En i on. Sci. P oc. 2021, 3, 30. h ps://doi.o g/10.3390/IECF2020-07973 www.mdpi.com/jou nal/en i onscip oc
P oceeding Pape
Mode nized Fo es Fi e Risk Assessmen Model Based on he
Case S udy o h ee Po uguese Municipali ies F equen ly
A ec ed by Fo es Fi es
†
Luis San os
1,2,
*, Vasco Lopes
1
and Cecília Bap is a
1,3
1
Poly echnic Ins i u e o Toma , School o Technology, 2300-313 Toma , Po ugal;
[email protected] (V.L.); ceci[email p o ec ed] (C.B.)
2
Geosciences Resea ch Cen e , Coimb a Uni e si y, 3030-790 Coimb a, Po ugal
3
Cen e o Technology, Res o a ion and A Enhancemen (Techn&A ), 2300-313 Toma , Po ugal
* Co espondence: lsan [email p o ec ed]; Tel.: +351-967743365
† P esen ed a he 1s In e na ional Elec onic Con e ence on Fo es s—Fo es s o a Be e Fu u e: Sus ainabili y,
Inno a ion, In e disciplina i y, 15–30 No embe 2020; A ailable online: h ps://iec 2020.sci o um.ne .
Abs ac : The numbe o o es i es igni ions has dec eased wo ldwide, hus obse ing inc eased
le els o in ensi y and des uc ion, endange ing u ban a eas and causing ma e ial damages and
dea hs (Po ugal, 2017). Fo es i e haza d mapping suppo ed by he su eillance s a egy a ge ed
a e y suscep ible a eas wi h high losses po en ial a e he common ools o i e p e en ion. Each
municipali y c ea es i s own Fo es Fi e Haza d Map, and so i is obse ed ha along he adminis-
a i e bounda ies, disc epancies occu , e en when iden ical ypes o land use a e in place. The e o-
lu ion o geog aphic in o ma ion sys ems echnology sus ained by he open-sou ce sa elli e im-
age y, along wi h he inno a i e Habi a Risk Assessmen model o he InVEST so wa e, allowed
he c ea ion o an easily applicable ans-adminis a i e bounda y i e haza d map, wi h equen
upda e capabili ies and ully open sou ce. This wo k conside ed h ee municipali ies (Toma , Ou-
ém, and Fe ei a do Zêze e) ha annually obse e a ious o es i e occu ences. Resul s enabled
he c ea ion o a homogeneous Fo es Fi e Risk Map, using landuse, slope, oad access ne wo k, i e
igni ions’ his o y, isualiza ion basins, and he No malized Di e ence Vege a ion Index (NDVI) as
a iables. All a iables co ela e wi h each o he using di e en weigh s, in which he di e en clas-
ses o land use a e conside ed as habi a s and he emaining a iables as i e haza d s esso s. The
esul s p oduce a cohe en mon hly upda ed Risk Map, which is an al e na i e o many isk assess-
men sys ems used wo ldwide.
Keywo ds: i e haza d; InVEST; NDVI; model; o es i es
1. In oduc ion
Wild i es a ise na u ally om sou ces such as spon aneous combus ion o d y ma -
e , unde high empe a u es and wind condi ions, o mos commonly ligh ning s ikes
[1]. Mos equen ly, wild i e igni ions a ise as acciden al consequences o human p ac-
ices, hough many ha e been p o en o be c iminal o negligen ac ions [2,3].
Medi e anean o es s a e egula ly subjec ed o a la ge numbe o i es [4–6],
changes in land use pa e ns [7], and he impac s o socio-economic ac o s on land man-
agemen p ac ices [8], u he agg a a ed by majo modi ica ions o o es ecosys ems
du ing he second hal o he 20 h cen u y [9] along wi h o es modi ica ion by i e ecu -
ence [10] con ibu ed owa ds wild i e in ensi ica ion.
Wi h a mland abandonmen and o es modi ica ion ends, i e concen a ion is be-
coming a he p edominan in Sou he n Eu opean egions [4]. In he pas yea s (2018–
2019), i es in Eu ope obse ed o e 800,000 hec a es o bu n a eas in Po ugal, I aly, and
Ci a ion: San os, L.; Lopes, V.;
Bap is a, C. Mode nized Fo es Fi e
Risk Assessmen Model Based on
he Case S udy o h ee Po uguese
Municipali ies F equen ly A ec ed
by Fo es Fi es. En i on. Sci. P oc.
2021, 3, 30. h ps://doi.o g/
10.3390/IECF2020-07973
Academic Edi o s: Angela Lo
Monaco, Ca e Macinnis-Ng and
Om P. Rajo a
Published: 12 No embe 2020
Publishe ’s No e: MDPI s ays neu-
al wi h ega d o ju isdic ional
claims in published maps and ins i u-
ional a ilia ions.
Copy igh : © 2020 by he au ho s.
Licensee MDPI, Basel, Swi ze land.
This a icle is an open access a icle
dis ibu ed unde he e ms and con-
di ions o he C ea i e Commons A -
ibu ion (CC BY) license (h p://c e-
a i ecommons.o g/licenses/by/4.0/).
En i on. Sci. P oc. 2021, 3, 30 2 o 9
Spain alone [11]. The numbe o occu ences ega ding o es i es has dec eased wo ld-
wide [12], hus obse ing inc eased le els o in ensi y and des uc ion, which in many
cases endange u ban a eas, causing no only ma e ial damage bu also dea hs, such as
hose ha occu ed in Po ugal in 2017 [9]. Wild i es, and in pa icula o es i es, a e
inc easing in in ensi y, eaching ca as ophic dimensions, as a esul o poo o es man-
agemen , e i o ial planning, and economic p essu es, also d i en by agg a a ed clima e
change condi ions [7,10,13].
Su ely i e ep esen s an impo an ecological unc ion and a s a egic ag icul u al
ool o e ilize soils [14,15]; howe e , he i e- igh ing associa ed cos s, he human li es
[16], he p ope y and na u al esou ces los [11], he en i onmen al g eenhouse gas emis-
sions and a mosphe ic pollu ion ac o s [17], and he ecosys em se ices los [15] ep esen
hea y coun e -weigh s o sus ainabili y scales.
Wild i es can be e med as o es i es, g ass i es, pea i es, and bush i es, depend-
ing on he p edominan ype o ege a ion ha is being bu n . Fo es i es in Po ugal, as
in he Medi e anean egion, can occu a any gi en ime; howe e , hey a e mo e p e a-
len du ing he d y season be ween he beginning o summe and he s a o he win e
ains [4,18]. Clima e change e ec s, inc easing d ough equency, and associa ed a e age
empe a u e ise, when combined wi h popula ion densi y and economic ac i i y, which
inc ease he u ban o es in e ace, will u he inc ease wild i e equency and se e i y
[8,9]. All ca as ophic i es in Po ugal occu wi hin he summe pe iod, when he associ-
a ion o h ee a iables, hea , wind, and lack o humidi y, align wi h he 30’s ule: o e 30 km/h
winds, o e 30° Celsius, and less han 30% humidi y.
The e olu ion o geog aphic in o ma ion sys ems (GIS) wi h a panoply o so wa e
and applica ions, bo h comme cial and open sou ce, enables easy i e- isk ca og aphy,
which equi es a dynamic p ocess, since a ce ain loca ion in a gi en ime quickly changes
ei he by seasonal changes, land occupa ion, o land use, hus p o iding a undamen al
ool in suppo o o es i e p e en ion, su eillance, and mi iga ion [19,20]. Wild i e mi -
iga ion analysis e ol ed conside ably since he u n o he cen u y: sa elli e image y and
emo e sensing (RS) analysis a e now widely used o de ec and moni o he beha io o ,
and aid ope a ional i e- igh ing ac i i ies, and acili a e he mapping o bu ned a eas.
These echnologies a e p o en o be mo e e icien and p ecise han adi ional su eying
me hods, educing isk and bu n a eas and sa ing human li es [20].
By using models, maps, and da abases in geospa ial analysis o i e a iables, e e y
coun y p one o wild i es p oduces isk maps as a s a egy o op imize moni o ing and
minimize losses du ing he seasonal i e- igh ing season. Risk mapping ocuses on low-
p obabili y, high-consequence ad e se e en s ha a e s ochas ic in space. As an ope a-
ional esul , coun ies alloca e wa ch owe s, open o es accessibili y acili ies, dis ibu e
acili ies, and su eying mobile eams in isky a eas [16].
The Po uguese municipali y isk map (PMDFCI) p oduc ion guidelines p oposed
by he go e nmen al Ins i u e o Na u e Conse a ion and o es y (ICNF) con empla es
Haza d as a mul iplica ion p oduc o p obabili y and suscep ibili y and po en ial losses
as a mul iplica i e p oduc o ulne abili y and economic alue, whe e isk is he mul i-
plica i e esul o haza d and po en ial damage (Figu e 1).
En i on. Sci. P oc. 2021, 3, 30 3 o 9
Figu e 1. PMDFCI p oduc ion ins uc ions (Sou ce ICNF).
The p obabili y c i e ion conside s i e his o y as he p oduc o occu ences mul i-
plied by 100, di ided by he numbe o se ies (yea s), whe eas suscep ibili y conside s
land-co e in eigh ca ego ies wi h o es and b ush wi h highe suscep ibili y and slope,
whe e s eepe slopes a e conside ed mos suscep ible. The ulne abili y c i e ion is ap-
plied o combus ion esilience o ma e ials and economic ma ke alue o goods.
Some o he unde lying p oblems obse ed wi h PMDFCI me hodology conce n he
indi idually implemen ed municipali y base isk maps which c ea e ha d bounda ies
whe e he same habi a ac oss wo municipali ies obse es a di e en isk class. Fu he -
mo e, he model is mos ly p oduced wi h s a ic cha ac e is ics whe e slow upda e a ia-
bles a e used (Figu e 2).
(a)
(b)
Figu e 2. (a) PMDFCI isk model o he s udy egion, (b) PMDFCI Risk de ail o municipali y bounda ies.
The objec i e o his pape is o p oduce an open sou ce, upda able and upg adable,
ansbounda y base solu ion o o es i e igni ion isk mapping and use i o iden i ying
loca ions, de e mined om a se o slow and dynamic a iables, hus mi iga ing suscep i-
bili y and social/in as uc u al ulne abili y.
En i on. Sci. P oc. 2021, 3, 30 4 o 9
2. Ma e ials and Me hods
Wi h he e olu ion o geog aphic in o ma ion echnology and he mul i ude o ee
sa elli e image y, he me hodology used in he Habi a Risk Assessmen model, one o he
InVEST (In eg a ed Valua ion o Ecosys em Se ices and T adeo s) models used o map
and alue he goods and se ices om na u e ha sus ain and ul il human li e. The In-
VEST Habi a and Species Risk Assessmen (HRA) model allows he assessmen o cumu-
la i e isk posed o habi a s by human ac i i ies, deli e ing ecosys em isk map, isk maps
o each indi idual habi a , esul ing om he con ibu ion o exposu e and consequence
o o e all isk [21].
This wo k enabled he c ea ion o a homogeneous Fi e Igni ion Risk map, which used
as a iables land occupa ion, slope, o es oad ne wo k, his o y o i e igni ions, isuali-
za ion basins, and he No malized Di e ence Vege a ion Index (NDVI). All a iables co -
ela ed wi h each o he using di e en weigh s, in which he di e en classes o land oc-
cupa ion en e as habi a s and he emaining as s esso s o he p oblem o i es.
2.1. S udy A ea
The s udy egion loca ed in he cen e o Po ugal co e ing he ansi ion be ween
he Tagus i e loodplains and he Mon ejun o-Es ela moun ain ange (Cen al Po ugal;
Figu e 3). Adminis a i ely, he egion comp ises he municipali ies o Toma , Ou ém,
and Fe ei a do Zêze e belonging o San a ém dis ic , Nomencla u e o Te i o ial Uni s
o S a is ical Pu poses (NUTS) le el II and III; he municipali ies a e inse ed in he Cen-
al egion and in he sub- egion o he Middle Tagus. They a e si ua ed in he Medi e a-
nean plu ioseasonal oceanic bio-clima ic egion [22] wi h a e age al i ude anging om
30 o 650 m s a l., mean annual p ecipi a ion o 66 mm, and a mean annual empe a u e o
14 °C. This egion obse es on a e age 2–3 mon hs o summe d ough when wild i es
a e equen , pa icula ly because o he associa ion wi h poo managemen , land aban-
donmen as a consequence o ageing popula ion and dese i ica ion, and he choice o
species in luenced by economic yields.
Figu e 3. Geog aphical loca ion o s udy egion.
En i on. Sci. P oc. 2021, 3, 30 5 o 9
The egion was once p ospe ous wi h Pinus pinas e plan a ions speckled wi h oli e
o cha ds and lowland ag icul u e; nowadays i is a mosaic o Eucalyp us globulus plan a-
ions, emnan s o Pinus pinas e , and sh ubland domina ed by Ulex spp., E ica spp., Pis a-
cea sp., My us sp., and Rubus spp., whe e once was ag icul u e, con i ming a end ha is
u he obse ed as we mo e no h-eas wa ds. Acco ding o he da a o he O icial Ad-
minis a i e Cha o Po ugal (DGT-2019), he h ee municipali ies co e an a ea o ap-
p oxima ely 95,823 ha, di ided as ollows: Toma (35,120 ha), Ou ém (41,666 ha), and Fe -
ei a do Zêze e (19,037 ha).
2.2. Me hods
The me hodology applied o he cu en s udy used image y om he sa elli e con-
s ella ion Sen inel-2 mul ispec al ins umen (MSI), which o e ed he possibili y o ob-
aining in o ma ion in medium and high spa ial esolu ion (10−20 m). In pa icula , he
MSI senso p o ides spec al in o ma ion in di e en bands, allowing o he calcula ion
o p e- i e ege a ion g eenness, he no malized di e ence ege a ion index (NDVI, The-
o em 1). Fo g eenness chlo ophyll concen a ion NDVI, we used Red and NIR acco ding
o he o mula desc ibed on Theo em 1.
𝑁𝐷𝑉𝐼 =
𝑁𝐼𝑅 − 𝑅𝑒𝑑
𝑁𝐼𝑅 + 𝑅𝑒𝑑 (1)
Theo em 1. No malized di e ence ege a ion index (NDVI), ed egion (Red: 650–680 nm, 10 m
esolu ion) being he nea in a ed (NIR: 785–899 nm, 10 m esolu ion).
The downloaded images o he s udy egion, le el 2A, a mosphe ically, adiome i-
cally, and geome ically co ec ed, om he July 24 o Augus 23, 2019, wi hou he p es-
ence o clouds, whe e he inal da a was selec ed o alues up o 0.4, conside ing loss o
chlo ophyll as he dynamic a iable wi h mon hly images equency [20,23,24].
The Mode nized Dynamic Igni ion Risk (MDIR) model (Figu e 4) conside s igni ion
isk as he igge o i e occu ence, bo h la ge and small; his ac o con eys a sa e ap-
p oach o moni o ing and mi iga ion ac ions o i e igh ing agen s. The slow geoda abase
a iables used we e land-co e wi h o es and b ush ep esen ing highe isk and an-
h opic e i o ies such as ci ies lowe om Na ional Landco e Maps Le el 1 (COS). Road
ne wo k selec ed 20 m high isk, conside ing mos igni ions s a wi hin he p oximi y o
oads, hus emphasizing he c iminal and acciden al majo i y o igni ions om he Na-
ional Ins i u e o Na u e conse a ion and o es y (ICNF) [25]. The his o y o i e igni-
ions, conside ing ha a bu n a ea will, acco ding o na u al egene a ion, ha e some
yea s be o e ecu en igni ion, means ha ecological succession will ep esen highe
isk. Da a used a e om 2011–2019 om he Na ional Se ice o Na u e and En i onmen
P o ec ion (SEPNA). Popula ion densi y (people pe sq. km) is acco ding o census 2011,
whe e a eas bellow 250 esiden s inc ease he isk, dese i ica ion, and aged popula ion,
om he Na ional S a is ics Ins i u e (INE). As low slopes a e conside ed p one a eas o
igni ion, his may cause con usion, as high slopes a e a o able o i e p opaga ion; how-
e e , igni ion da a om he pas nine yea s e eals ha he majo i y o igni ions occu in
lowe slope a eas; slope was calcula ed wi h QGIS om DTM 25 m esolu ion. Visualiza-
ion h ough he cons uc ion o wa ch owe s is one o he moni o ing s a egies imple-
men ed wo ldwide; o his s udy, all a eas no isible by a leas h ee owe s a e consid-
e ed o high isk; isualiza ion was calcula ed using he wa ch owe loca ions, in u n ap-
plying iewshed analysis QGIS plug-in (SEPNA).
The MDIR model adop s ee open sou ce so wa e QGIS and GRASS o p epa e all
ec o and as e inpu s o he InVEST HRA, applying Landco e as he cu en habi a
spli in o indi idual ca ego ies, wi h slope, isibili y, oads, i e igni ion his o y, and pop-
ula ion densi y as s esso s. Requi ing a classi ica ion c i e ia able, HRA, classi ies each
a iable as exposu e o consequence; he habi a s will be a ed o g ow h a e, o a ion,

En i on. Sci. P oc. 2021, 3, 30 6 o 9
connec i i y, and na u al eco e y ime based on li e a u e in o ma ion, e alua ing da a
quali y and weigh . Fo each s esso , a ings we e conside ed o dis ibu ion equency,
change in classi ica ion, managemen e iciency, and neighbo hood, also e alua ing da a
quali y and weigh .
Figu e 4. Mode nized Dynamic Igni ion Risk model (MDIR), elabo a ion p ocess.
The InVEST HRA model cumula i ely analyses habi a cha ac e is ics as mi iga ion
ac o s and s esso s as de e io a ing ac o o isk p oducing isk maps o he s udy egion.
3. Resul s and Discussion
Resul s deli e ed as isk maps om he HRA o he s udy egion we e es ed using
he di e en se s o s esso s o unde s and he di e ences be ween slow and dynamic
a iables. As a i s un o he HRA, only slow a iables we e analyzed and isk classes
we e a ibu ed as Mino , Gua ded, Ele a ed, and Se e e, using designa ed Mode nized
Igni ion Risk (MIR) (Figu e 5a). Classi ica ion conside s isk is always p esen excep o
wa e su aces, and he e o e no isk classi ica ion was no used in he isk map.
Resul s gene a ed om he InVEST HRA in he p oduc ion o he MIR map (Figu e 5a)
a e conside ably di e en om he PMDFCI; when we ocus on municipal bounda ies, a
so ansbounda y isk classi ica ion can be obse ed (Figu e 5b).
The inclusion o he dynamic (mon hly) g eenness (NDVI) s ess, denomina ed Mod-
e nized Dynamic Igni ion Risk model (MDIR), deli e s a subs an ially di e en igni ion
isk map (Figu e 6a,b).
En i on. Sci. P oc. 2021, 3, 30 7 o 9
(a)
(b)
Figu e 5. Igni ion Risk Model (MIR) map conside ing slow a iables (a) MIR o he en i e egion, (b) MIR de ail o in e
municipali y bounda ies.
(a)
(b)
Figu e 6. Igni ion isk map wi h slow a iables using dynamic NDVI as a s esso , (a) NDVI—July 24, 2019, (b) NDVI—
Augus 23, 2019.
When compa ing MIR and MDIR model esul s wi h he PMDFCI, he 10% inc ease
o se e e isk class in MIR and he 7% dec ease in MDIR (Figu e 7a) a e clea ly isible; he
o me is based on slow a iables, and one can jus a ibu e he la e educ ion o he
dynamic NDVI s ess, which epe cusses he clima ic in luence, hus pinpoin ing se e e
En i on. Sci. P oc. 2021, 3, 30 8 o 9
isk a eas. This ac may e eal o be an in e es ing cue o moni o ing and su eying s a -
egies.
On he o he hand, lowe classi ica ions, mino and gua ded, when added, a e 49%
o PMDFCI, 48% o MDIR, and 41% o MIR. The close pe cen age obse ed be ween
PMDFCI and MDIR indica es ha na u al condi ions a e equally conside ed in bo h mod-
els, u he e idenced in he geog aphical span; howe e , MDIR may, due o i s dynamic
na u e (mon hly), exp ess clima e change condi ions, whe eas PMDFCI is s a ic.
Figu e 7b compa es MDIR esul s be ween July 24 and he Augus 23, 2019, whe e
he e is a sligh isk inc ease (0.1%), excep o he mino isk class ha dec eased 0.5%;
alues may seem esidual; howe e , chlo ophyll loss in some well adap ed plan s is no a
apid p ocess.
(a)
(b)
Figu e 7. (a) Compa ison ba cha be ween he wo a ia ions o he mode nized model and
PMDFCI, (b) Compa ison pie cha be ween he MDIR o July 24 and Augus 23, 2019.
Model unc ional p ac icabili y is p obably he mos impo an cha ac e is ic o p e-
dic ed esul s, hence he need o u he exploi s esso s and mi iga ion a iables o op i-
mize MDIR.
4. Conclusions
The in eg a ed analysis o spa ial slow and dynamic a iables enabled by he InVes
HRA model a e aluable esou ces o o es i e esea ch, and when used wi h emo e
sensing sa elli e image y combined wi h GIS p ocessing, made possible he c ea ion o i e
igni ion models MIR and MDIR. The p oduced models used li e a u e p o en alid a i-
ables and pe o med p ope ly in iden i ying a eas o high igni ion isk.
The MIR model o e s a gene ic well-de ined se e e isk a ea and may be use ul as a
gene al-pu pose igni ion isk map, whe eas he MDIR o e s a possible solu ion o mon-
i o ing and su eillance, despi e he needed esea ch and alida ion. MDIR model o e s
a solu ion ha using NDVI indi ec ly includes he clima e change, hough being he only
dynamic a iable would bene i om sho wa e in a ed, No malized Dis ance Wa e
Index (NDWI).
Al hough much esea ch is needed, he p oposed models a e ully upda able and
buil en i ely wi h open sou ce so wa e and hence a e accessible o anyone.
Da a A ailabili y S a emen : All da a used in he cu en s udy is unde access ia public In e ne
and GEANT ne wo k, wi h commi ed eliabili y and pe o mance on h ps://sen inels.cope ni-
cus.eu/web/sen inel/home.
Re e ences
1. Cou o, F.T.; Iakunin, M.; Salgado, R.; Pin o, P.; Viegas, T.; Pin y, J.P. Ligh ning modelling o he esea ch o o es i e igni ion
in Po ugal. A mos. Res. 2020, 242, 104993.
2. Koun ou is, Y. Human ac i i y, dayligh sa ing ime and wild i e occu ence. Sci. To al En i on. 2020, 727, 138044.
3. Balch, J.K.; B adley, B.A.; Aba zoglou, J.T.; Nagy, R.C.; Fusco, E.J.; Mahood, A.L. Human-s a ed wild i es expand he i e niche
ac oss he Uni ed S a es. P oc. Na l. Acad. Sci. USA 2017, 114, 2946–2951.
En i on. Sci. P oc. 2021, 3, 30 9 o 9
4. Ga cía-Llamas, P.; Suá ez-Seoane, S.; Taboada, A.; Fe nández-Manso, A.; Quin ano, C.; Fe nández-Ga cía, V.; Fe nández-
Guisu aga, J.M.; Ma cos, E.; Cal o, L. En i onmen al d i e s o i e se e i y in ex eme i e e en s ha a ec Medi e anean
pine o es ecosys ems. Fo . Ecol. Manag. 2019, 433, 24–32.
5. Oli ei a, S.; Oehle , F.; San-Miguel-Ayanz, J.; Camia, A.; Pe ei a, J.M.C. Modeling spa ial pa e ns o i e occu ence in
Medi e anean Eu ope using Mul iple Reg ession and Random Fo es . Fo . Ecol. Manag. 2012, 275, 117–129.
6. Quin ano, C.; Fe nández-Manso, A.; S ein, A.; Bijke , W. Es ima ion o a ea bu ned by o es i es in Medi e anean coun ies:
A emo e sensing da a mining pe spec i e. Fo . Ecol. Manag. 2011, 262, 1597–1607.
7. Fe nandes, P.M. Fi e-sma managemen o o es landscapes in he Medi e anean basin unde global change. Landsc. U ban
Plan. 2013, 110, 175–182.
8. Galizia, L.F.D.; Rod igues, M. Modeling he in luence o eucalyp plan a ion on wild i e occu ence in he B azilian sa anna
biome. Fo es s 2019, 10, 844.
9. Na alho, I.; Aleg ia, C.; Quin a-No a, L.; Fe nandez, P. In eg a ed planning o landscape di e si y enhancemen , i e haza d
mi iga ion and o es p oduc ion egula ion: A case s udy in cen al Po ugal. Land Use Policy 2017, 61, 398–412.
10. Pue a-Piñe o, C.; Espel a, J.M.; Sánchez-Humanes, B.; Rod igo, A.; Coll, L.; B o ons, L. His o y ma e s: P e ious land use
changes de e mine pos - i e ege a ion eco e y in o es ed Medi e anean landscapes. Fo . Ecol. Manag. 2012, 279, 121–127.
11. Alló, M.; Lou ei o, M.L. Assessing p e e ences o wild i e p e en ion policies in Spain. Fo . Policy Econ. 2020, 115, 102145.
12. Andela, N.; Mo on, D.C.; Giglio, L.; Chen, Y.; an de We , G.R.; Kasibha la, P.S.; DeF ies, R.S.; Colla z, G.J.; Han son, S.;
Klos e , S.; e al. A human-d i en decline in global bu ned a ea. Science 2017, 356, 1356–1362.
13. Be key, J.K.; Belo e, R.T.; Mahe , C.T.; La son, A.J. S uc u al di e si y and de elopmen in ac i e i e egime mixed-coni e
o es s. Fo . Ecol. Manag. 2021, 479, 118548.
14. Russo, A.; Gou eia, C.M.; Páscoa, P.; DaCama a, C.C.; Sousa, P.M.; T igo, R.M. Assessing he ole o d ough e en s on wild i es
in he Ibe ian Peninsula. Ag ic. Fo . Me eo ol. 2017, 237–238, 50–59.
15. Zai se , A.S.; Gongalsky, K.B.; Malms öm, A.; Pe sson, T.; Beng sson, J. Why a e o es i es gene ally neglec ed in soil auna
esea ch? A mini- e iew. Appl. Soil Ecol. 2016, 98, 261–271.
16. Doe , S.H.; San ín, C. Global ends in wild i e and i s impac s: Pe cep ions e sus eali ies in a changing wo ld. Philos. T ans.
R. Soc. B Biol. Sci. 2016, 371, 1696.
17. Sannig ahi, S.; Pilla, F.; Basu, B.; Basu, A.S.; Sa ka , K.; Chak abo i, S.; Joshi, P.K.; Zhang, Q.; Wang, Y.; Bha , S.; e al. Examining
he e ec s o o es i e on e es ial ca bon emission and ecosys em p oduc ion in India using emo e sensing app oaches. Sci.
To al En i on. 2020, 725, 138331.
18. Häusle , M.; Nunes, J.P.; Sil a, J.M.N.; Keize , J.J.; Wa neke, T.; Pe ei a, J.M.C. A p omising new app oach o es ima e d ough
indices o i e dange assessmen using emo ely sensed da a. Ag ic. Fo . Me eo ol. 2019, 274, 195–209.
19. He nandez-Leal, P.A.; A belo, M.; Gonzalez-Cal o, A. Fi e isk assessmen using sa elli e da a. Ad . Sp. Res. 2006, 37, 741–746.
20. Chu ieco, E.; Congal on, R.G. Applica ion o emo e sensing and geog aphic in o ma ion sys ems o o es i e haza d mapping.
Remo e Sens. En i on. 1989, 29, 147–159.
21. Gaglio, M.; Aschoni is, V.; Pie e i, L.; San os, L.; Gissi, E.; Cas aldelli, G.; Fano, E.A. Modelling pas , p esen and u u e
Ecosys em Se ices supply in a p o ec ed loodplain unde land use and clima e changes. Ecol. Modell. 2019, 403, 23–34.
22. Ri as-Ma ínez, S.; Ri as-Sáenz, S.; Penas-Me ino, A. Wo ldwide bioclima ic classi ica ion sys em. Glob. Geobo . 2011, 1, 1–638.
23. Fe nández-Manso, A.; Fe nández-Manso, O.; Quin ano, C. SENTINEL-2A ed-edge spec al indices sui abili y o
disc imina ing bu n se e i y. In . J. Appl. Ea h Obs. Geoin . 2016, 50, 170–175.
24. Ch ysa is, I.; Mallinis, G.; Tsaki i, M.; Pa ias, P. E alua ion o single-da e and mul i-seasonal spa ial and spec al in o ma ion o
Sen inel-2 image y o assess g owing s ock olume o a Medi e anean o es . In . J. Appl. Ea h Obs. Geoin . 2019, 77, 1–14.
25. Gó iz-Mi sud, E.; Bu ns, M.; Go igli, V.M. Ci il socie y engaged in wild i es: Medi e anean o es i e olun ee g oupings.
Fo . Policy Econ. 2019, 102, 119–129.