareas for tourism and conservation, and link concepts of heritageand identity to the landscape (Turner et al., 2003; Naidoo andRicketts, 2006; Cabello et al., 2012). However, assessment and val-uation of cultural ecosystem services through remote sensing havebeen focused mainly on extracting data to feed into GeographicInformation Systems (GIS) (Bender et al., 2005; Gret-Regamey et al.,2008; Sherrouse et al., 2011). The GIS object based approach is usu-ally not sufficient to reveal the complex behaviour emerging fromthe interactions between social and ecological systems, which in
turn is critical to define and constrain any given cultural ecosystemservice (Brown, 2005). In this context, cultural services and com-munity values tend to be overlooked and not entirely considered(Raymond et al., 2009; Daniel et al., 2012).

4 ologica

ste(fmcpuc

ssrsrc

emat

6

topaw

fpprtwkitn

qbpstptg

ptcft

A

HiC


The application of remote sensing combined with economic andocial data (both quantitative and qualitative) may help to stablishhe necessary abstractions we need when accounting for culturalcosystem services. For example, the combination of satellite datae.g. crop productivity, land cover, precipitation, and land sur-ace temperatures) individual and group choices, in a multi-agent

odelling environment, has the potential to provide estimates ofultural ecosystem services. Inevitably, the modelling exercise willroduce valuable data that can be used as input into dynamic sim-lation models, improving our understanding about the nature ofhanges occurred at varying spatial scales through time.

Assessment and valuation of cultural ecosystem servicestrongly depends on perceptions and expectations of users andtakeholders. Medium resolution satellite imagery (e.g. 30 mesolution Landsat images) can be used to inform users andtakeholders, creating more intuitive representations of the envi-onment, influencing expectations and choices, and inevitablyontributing to participatory and deliberative decision making.

In future studies, interdisciplinarity integrating more of RS intostimating cultural ecosystem services may provide the necessaryeans to move beyond measuring the value of physical stocks

bove the Earth’s surface, and start systematically valuing the cul-ural services provided to society.

. Conclusions

This systematic review summarises research progress towardshe use and integration of remote sensing data within the contextf ecosystem services valuation and assessment. Individual exam-les from the literature were reviewed to suggest what has beenchieved so far, what can be done and what can still be improvedhen using remote sensing data to estimate ecosystem services.

Through a systematic review of the literature on remote sensingor ecosystem services assessment and valuation studies, it wasossible to describe quantitatively both the overall numbers ofublications through time, and the relation between the range ofemotely sensed biophysical properties predicted and the ecosys-em services which they are used to estimate. The results showhich ecosystem services are the most measured from space, what

ind of remote sensing data have been applied more frequentlyn estimating these ecosystem services, the temporal resolution ofhese research efforts, and where in the world more research is stilleeded.

Several basic technical issues need to be addressed more fre-uently, as these issues will influence the overall uncertaintyudget, as follows: the accuracies of classification results and bio-hysical data used as proxies, the different spatial, temporal andpectral resolutions of the satellite sensor system (in many caseshis restricts the utilisation of the remotely sensed data) and, whenossible, the need to account for the dominant conditions at theime of image acquisition (e.g., atmospheric interference, viewingeometry, etc.).

Future research needs to account for the fact that effective map-ing of the full range of ecosystem services is currently beyondhe capabilities of remote sensing alone and, therefore, more con-erted research is needed to overcome this and other limitations,or example, through fusion of remotely sensed data with informa-ion from other sources.

cknowledgements

The authors thank the Coordination for the Improvement ofigher Education Personnel (CAPES) Foundation within the Min-

stry of Education in Brazil for the funding and support provided to.C. de Araujo Barbosa. We also would like to extend our gratitude

l Indicators 52 (2015) 430–443

to the reviewers for their helpful comments and suggestions on themanuscript.

References

Aguirre-Gutierrez, J., Seijmonsbergen, A.C., Duivenvoorden, J.F., 2012. Optimizingland cover classification accuracy for change detection, a combined pixel-basedand object-based approach in a mountainous area in Mexico. Appl. Geogr. 34,29–37.

Aighewi, I.T., Nosakhare, O.K., Ishaque, A.B., 2013. Land use-land cover changesand sewage loading in the lower eastern shore watersheds and coastal bays ofMaryland: implications for surface water quality. J. Coast. Res. 29, 1073–1082.

Aitkenhead, M.J., Aalders, I.H., 2011. Automating land cover mapping of Scotlandusing expert system and knowledge integration methods. Remote Sens. Environ.115, 1285–1295.

Alamgir, M., Pert, P.L., Turton, S.M., 2014. A review of ecosystem services research inAustralia reveals a gap in integrating climate change and impacts on ecosystemservices. Int. J. Biodivers. Sci. Ecosyst. Serv. Manage. 10, 112–127.

Alcaraz-Segura, D., Paruelo, J.M., Epstein, H.E., Cabello, J., 2013. Environmental andhuman controls of ecosystem functional diversity in temperate South America.Remote Sens. 5, 127–154.

Asner, G.P., Townsend, A.R., Bustamante, M.M.C., Nardoto, G.B., Olander, L.P., 2004.Pasture degradation in the central Amazon: linking changes in carbon and nutri-ent cycling with remote sensing. Global Change Biol. 10, 844–862.

Atzberger, C., Rembold, F., 2013. Mapping the spatial distribution of winter crops atsub-pixel level using AVHRR NDVI time series and neural nets. Remote Sens. 5,1335–1354.

Bagstad, K.J., Johnson, G.W., Voigt, B., Villa, F., 2013. Spatial dynamics of ecosystemservice flows: a comprehensive approach to quantifying actual services. Ecosyst.Serv. 4, 117–125.

Bagstad, K.J., Villa, F., Batker, D., Harrison-Cox, J., Voigt, B., Johnson, G.W., 2014. Fromtheoretical to actual ecosystem services: mapping beneficiaries and spatial flowsin ecosystem service assessments. Ecol. Soc., 19.

Baraldi, A., Boschetti, L., 2012. Operational automatic remote sensing imageunderstanding systems: beyond geographic object-based and object-orientedimage analysis (GEOBIA/GEOOIA). Part 2: Novel system architecture, informa-tion/knowledge representation, algorithm design and implementation. RemoteSens. 4, 2768–2817.

Barman, D., Sehgal, V.K., Sahoo, R.N., Nagarajan, S., 2010. Relationship of bidirec-tional reflectance of wheat with biophysical parameters and its radiative transfermodeling using PROSAIL. J. Ind. Soc. Remote Sens. 38, 35–44.

Basnyat, P., McConkey, B., Lafond, G.R., Moulin, A., Pelcat, Y., 2004. Optimal time forremote sensing to relate to crop grain yield on the Canadian prairies. Can. J. PlantSci. 84, 97–103.

Behrenfeld, M.J., Boss, E., 2006. Beam attenuation and chlorophyll concentration asalternative optical indices of phytoplankton biomass. J. Mar. Res. 64, 431–451.

Bender, O., Boehmer, H.J., Jens, D., Schumacher, K.P., 2005. Using GIS to analyse long-term cultural landscape change in Southern Germany. Landsc. Urban Plann. 70,111–125.

Beresford, L., 1990. Perspectives. Medicare rate hike challenges hospices. FaulknerGrays Med. Health 44, suppl. 4 p.

Bhaskaran, P.K., Nayak, S., Bonthu, S.R., Murty, P.L.N., Sen, D., 2013. Performance andvalidation of a coupled parallel ADCIRC-SWAN model for THANE cyclone in theBay of Bengal. Environ. Fluid Mech. 13, 601–623.

Bhattacharya, B.K., Padmanabhan, N., Mahammed, S., Ramakrishnan, R., Parihar,J.S., 2013. Assessing solar energy potential using diurnal remote-sensing obser-vations from Kalpana-1 VHRR and validation over the Indian landmass. Int. J.Remote Sens. 34, 7069–7090.

Boumans, R., Costanza, R., Farley, J., Wilson, M.A., Portela, R., Rotmans, J., Villa, F.,Grasso, M., 2002. Modeling the dynamics of the integrated earth system andthe value of global ecosystem services using the GUMBO model. Ecol. Econ. 41,529–560.

Boyd, J., Banzhaf, S., 2007. What are ecosystem services? The need for standardizedenvironmental accounting units. Ecol. Econ. 63, 616–626.

Brakenridge, G.R., Syvitski, J.P.M., Overeem, I., Higgins, S.A., Kettner, A.J., Stewart-Moore, J.A., Westerhoff, R., 2013. Global mapping of storm surges and theassessment of coastal vulnerability. Nat. Hazards 66, 1295–1312.

Brandt, J.S., Kuemmerle, T., Li, H.M., Ren, G.P., Zhu, J.G., Radeloff, V.C., 2012. UsingLandsat imagery to map forest change in southwest China in response to thenational logging ban and ecotourism development. Remote Sens. Environ. 121,358–369.

Brown, D.G., Robinson, D.T., An, L., Nassauer, J.I., Zellner, M., Rand, W., Riolo, R., Page,S.E., Low, B., Wang, Z., 2008. Exurbia from the bottom-up: confronting empiricalchallenges to characterizing a complex system. Geoforum 39, 805–818.

Brown, G., 2005. Mapping spatial attributes in survey research for natural resourcemanagement: methods and applications. Soc. Nat. Resour. 18, 17–39.

Brown, M.E., de Beurs, K.M., 2008. Evaluation of multi-sensor semi-arid crop seasonparameters based on NDVI and rainfall. Remote Sens. Environ. 112, 2261–2271.

Cabello, J., Fernandez, N., Alcaraz-Segura, D., Oyonarte, C., Pineiro, G., Altesor, A.,Delibes, M., Paruelo, J.M., 2012. The ecosystem functioning dimension in con-
servation: insights from remote sensing. Biodivers. Conserv. 21, 3287–3305.
Cai, Y.B., Zhang, H., Pan, W.B., Chen, Y.H., Wang, X.R., 2013. Land use pattern, socio-economic development, and assessment of their impacts on ecosystem servicevalue: study on natural wetlands distribution area (NWDA) in Fuzhou city,southeastern China. Environ. Monit. Assess. 185, 5111–5123.

ologica

C

C

C

C

C

C

C

C

C

C

C

C

C

C

C

C

C

C

D

D

D

D

D

d

d

d

d


airns, B., Carlson, B.E., Ying, R., Laveigne, J., 2002. Accuracy vs speed: evaluation oftradeoffs in atmospheric correction methods. Algorithms Technol. Multispectr.Hyperspectr. Ultraspectr. Imag. Viii 4725, 427–437.

amacho-Valdez, V., Ruiz-Luna, A., Ghermandi, A., Nunes, P.A.L.D., 2013. Valuationof ecosystem services provided by coastal wetlands in northwest Mexico. OceanCoast. Manage. 78, 1–11.

astella, J.C., Lestrelin, G., Hett, C., Bourgoin, J., Fitriana, Y.R., Heinimann, A., Pfund,J.L., 2013. Effects of landscape segregation on livelihood vulnerability: movingfrom extensive shifting cultivation to rotational agriculture and natural forestsin northern Laos. Hum. Ecol. 41, 63–76.

hambers, J.Q., Asner, G.P., Morton, D.C., Anderson, L.O., Saatchi, S.S., Espirito-Santo,F.D., Palace, M., Souza Jr., C., 2007. Regional ecosystem structure and function:ecological insights from remote sensing of tropical forests. Trends Ecol. Evol. 22,414–423.

hang, N.-B., Xuan, Z., 2011. Exploring the nutrient inputs and cycles in Tampa Bayand coastal watersheds using MODIS images and data mining. In: Gao, W., Jack-son, T.J., Wang, J., Chang, N.B. (Eds.), Remote Sensing and Modeling of Ecosystemsfor Sustainability, vol. Viii.

hen, S., Fang, L., Zhang, L., Huang, W., 2009. Remote sensing of turbidity in seawaterintrusion reaches of Pearl river estuary – a case study in Modaomen water way,China. Estuar. Coast. Shelf Sci. 82, 119–127.

hen, Z.H., Ma, Q.Y., Wang, J., Yang, Z., 2007. Assessing value of grassland ecosystemservices in Gansu Province, northwest of China. In: Igarss: 2007 IEEE Interna-tional Geoscience and Remote Sensing Symposium, vols. 1–12, pp. 1782–1785.

hopra, R., Verma, V.K., Sharma, P.K., 2001. Mapping, monitoring and conservation ofHarike wetland ecosystem, Punjab, India, through remote sensing. Int. J. RemoteSens. 22, 89–98.

hu, D.A., Kaufman, Y.J., Zibordi, G., Chern, J.D., Mao, J., Li, C.C., Holben, B.N., 2003.Global monitoring of air pollution over land from the Earth Observing System-Terra Moderate Resolution Imaging Spectroradiometer (MODIS). J. Geophys.Res.-Atmos., 108.

loern, J.E., Jassby, A.D., 2012. Drivers of change in estuarine–coastal ecosystems:discoveries from four decades of study in San Francisco bay. Rev. Geophys., 50.

olesanti, C., Wasowski, J., 2006. Investigating landslides with space-borne syntheticaperture radar (SAR) interferometry. Eng. Geol. 88, 173–199.

ongalton, R.G., 1991. A review of assessing the accuracy of classifications ofremotely sensed data. Remote Sens. Environ. 37, 35–46.

ongalton, R.G., Green, K., 1993. A practical look at the sources of confusion in errormatrix generation. Photogrammet. Eng. Remote Sens. 59, 641–644.

oppin, P., Jonckheere, I., Nackaerts, K., Muys, B., Lambin, E., 2004. Digital changedetection methods in ecosystem monitoring: a review. Int. J. Remote Sens. 25,1565–1596.

ostanza, R., dArge, R., deGroot, R., Farber, S., Grasso, M., Hannon, B., Limburg, K.,Naeem, S., ONeill, R.V., Paruelo, J., Raskin, R.G., Sutton, P., vandenBelt, M., 1997.The value of the world’s ecosystem services and natural capital. Nature 387,253–260.

ostanza, R., Kubiszewski, I., 2012. The authorship structure of ecosystem servicesas a transdisciplinary field of scholarship. Ecosyst. Serv. 1, 16–25.

ostanza, R., Perez-Maqueo, O., Martinez, M.L., Sutton, P., Anderson, S.J., Mulder,K., 2008. The value of coastal wetlands for hurricane protection. Ambio 37,241–248.

owling, R.M., Egoh, B., Knight, A.T., O’Farrell, P.J., Reyers, B., Rouget’ll, M., Roux, D.J.,Welz, A., Wilhelm-Rechman, A., 2008. An operational model for mainstream-ing ecosystem services for implementation. Proc. Natl. Acad. Sci. U. S. A. 105,9483–9488.

agg, J., Lafleur, P., 2010. An application of plot-scale NDVI in predicting carbondioxide exchange and leaf area index in heterogeneous subarctic tundra. Can. J.Remote Sens. 36, S111–S123.

aily, G.C., 1997. Introduction: What are Ecosystem Services? Island Press, Wash-ington, DC.

alezios, N.R., Domenikiotis, C., Loukas, A., Tzortzios, S.T., Kalaitzidis, C., 2001. Cottonyield estimation based on NOAA/AVHRR produced NDVI. Phys. Chem. Earth PartB-Hydrol. Oceans Atmos. 26, 247–251.

aniel, T.C., Muhar, A., Arnberger, A., Aznar, O., Boyd, J.W., Chan, K.M., Costanza,R., Elmqvist, T., Flint, C.G., Gobster, P.H., Gret-Regamey, A., Lave, R., Muhar, S.,Penker, M., Ribe, R.G., Schauppenlehner, T., Sikor, T., Soloviy, I., Spierenburg,M., Taczanowska, K., Tam, J., von der Dunk, A., 2012. Contributions of culturalservices to the ecosystem services agenda. Proc. Natl. Acad. Sci. U. S. A. 109,8812–8819.

e Clercq, E.M., Clement, L., De Wulf, R.R., 2009. Monte Carlo simulation of falsechange in the overlay of misregistered forest vector maps. Landsc. Urban Plann.91, 36–45.

e Groot, R., 2006. Function-analysis and valuation as a tool to assess land use con-flicts in planning for sustainable, multi-functional landscapes. Landsc. UrbanPlann. 75, 175–186.

e Groot, R., Brander, L., van der Ploeg, S., Costanza, R., Bernard, F., Braat, L., Christie,M., Crossman, N., Ghermandi, A., Hein, L., Hussain, S., Kumar, P., McVittie, A.,Portela, R., Rodriguez, L.C., ten Brink, P., van Beukering, P., 2012. Global estimatesof the value of ecosystems and their services in monetary units. Ecosyst. Serv. 1,50–61.

e Groot, R.S., Alkemade, R., Braat, L., Hein, L., Willemen, L., 2010. Challenges in
integrating the concept of ecosystem services and values in landscape planning,management and decision making. Ecol. Complex. 7, 260–272.
e Groot, R.S., Wilson, M.A., Boumans, R.M.J., 2002. A typology for the classification,description and valuation of ecosystem functions, goods and services. Ecol. Econ.41, 393–408.

l Indicators 52 (2015) 430–443 441

Deegan, L.A., Johnson, D.S., Warren, R.S., Peterson, B.J., Fleeger, J.W., Fagherazzi, S.,Wollheim, W.M., 2012. Coastal eutrophication as a driver of salt marsh loss.Nature 490, 388–392.

Doan, H.T.X., Foody, G.M., 2007. Increasing soft classification accuracy through theuse of an ensemble of classifiers. Int. J. Remote Sens. 28, 4609–4623.

Doron, M., Belanger, S., Doxaran, D., Babin, M., 2011. Spectral variations in the near-infrared ocean reflectance. Remote sens. Environ. 115, 1617–1631.

Drolet, G.G., Middleton, E.M., Huemmrich, K.F., Hall, F.G., Amiro, B.D., Barr, A.G., Black,T.A., McCaughey, J.H., Margolis, H.A., 2008. Regional mapping of gross light-useefficiency using MODIS spectral indices. Remote Sens. Environ. 112, 3064–3078.

Egoh, B., Rouget, M., Reyers, B., Knight, A.T., Cowling, R.M., van Jaarsveld, A.S.,Welz, A., 2007. Integrating ecosystem services into conservation assessments:a review. Ecol. Econ. 63, 714–721.

Eigenbrod, F., Armsworth, P.R., Anderson, B.J., Heinemeyer, A., Gillings, S., Roy, D.B.,Thomas, C.D., Gaston, K.J., 2010. The impact of proxy-based methods on mappingthe distribution of ecosystem services. J. Appl. Ecol. 47, 377–385.

El-ossta, E., Qahwaji, R., Ipson, S.S., 2013. Detection of dust storms using MODISreflective and emissive bands. IEEE J. Sel. Top. Appl. Earth Observ. Remote Sens.6, 2480–2485.

Elmahboub, W.M., Scarpace, F., Smith, B., 2006. Algorithm and mathematical mod-eling for atmospheric correction and classification accuracy for hyperspectraldata. In: Wmsci 2006: 10th World Multi-Conference on Systemics, Cyberneticsand Informatics, vol. Iii, Proceedings, pp. 56–59.

Emeis, S., Schaefer, K., 2006. Remote sensing methods to investigate boundary-layer structures relevant to air pollution in cities. Bound. Layer Meteorol. 121,377–385.

Estoque, R.C., Murayama, Y., 2013. Landscape pattern and ecosystem service valuechanges: implications for environmental sustainability planning for the rapidlyurbanizing summer capital of the Philippines. Landsc. Urban Plann. 116, 60–72.

Estreguil, C., Lambin, E.F., 1996. Mapping forest-cover disturbances in Papua NewGuinea with AVHRR data. J. Biogeogr. 23, 757–773.

Evans, T.P., Kelley, H., 2008. Assessing the transition from deforestation to forestregrowth with an agent-based model of land cover change for south-centralIndiana (USA). Geoforum 39, 819–832.

Faber, J.H., 2006. European experience on application of site-specific ecologi-cal risk assessment in terrestrial ecosystems. Hum. Ecol. Risk Assess. 12,39–50.

Fegraus, E.H., Zaslavsky, I., Whitenack, T., Dempewolf, J., Ahumada, J.A., Lin, K., Andel-man, S.J., 2012. Interdisciplinary decision support dashboard: a new frameworkfor a Tanzanian agricultural and ecosystem service monitoring system pilot. IEEEJ. Sel. Top. Appl. Earth Observ. Remote Sens. 5, 1700–1708.

Fish, R.D., 2011. Environmental decision making and an ecosystems approach:some challenges from the perspective of social science. Progr. Phys. Geogr. 35,671–680.

Fisher, B., Turner, R.K., Morling, P., 2009. Defining and classifying ecosystem servicesfor decision making. Ecol. Econ. 68, 643–653.

Frank, S., Fuerst, C., Koschke, L., Makeschin, F., 2012. A contribution towards a trans-fer of the ecosystem service concept to landscape planning using landscapemetrics. Ecol. Indic. 21, 30–38.

Fuller, D.O., 1998. Trends in NDVI time series and their relation to range-land and crop production in Senegal, 1987–1993. Int. J. Remote Sens. 19,2013–2018.

Gahegan, M., Ehlers, M., 2000. A framework for the modelling of uncertainty betweenremote sensing and geographic information systems. ISPRS J. Photogram. Rem.Sens. 55, 176–188.

Ghobadi, Y., Pradhan, B., Kabiri, K., Pirasteh, S., Shafri, H.Z.M., Sayyad, G.A., 2012.Use of multi-temporal remote sensing data and GIS for wetland change mon-itoring and degradation. In: 2012 IEEE Colloquium on Humanities, Science &Engineering Research (Chuser 2012).

Gong, P., Niu, Z., Cheng, X., Zhao, K., Zhou, D., Guo, J., Liang, L., Wang, X., Li, D., Huang,H., Wang, Y., Wang, K., Li, W., Wang, X., Ying, Q., Yang, Z., Ye, Y., Li, Z., Zhuang, D.,Chi, Y., Zhou, H., Yan, J., 2010. China’s wetland change (1990–2000) determinedby remote sensing. Sci. China Earth Sci. 53, 1036–1042.

Gret-Regamey, A., Bebi, P., Bishop, I.D., Schmid, W.A., 2008. Linking GIS-based mod-els to value ecosystem services in an Alpine region. J. Environ. Manage. 89,197–208.

Hadjimitsis, D.G., Papadavid, G., Agapiou, A., Themistocleous, K., Hadjimitsis, M.G.,Retalis, A., Michaelides, S., Chrysoulakis, N., Toulios, L., Clayton, C.R.I., 2010.Atmospheric correction for satellite remotely sensed data intended for agricul-tural applications: impact on vegetation indices. Nat. Hazards Earth Syst. Sci. 10,89–95.

Hammond, T.O., Verbyla, D.L., 1996. Optimistic bias in classification accuracy assess-ment. Int. J. Remote Sens. 17, 1261–1266.

Han, G.Q., Ma, Z.M., Chen, D., deYoung, B., Chen, N., 2012. Observing storm surgesfrom space: hurricane Igor off newfoundland. Sci. Rep., 2.

Hao, F., Zhang, X., Ouyang, W., Skidmore, A.K., Toxopeus, A.G., 2012. Vegetation NDVIlinked to temperature and precipitation in the upper catchments of yellow river.Environ. Model. Assess. 17, 389–398.

Hasager, C.B., Frank, H.P., Furevik, B.R., 2002. On offshore wind energy mapping usingsatellite SAR. Can. J. Remote Sens. 28, 80–89.

Hochschild, V., Marker, M., Rodolfi, G., Staudenrausch, H., 2003. Delineation of ero-
sion classes in semi-arid southern African grasslands using vegetation indicesfrom optical remote sensing data. Hydrol. Process. 17, 917–928.
Hyun-Joo, O., No-Wook, P., Sung-Soon, L., Lee, S., 2012. Extraction of landslide-related factors from ASTER imagery and its application to landslide susceptibilitymapping. Int. J. Remote Sens. 33, 3211–3231.

4 ologica

I

I

J

J

J

J

J

K

K

K

K

K

K

K

K

K

K

K

K

K

L

L

L

L

L

L

L

L

L

M

M


mhoff, M.L., Bounoua, L., Ricketts, T., Loucks, C., Harriss, R., Lawrence, W.T., 2004.Global patterns in human consumption of net primary production. Nature 429,870–873.

vits, E., Cherlet, M., Toth, G., Sommer, S., Mehl, W., Vogt, J., Micale, F., 2012. Com-bining satellite derived phenology with climate data for climate change impactassessment. Global Planet. Change 88-89, 85–97.

epsen, M.R., Leisz, S., Rasmussen, K., Jakobsen, J., Moller-Jensen, L., Christiansen, L.,2006. Agent-based modelling of shifting cultivation field patterns, Vietnam. Int.J. Geogr. Infor. Sci. 20, 1067–1085.

i, W.J., Shi, Z., Zhou, Q., Zhou, L.Q., 2012. VIS-NIR reflectance spectroscopy of theorganic matter in several types of soils. J. Infrared Millim. Waves 31, 277–282.

ing, J.J., Liu, L.Y., Wang, J.H., Wang, J.D., Zhao, C.J., 2004. Uncertainty analysis forNDVI using the physical models. In: Igarss 2004: IEEE International Geoscienceand Remote Sensing Symposium Proceedings, vols. 1–7, pp. 4321–4324.

ohannessen, O.M., Espedal, H., Bjorgo, E., 1999. Wind energy mapping from syn-thetic aperture radar. In: IEEE 1999 International Geoscience and RemoteSensing Symposium. IGARSS’99 (Cat. No. 99CH36293), vol. 1613, pp. 1619–1621.

ohnson, K.A., Polasky, S., Nelson, E., Pennington, D., 2012. Uncertainty in ecosys-tem services valuation and implications for assessing land use tradeoffs: anagricultural case study in the Minnesota River Basin. Ecol. Econ. 79, 71–79.

arl, D.M., 2002. Nutrient dynamics in the deep blue sea. Trends Microbiol. 10,410–418.

astens, J.H., Kastens, T.L., Kastens, D.L.A., Price, K.P., Martinko, E.A., Lee, R.Y., 2005.Image masking for crop yield forecasting using AVHRR NDVI time series imagery.Remote Sens. Environ. 99, 341–356.

awata, Y., Ohtani, A., Kusaka, T., Ueno, S., 1990. Classification accuracy for the Mos-1 Messr data before and after the atmospheric correction. IEEE Trans. Geosci.Remote Sens. 28, 755–760.

e, C.Q., Zhang, D., Wang, F.Q., Chen, S.X., Schmullius, C., Boerner, W.M., Wang,H., 2011. Analyzing coastal wetland change in the Yancheng National NatureReserve, China. Reg. Environ. Change 11, 161–173.

emper, T., Sommer, S., 2002. Estimate of heavy metal contamination in soils aftera mining accident using reflectance spectroscopy. Environ. Sci. Technol. 36,2742–2747.

ienast, F., Bolliger, J., Potschin, M., de Groot, R.S., Verburg, P.H., Heller, I., Wascher,D., Haines-Young, R., 2009. Assessing landscape functions with broad-scale envi-ronmental data: insights gained from a prototype development for Europe.Environ. Manage. 44, 1099–1120.

lepeis, P., Scull, P., LaLonde, T., Svajlenka, N., Gill, N., 2013. Changing forest recoverydynamics in the northeastern United States. Area 45, 239–248.

naeps, E., Sterckx, S., Raymaekers, D., 2010. A seasonally robust empirical algorithmto retrieve suspended sediment concentrations in the scheldt river. Remote Sens.2, 2040–2059.

onarska, K.M., Sutton, P.C., Castellon, M., 2002. Evaluating scale dependence ofecosystem service valuation: a comparison of NOAA-AVHRR and Landsat TMdatasets. Ecol. Econ. 41, 491–507.

ozai, K., Ohsawa, T., 2007. Evaluation of offshore wind energy potential using SARand MM5 – art. no. 66800Y. In: Frouin, R.J., Lee, Z. (Eds.), Coastal Ocean RemoteSensing. , p. Y6800.

reuter, U.P., Harris, H.G., Matlock, M.D., Lacey, R.E., 2001. Change in ecosystemservice values in the San Antonio area, Texas. Ecol. Econ. 39, 333–346.

rishnaswamy, J., Bawa, K.S., Ganeshaiah, K.N., Kiran, M.C., 2009. Quantifying andmapping biodiversity and ecosystem services: Utility of a multi-season NDVIbased Mahalanobis distance surrogate. Remote Sens. Environ. 113, 857–867.

ross, A., Seaquist, J.W., Roulet, N.T., Fernandes, R., Sonnentag, O., 2013. Estimatingcarbon dioxide exchange rates at contrasting northern peatlands using MODISsatellite data. Remote Sens. Environ. 137, 234–243.

a Puma, I.P., Philippi, T.E., Oberbauer, S.F., 2007. Relating NDVI to ecosystem CO2

exchange patterns in response to season length and soil warming manipulationsin arctic Alaska. Remote Sens. Environ. 109, 225–236.

ambin, E.F., 1996. Change detection at multiple temporal scales: Seasonal andannual variations in landscape variables. Photogrammet. Eng. Remote Sens. 62,931–938.

ewis, D., Phinn, S., Arroyo, L., 2013. Cost-effectiveness of seven approaches to mapvegetation communities – a case study from northern Australia’s tropical savan-nas. Remote Sens. 5, 377–414.

in, J., Tang, D., Alpers, W., Wang, S., 2014. Response of dissolved oxygen and relatedmarine ecological parameters to a tropical cyclone in the South China Sea. Adv.Space Res. 53, 1081–1091.

inard, C., Tatem, A.J., Gilbert, M., 2013. Modelling spatial patterns of urban growthin Africa. Appl. Geogr. 44, 23–32.

ing, F., Li, Z., Chen, E., Wang, Q., 2009. Comparison of ALOS PALSAR RVI and LandsatTM NDVI for forest area mapping. In: Proceedings of the 2009 2nd Asian-PacificConference on Synthetic Aperture Radar (APSAR 2009), pp. 132–135.

iu, J., Li, Y., Liu, S.B., 2013a. Prediction models of soil organic matter based on spectralcurve in the upstream of Heihe basin. Spectrosc. Spectr. Anal. 33, 3354–3358.

iu, J., Sun, D., Zhang, Y., Li, Y., 2013b. Pre-classification improves relationshipsbetween water clarity, light attenuation, and suspended particulates in turbidinland waters. Hydrobiologia 711, 71–86.

iu, S., Costanza, R., Troy, A., D’Aagostino, J., Mates, W., 2010. Valuing New Jersey’secosystem services and natural capital: a spatially explicit benefit transfer
approach. Environ. Manage. 45, 1271–1285.
A, 2005. Millennium Ecosystem Assessment. Ecosystems and Human Well-being:Synthesis. Washington, DC, Island Press.

aes, J., Egoh, B., Willemen, L., Liquete, C., Vihervaara, P., Schägner, J.P., Grizzetti, B.,Drakou, E.G., Notte, A.L., Zulian, G., Bouraoui, F., Luisa Paracchini, M., Braat, L.,

l Indicators 52 (2015) 430–443

Bidoglio, G., 2012. Mapping ecosystem services for policy support and decisionmaking in the European Union. Ecosyst. Serv. 1, 31–39.

Marghany, M., Hashim, M., 2010. MODIS satellite data for modeling chlorophyll-aconcentrations in Malaysian coastal waters. Int. J. Phys. Sci. 5, 1489–1495.

Melendez-Pastor, I., Navarro-Pedreno, J., Koch, M., Gomez, I., 2010. Applying imagingspectroscopy techniques to map saline soils with ASTER images. Geoderma 158,55–65.

Mkhabela, M.S., Bullock, P., Raj, S., Wang, S., Yang, Y., 2011. Crop yield forecastingon the Canadian Prairies using MODIS NDVI data. Agric. Forest Meteorol. 151,385–393.

Moran, M.S., Jackson, R.D., Slater, P.N., Teillet, P.M., 1992. Evaluation of simplifiedprocedures for retrieval of land surface reflectance factors from satellite sensoroutput. Remote Sens. Environ. 41, 169–184.

Morelli, M., Masini, A., Potenza, M.A.C., 2012. A new method for the performanceanalysis of a concentrating solar power energy plant using remotely sensedoptical images. In: Proceedings of the Advances in Atmospheric Science andApplications (ATMOS 2012), pp. 53–55.

Morgan, D.G., Ben Abdallah, S., Lasserre, P., 2008. A real options approach to forest-management decision making to protect caribou under the threat of extinction.Ecol. Soc., 13.

Moriondo, M., Maselli, F., Bindi, M., 2007. A simple model of regional wheat yieldbased on NDVI data. Eur. J. Agron. 26, 266–274.

Mozumder, C., Reddy, K.V., Pratap, D., 2013. Air pollution modeling from remotelysensed data using regression techniques. J. Ind. Soc. Remote Sens. 41,269–277.

Muraoka, H., Koizumi, H., 2009. Satellite ecology (SATECO)-linking ecology, remotesensing and micrometeorology, from plot to regional scale, for the study ofecosystem structure and function. J. Plant Res. 122, 3–20.

Murray, N.J., Phinn, S.R., Clemens, R.S., Roelfsema, C.M., Fuller, R.A., 2012. Continentalscale mapping of tidal flats across east Asia using the landsat archive. RemoteSens. 4, 3417–3426.

Muthuwatta, L.P., Ahmad, M.U.D., Bos, M.G., Rientjes, T.H.M., 2010. Assessmentof water availability and consumption in the Karkheh river basin, Iran-usingremote sensing and geo-statistics. Water Resour. Manage. 24, 459–484.

Muukkonen, P., Heiskanen, J., 2005. Estimating biomass for boreal forests usingASTER satellite data combined with standwise forest inventory data. RemoteSens. Environ. 99, 434–447.

Naidoo, R., Ricketts, T.H., 2006. Mapping the economic costs and benefits of conser-vation. PLoS Biol. 4, e360.

Nedkov, S., Burkhard, B., 2012. Flood regulating ecosystem services-Mappingsupply and demand, in the Etropole municipality, Bulgaria. Ecol. Indic. 21,67–79.

Nelson, E., Mendoza, G., Regetz, J., Polasky, S., Tallis, H., Cameron, D.R., Chan, K.M.A.,Daily, G.C., Goldstein, J., Kareiva, P.M., Lonsdorf, E., Naidoo, R., Ricketts, T.H.,Shaw, M.R., 2009. Modeling multiple ecosystem services, biodiversity conser-vation, commodity production, and tradeoffs at landscape scales. Front. Ecol.Environ. 7, 4–11.

Nemani, R., Hashimoto, H., Votava, P., Melton, F., Wang, W.L., Michaelis, A., Mutch,L., Milesi, C., Hiatt, S., White, M., 2009. Monitoring and forecasting ecosys-tem dynamics using the Terrestrial Observation and Prediction System (TOPS).Remote Sens. Environ. 113, 1497–1509.

Niu, X., Wang, B., Liu, S.R., Liu, C.J., Wei, W.J., Kauppi, P.E., 2012. Economical assess-ment of forest ecosystem services in China: characteristics and implications.Ecol. Complex. 11, 1–11.

Noferini, L., Pieraccini, M., Mecatti, D., Macaluso, G., Atzeni, C., Mantovani, M., Mar-cato, G., Pasuto, A., Silvano, S., Tagliavini, F., 2007. Using GB-SAR technique tomonitor slow moving landslide. Eng. Geol. 95, 88–98.

Notaro, M., Alkolibi, F., Fadda, E., Bakhrjy, F., 2013. Trajectory analysis of Saudi Ara-bian dust storms. J. Geophys. Res.-Atmos. 118, 6028–6043.

Numata, I., Soares, J.V., Roberts, D.A., Leonidas, F.C., Chadwick, O.A., Batista, G.T.,2003. Relationships among soil fertility dynamics and remotely sensed meas-ures across pasture chronosequences in Rondonia, Brazil. Remote Sens. Environ.87, 446–455.

Olivas, P.C., Oberbauer, S.F., Tweedie, C.E., Oechel, W.C., Kuchy, A., 2010. Responsesof CO2 flux components of Alaskan Coastal Plain tundra to shifts in water table.J. Geophys. Res.-Biogeosci., 115.

Olofsson, P., Foody, G.M., Stehman, S.V., Woodcock, C.E., 2013. Making better useof accuracy data in land change studies: Estimating accuracy and area andquantifying uncertainty using stratified estimation. Remote Sens. Environ. 129,122–131.

Paerl, H.W., Paul, V.J., 2012. Climate change: links to global expansion of harmfulcyanobacteria. Water Res. 46, 1349–1363.

Palacios-Orueta, A., Huesca, M., Whiting, M.L., Litago, J., Khanna, S., Garcia, M., Ustin,S.L., 2012. Derivation of phenological metrics by function fitting to time-series ofspectral shape indexes AS1 and AS2: Mapping cotton phenological stages usingMODIS time series. Remote Sens. Environ. 126, 148–159.

Perez, R., Seals, R., Stewart, R., Zelenka, A., Estradacajigal, V., 1994. Using satellite-derived insolation data for the site/time specific simulation of solar-energysystems. Solar Energy 53, 491–495.

Petit, C.C., Lambin, E.F., 2001. Integration of multi-source remote sensing data forland cover change detection. Int. J. Geogr. Inf. Sci. 15, 785–803.

Porras, I., 2012. Environment: Costa Rica pioneers ecosystem services. Nature 487,302.

Potter, C., Gross, P., Klooster, S., Fladeland, M., Genovese, V., 2008. Storage of carbonin US forests predicted from satellite data, ecosystem modeling, and inventorysummaries. Clim. Change 90, 269–282.

ologica

P

R

R

R

R

R

R

R

R

R

S

S

S

S

S

S

S

S

S

S

S

S

S

S

S

S

S

T

T


rabakaran, C., Singh, C.P., Panigrahy, S., Parihar, J.S., 2013. Retrieval of forest phe-nological parameters from remote sensing-based NDVI time-series data. Curr.Sci. 105, 795–802.

ahman, M.R., Shi, Z.H., Chongfa, C., 2009. Soil erosion hazard evaluation – an inte-grated use of remote sensing, GIS and statistical approaches with biophysicalparameters towards management strategies. Ecol. Model. 220, 1724–1734.

aucoules, D., de Michele, M., Malet, J.P., Ulrich, P., 2013. Time-variable 3D grounddisplacements from high-resolution synthetic aperture radar (SAR). Applicationto La Valette landslide (South French Alps). Remote Sens. Environ. 139, 198–204.

aymond, C.M., Bryan, B.A., MacDonald, D.H., Cast, A., Strathearn, S., Grandgirard, A.,Kalivas, T., 2009. Mapping community values for natural capital and ecosystemservices. Ecol. Econ. 68, 1301–1315.

eid, W.V., Mooney, H.A., Capistrano, D., Carpenter, S.R., Chopra, K., Cropper, A.,Dasgupta, P., Hassan, R., Leemans, R., May, R.M., Pingali, P., Samper, C., Scholes,R., Watson, R.T., Zakri, A.H., Shidong, Z., 2006. Nature: the many benefits ofecosystem services. Nature 443, 749, author reply 750.

eyers, B., O’Farrell, P.J., Cowling, R.M., Egoh, B.N., Le Maitre, D.C., Vlok, J.H.J., 2009.Ecosystem services, land-cover change, and stakeholders: finding a sustainablefoothold for a semiarid biodiversity hotspot. Ecol. Soc., 14.

ighini, G., Pancioli, V., Casagli, N., 2012. Updating landslide inventory maps usingPersistent Scatterer Interferometry (PSI). Int. J. Remote Sens. 33, 2068–2096.

occhini, D., Foody, G.M., Nagendra, H., Ricotta, C., Anand, M., He, K.S., Amici, V.,Kleinschmit, B., Forster, M., Schmidtlein, S., Feilhauer, H., Ghisla, A., Metz, M.,Neteler, M., 2013. Uncertainty in ecosystem mapping by remote sensing. Com-put. Geosci. 50, 128–135.

oots, O.O., Roose, A., Eerme, K., 2011. Remote sensing of climate change, long-termmonitoring of air pollution and stone material corrosion in Estonia. Int. J. RemoteSens. 32, 9691–9705.

oujean, J.L., 1998. Two-storey Equations of the Transmission of Solar Energy(TSETSE) in Open Boreal Conifer Tree Stands.

aatchi, S.S., Houghton, R.A., Alvala, R.C.D.S., Soares, J.V., Yu, Y., 2007. Distribution ofaboveground live biomass in the Amazon basin. Global Change Biol. 13, 816–837.

adhuram, Y., Ramana Murthy, T.V., Somayajulu, Y.K., 2006. Estimation of tropi-cal cyclone heat potential in the Bay of Bengal and its role in the genesis andintensification of storms. Indian J. Mar. Sci. 35, 132–138.

chmitt, K., Albers, T., Pham, T.T., Dinh, S.C., 2013. Site-specific and integrated adap-tation to climate change in the coastal mangrove zone of Soc Trang Province,Viet Nam. J. Coast. Conserv. 17, 545–558.

erra, P., Pons, X., Sauri, D., 2003. Post-classification change detection with datafrom different sensors: some accuracy considerations. Int. J. Remote Sens. 24,3311–3340.

esnie, S.E., Gessler, P.E., Finegan, B., Thessler, S., 2008. Integrating Landsat TM andSRTM-DEM derived variables with decision trees for habitat classification andchange detection in complex neotropical environments. Remote Sens. Environ.112, 2145–2159.

exton, J.O., Urban, D.L., Donohue, M.J., Song, C.H., 2013. Long-term land coverdynamics by multi-temporal classification across the Landsat-5 record. RemoteSens. Environ. 128, 246–258.

herrouse, B.C., Clement, J.M., Semmens, D.J., 2011. A GIS application for assessing,mapping, and quantifying the social values of ecosystem services. Appl. Geogr.31, 748–760.

huwen, Y., Yikun, L., Guangsheng, F., Tao, L., Zhihua, Z., Lifeng, Z., 2012. Bareland extraction method of rain-induced shallow landslides based on SPOT5images. 2012. In: International Symposium on Geomatics for Integrated WaterResources Management (GIWRM 2012), 4 pp.

kourtos, M., Kontogianni, A., Harrison, P.A., 2010. Reviewing the dynamics of eco-nomic values and preferences for ecosystem goods and services. Biodivers.Conserv. 19, 2855–2872.

mall, C., Steckler, M., Seeber, L., Akhter, S.H., Goodbred, S., Mia, B., Imam, B., 2009.Spectroscopy of sediments in the Ganges-Brahmaputra delta: spectral effects ofmoisture, grain size and lithology. Remote Sens. Environ. 113, 342–361.

ong, C., Woodcock, C.E., Seto, K.C., Lenney, M.P., Macomber, S.A., 2001. Classifi-cation and change detection using landsat TM data. Remote Sens. Environ. 75,230–244.

oojeong, M., Nowak, D.J., Duggin, M.J., 2006. A temporal analysis of urban forestcarbon storage using remote sensing. Remote Sens. Environ. 101, 277–282.

pillman, C.M., Imberger, J., Hamilton, D.P., Hipsey, M.R., Romero, J.R., 2007.Modelling the effects of Po River discharge, internal nutrient cycling and hydro-dynamics on biogeochemistry of the Northern Adriatic Sea. J. Mar. Syst. 68,167–200.

tolle, F., Dennis, R.A., Kurniwan, I., Lambin, E.F., 2004. Evaluation of remote sensing-based active fire datasets in Indonesia. Int. J. Remote Sens. 25, 471–479.

turck, J., Poortinga, A., Verburg, P.H., 2014. Mapping ecosystem services: the supplyand demand of flood regulation services in Europe. Ecol. Indic. 38, 198–211.

u, S.L., Xiao, R., Jiang, Z.L., Zhang, Y., 2012. Characterizing landscape pattern andecosystem service value changes for urbanization impacts at an eco-regionalscale. Appl. Geogr. 34, 295–305.

utton, P.C., Costanza, R., 2002. Global estimates of market and non-market valuesderived from nighttime satellite imagery, land cover, and ecosystem servicevaluation. Ecol. Econ. 41, 509–527.

antianuparp, P., Shi, X., Zhang, L., Balz, T., Liao, M., 2013. Characterization of land-
slide deformations in three gorges area using multiple InSAR data stacks. RemoteSens. 5, 2704–2719.
aylor, G.R., Hemphill, P., Currie, D., Broadfoot, T., Dehaan, R.L., 2001. Mapping dry-land salinity with hyperspectral imagery. In: Igarss 2001: Scanning the Presentand Resolving the Future, vols. 1–7, Proceedings, pp. 302–304.

l Indicators 52 (2015) 430–443 443

Turner, W., Spector, S., Gardiner, N., Fladeland, M., Sterling, E., Steininger, M., 2003.Remote sensing for biodiversity science and conservation. Trends Ecol. Evol. 18,306–314.

Ustin, S.L., Darling, D., Kefauver, S., Greenberg, J., Cheng, Y.B., Whiting, M.L., 2004.Remotely sensed estimates of crop water demand. Remote Sens. Model. Ecosyst.Sustain. 5544, 230–240.

Van Tuyl, S., Law, B.E., Turner, D.P., Gitelman, A.I., 2005. Variability in net primaryproduction and carbon storage in biomass across Oregon forests – an assessmentintegrating data from forest inventories, intensive sites, and remote sensing.Forest Ecol. Manage. 209, 273–291.

Vanonckelen, S., Lhermitte, S., Van Rompaey, A., 2013. The effect of atmosphericand topographic correction methods on land cover classification accuracy. Int.J. Appl. Earth Observ. Geoinfor. 24, 9–21.

Verburg, P.H., van de Steeg, J., Veldkamp, A., Willemen, L., 2009. From land coverchange to land function dynamics: a major challenge to improve land charac-terization. J. Environ. Manage. 90, 1327–1335.

Volante, J.N., Alcaraz-Segura, D., Mosciaro, M.J., Viglizzo, E.F., Paruelo, J.M., 2012.Ecosystem functional changes associated with land clearing in NW Argentina.Agric. Ecosyst. Environ. 154, 12–22.

von Haaren, C., Albert, C., 2011. Integrating ecosystem services and environmentalplanning: limitations and synergies. Int. J. Biodivers. Sci. Ecosyst. Serv. Manage.7, 150–167.

Wada, Y., Rajan, K.S., Shibasaki, R., 2007. Modelling the spatial distribution of shiftingcultivation in Luangprabang, Lao PDR. Environ. Plann. B-Plann. Des. 34, 261–278.

Wall, L., Larocque, D., Leger, P.-M., 2008a. The early explanatory power of NDVI incrop yield modelling. Int. J. Remote Sens. 29, 2211–2225.

Wall, L., Larocque, D., Leger, P.M., 2008b. The early explanatory power of NDVI incrop yield modelling. Int. J. Remote Sens. 29, 2211–2225.

Wallace, K.J., 2007. Classification of ecosystem services: problems and solutions.Biol. Conserv. 139, 235–246.

Wang, B., An, H.-J., Lu, C.-W., 2013a. Inversion modeling of dissolved oxygen in HulunLake of Northeast China based on multi-source remote sensing. ShengtaixueZazhi 32, 993–998.

Wang, L., Huang, J., Du, Y., Hu, Y., Han, P., 2013b. Dynamic Assessment of soil erosionrisk using landsat TM and HJ satellite data in Danjiangkou Reservoir Area, China.Remote Sens. 5, 3826–3848.

Wang, Z., Zhang, B., Zhang, S., Li, X., Liu, D., Song, K., Li, J., Li, F., Duan, H., 2006. Changesof land use and of ecosystem service values in Sanjiang Plain, Northeast China.Environ. Monit. Assess. 112, 69–91.

Wardlow, B.D., Egbert, S.L., 2008. Large-area crop mapping using time-series MODIS250 m NDVI data: an assessment for the US Central Great Plains. Remote Sens.Environ. 112, 1096–1116.

White, P.C.L., Godbold, J.A., Solan, M., Wiegand, J., Holt, A.R., 2010. Ecosystemservices and policy: a review of coastal wetland ecosystem services and anefficiency-based framework for implementing the ecosystem approach. Ecosys-tem Services-Book, vol. 30., pp. 29–51.

Wu, C.G., Zhou, Z.X., Xiao, W.F., Wang, P.C., Teng, M.J., Huang, Z.J., 2011. Estimationof soil erosion in the Three Gorges Reservoir Area of China using RUSLE, remotesensing and GIS. J. Food Agric. Environ. 9, 728–734.

Wu, Y.Z., Chen, J., Ji, J.F., Tian, Q.J., Wu, X.M., 2005. Feasibility of reflectancespectroscopy for the assessment of soil mercury contamination. Environ. Sci.Technol. 39, 873–878.

Xie, H., Tong, X.-H., Qiu, Y.-L., Zhang, H.-E., Mang, Y.-L., Zhao, J.-F., 2007. Monitor-ing model and temporal-spatial changes of dissolved oxygen based on remotesensing: a case study in Huangpu river. Spectrosc. Spectr. Anal. 27, 1574–1579.

Xie, Y.J., Ng, C.N., 2013. Exploring spatio-temporal variations of habitat loss and itscausal factors in the Shenzhen River cross-border watershed. Appl. Geogr. 39,140–150.

Xiulin, L., Junfang, C., Aiqin, S., 2013. Coastal wind energy assessment in the EastChina sea with SAR imagery. In: Proceedings of Dragon 2 Final Results andDragon 3 Kick-Off Symposium, p. 675.

Yang, W., Wang, M., Shi, P., 2013. Using MODIS NDVI time series to identify geo-graphic patterns of landslides in vegetated regions. IEEE Geosci. Remote Sens.Lett. 10, 707–710.

Yong, Z., Gao, J., Jianjun, J., Heng, L., Jiazhu, H., 2010. Monitoring of urban air pollutionfrom MODIS aerosol data: effect of meteorological parameters. Tellus, Ser. B(Chem. Phys. Meteorol.) 62B, 109–116.

Zeng, Y., Li, H.T., Wang, G.L., van Genderen, J.L., 2012. A rule-based soft-decisionmethod for change detection. In: 2011 International Conference on ComputerScience and Network Technology (Iccsnt), vols. 1–4, pp. 315–318.

Zhang, D.S., Lan, Z.R., Wang, Q.M., Wang, X.Q., Zhang, W., Li, Z., 2007. The evaluationof the Mangrove ecosystem services value change in Zhangjiang River Estuarybased on remote sensing. In: Igarss: 2007 IEEE International Geoscience andRemote Sensing Symposium, vols. 1–12, pp. 2302–2305.

Zhao, B., Kreuter, U., Li, B., Ma, Z.J., Chen, J.K., Nakagoshi, N., 2004. An ecosystemservice value assessment of land-use change on Chongming Island, China. LandUse Policy 21, 139–148.

Zhao, D., Cai, Y., Jiang, H., Xu, D., Zhang, W., An, S., 2011. Estimation of water clarityin Taihu Lake and surrounding rivers using Landsat imagery. Adv. Water Resour.34, 165–173.

Zhu, L., Gong, H., Zhao, W., Li, X., Zhang, J., Lin, F., Zhou, B., Xiao, Y., 2009. Spectral
features analysis of water quality in Baiyangdian wetland, China. In: 2009 17thInternational Conference on Geoinformatics, 4 pp.
Zinnert, J.C., Shiflett, S.A., Vick, J.K., Young, D.R., 2011. Woody vegetative coverdynamics in response to recent climate change on an Atlantic coast barrierisland: a remote sensing approach. Geocarto Int. 26, 595–612.

Remote sensing of ecosystem services: A systematic review

Documents

Transcript of Remote sensing of ecosystem services: A systematic review