LETTER Spatially explicit estimates of forest carbon ... · swamp forest, lowland forest, lower...

LETTER

Spatially explicit estimates of forest carbon emissionsmitigation costs and REDD+ opportunities in Indonesia

Victoria Graham14 Susan G Laurance1 Alana Grech2 and Oscar Venter3

1 College of Marine and Environmental Sciences and the Centre for Tropical Environmental and Sustainability Science James CookUniversity Cairns Queensland 4870 Australia

2 Australian Research Council Centre of Excellence for Coral Reef Studies James Cook University Townsville Queensland 4811Australia

3 Ecosystem Science and Management University of Northern British Columbia Prince George Canada4 Author to whom any correspondence should be addressed

E-mail victoriagrahammyjcueduau

Keywords carbon REDD+ prioritize spatial targets Indonesia forests

Supplementary material for this article is available online

AbstractCarbon emissions from the conversion and degradation of tropical forests contribute toanthropogenic climate change Implementing programs to reduce emissions from tropical forestloss in Southeast Asia are perceived to be expensive due to high opportunity costs of avoideddeforestation However these costs are not representative of all REDD+ opportunities as they aretypically based on average costs across large land areas and are primarily for reducingdeforestation from oil palm or pulp concessions As mitigation costs and carbon benefits canvary according to site characteristics spatially-explicit information should be used to assess cost-effectiveness and to guide the allocation of scarce REDD+ resources We analyzed the cost-effectiveness of the following REDD+ strategies in Indonesia one of the worldrsquos largest sources ofcarbon emissions from deforestation halting additional deforestation in protected areas timberand oil palm concessions reforesting degraded land and employing reduced-impact loggingtechniques in logging concessions We discover that when spatial variation in costs and benefits isconsidered low-cost options emerged even for the two most expensive strategies protectingforests from conversion to oil palm and timber plantations To achieve a low emissions reductiontarget of 25 we suggest funding should target deforestation in protected areas and oil palmand timber concessions to maximize emissions reductions at the lowest cumulative cost Low-costopportunities for reducing emissions from oil palm are where concessions have been granted ondeep peat deposits or unproductive land To achieve a high emissions reduction target of 75funding is allocated across all strategies emphasizing that no single strategy can reduce emissionscost-effectively across all of Indonesia These findings demonstrate that by using a spatially-targeted approach to identify high priority locations for reducing emissions from deforestationand forest degradation REDD+ resources can be allocated cost-effectively across Indonesia

1 Introduction

Tropical forests are important reservoirs of carboncontaining around half (55) of the carbon stored inforests worldwide (Pan et al 2011) Globally tropicalforests declined at a rate of sim05 pa for the period1990ndash2010 which equated to sim120 million ha (Achardet al 2014) and contributed tosim15 of anthropogeniccarbon emissions (Houghton 2013) Indonesia is one

of the largest contributors of carbon emissions fromtropical deforestation and degradation (Baccini et al2012) The Indonesian government have pledged tocurb the conversion of tropical lowland forests and oneof the initiatives they are supporting to achieve thisgoal is REDD+ (for Reducing Emissions fromDeforestation and forest Degradation plus conservingsustainably managing forests and enhancing forestcarbon stocks) REDD+ payments are intended to

OPEN ACCESS

RECEIVED

20 October 2016

REVISED

8 March 2017

ACCEPTED FOR PUBLICATION

13 March 2017

PUBLISHED

5 April 2017

Original content fromthis work may be usedunder the terms of theCreative CommonsAttribution 30 licence

Any further distributionof this work mustmaintain attribution tothe author(s) and thetitle of the work journalcitation and DOI

Environ Res Lett 12 (2017) 044017 httpsdoiorg1010881748-9326aa6656

copy 2017 IOP Publishing Ltd

provide the economic incentives needed to conserveforests by linking financial rewards to emissionsreduced or carbon sequestered (Agrawal et al 2011)When REDD+ was first conceived in 2005 it sought toReduce Emissions from Deforestation (RED see denBesten et al 2014) at which point it was chieflyconcerned with limiting tropical deforestation Duringearly-stage discussions the scope of REDD+ wasbroadened to include reducing degradation (REDD)as well as conserving and sustainably managing forestsand enhancing forest carbon stocks (REDD+) Thisdevelopment opened up a range of new opportunitiesfor addressing forest carbon loss including activitiesthat sequester carbon such as reforestation and thatreduce degradation such as reduced-impact logging(RIL Putz et al 2008 Alexander et al 2011)

Through its range of strategies REDD+ has thepotential to reduce carbon dioxide (CO2) concen-trations in the atmosphere which will aid in thetransition to a low fossil fuel global economy(Houghton et al 2015) Since its inception REDD+has attracted over US$73 billion in fundingincluding pledges of over US$2 billion to Indonesiaalone (Forest Trends Association 2016) A key issuehindering the implementation of REDD+ is how wellcost-effective climate mitigation activities align withthe rights of local forest users with concerns raisedthat the priorities of international investors will beprivileged over those of local communities (Howsonand Kindon 2015) Additionally economic concernshave been centred on the unlikelihood that REDD+will generate sufficient finance to off-set lost revenuesfrom alternative land-use activities drawing compar-isons against the moderate level of funding directedtowards REDD+ relative to the high profits generatedfrom deforestation-dependent activities such astimber and oil palm production (Venter and Koh2011) The literature shows that projects aimed atlimiting deforestation from large-scale oil palmproduction are expensive due to the high forgonerevenues (ie opportunity cost) from convertingforest into oil palm (eg Butler et al 2009 Venter et al2009 Fisher et al 2011a Irawan et al 2011 Ruslandiet al 2011)

An alternative and potentially cheaper pathway forREDD+ to contribute towards carbonmitigation is viareforestation reducing illegal deforestation in pro-tected areas (PAs) and reducing forest degradationThe optimal approach to allocating REDD+ resourceswill be influenced by the spatial context in which eachproject is applied as costs and carbon benefits can varyspatially (Pagiola and Bosquet 2009) Site-specificfactors that influence costs and benefits includeterrain distance to markets and soil type (Gibbset al 2007 Pagiola and Bosquet 2009) Recent studiesundertaken in Indonesia highlight how applying aspatially-targeted approach to regional developmentplans can reduce the trade-offs of agricultural ortimber expansion and forest protection (Koh and

Ghazoul 2010 Venter et al 2012) A key questiontherefore is how does spatial variation influence theeffectiveness of REDD+ strategies to mitigate forest-based carbon emissions at low-cost across Indonesia

To address this question we used spatial analysesto assess the variation in costs and carbon benefits ofvarious REDD+ strategies in Indonesia and identifiedthe factors that drive cost-effectiveness We used mapsof carbon stocks forest cover peatlands and cropsuitability to estimate the potential for REDD+ to slowor reverse carbon emissions from oil palm timber andlogging permits PAs and on degraded land Weexplored the cost-effectiveness of REDD+ strategiesfor reducing one tonne of carbon and for achieving arange of emissions targets We compared the resultsfrom this spatial analysis to estimates from a cost-benefit analysis of REDD+ that used average costs andbenefits (Graham et al 2016) This paper is designed todeliver fine-scale information to policy makers onspatially-targeted opportunities for mitigating carbonemissions from deforestation and forest degradationin Indonesia

2 Materials and methods

In this paper we estimated the 30 yr carbon emissionsand financial costs from anticipated land conversionand determined the carbon sequestered from restoringland that is not slated for urban development oragriculture across Indonesia The term lsquopermitsrsquo refersto land use rights issued to companies for logging oilpalm or timber concessions We ranked all permitsPAs and reforestation sites by the cost of reducing onetonne of carbon (from low to high) to determine thecombination of strategies that achieve emissionstargets (25 50 75 100) most cost-effectivelyAll carbon values are in tonnes (1 tonne frac14 1 Mg) ofcarbon (C) Carbon dioxide (CO2) was converted tocarbon by dividing by 367 (van Kooten et al 2004)Biomass was converted to carbon by multiplying by0492 (Pinard and Putz 1996) All financial figures arein 2010 US dollars Here we present a summarizedversion of the steps involved in calculating the spatiallyexplicit emissions and costs individually for eachpermit PA or reforestation site See appendix S1(available at httpsstacksioporgERL12044017mmedia) in supporting information for details onthe input data and detailed methods

21 Estimating carbon benefits of REDD+ strategiesSpatial analysis was performed in ArcGIS v103 (ESRI2014) We used 250 m spatial resolution land covermaps for 2000 and 2010 that were produced usingModerate Resolution Imaging Spectroradiometer(MODIS) images and Daichi-Advanced Land Observ-ing Satellite data (Miettinen et al 2012b) We createdbinary maps of 2000 and 2010 natural forest cover(figure 1(a)) by classifying mangrove forest peat

Environ Res Lett 12 (2017) 044017

2

swamp forest lowland forest lower montane forestand upper montane forest as natural forest (hereafterreferred to as lsquoforestrsquo) To create a layer of deforesta-tion we used the erase function to estimate net forestloss for the decade 2000 to 2010 (figure 1(b)) Weresampled all layers to a sim250 m resolution to matchthe Miettinen et al (2012b) land cover dataset andprojected all spatial data into Asia South Albers EqualArea Conic We measured carbon emissions from lossof above- and below-ground carbon (AGC BGCfigures 1(c) and (d)) Baccini et al (2012) used fielddata and remote sensing to estimate and map AGC forall of Indonesia We used a rootshoot ratio of 21100to convert the AGC estimates from the Baccini map tototal carbon in natural forests and timber plantations(Saatchi et al 2011 Kotowska et al 2015) and 32100 inoil palm concessions and mixed-crops (Kotowska et al2015) We tested the sensitivity of our results tochanges in the forest cover and carbon input data byanalyzing all of the scenarios using high spatialresolution (30 m) land cover maps for 2000 and 2010(Hansen et al 2013) and a map of above- and below-ground biomass produced circa 2000 (Saatchi et al2011)mdashrefer to supporting information for detailsCarbon benefits refer to emissions reduced fromavoided deforestation and degradation as well ascarbon accrued from reforestation

211 Oil palm and timber concessionsWe overlayed maps of timber (Minnemeyer et al 2009figure 2(a)) and oil palm concessions (Greenpeace2011 figure 2(b)) with our 2010 forest cover and AGCmaps to estimate the total carbon contained within theforested part of each permit and estimate theemissions that would result from clearing the forest

and replacing it with plantations (see appendix S1) If apermit was highly suitable for growing oil palm weaccounted for carbon stocks in the replacementvegetation (a carbon benefit) whereas if a permitwas unsuitable for oil palm we accounted for nocarbon benefit following deforestation if oil palmcould not be grown Prior to the establishment of oilpalm and timber plantations peat swamps are firstlydrained (FAO 2014) which leads to additional carbonemissions from the oxidation and increased probabili-ty of fires after draining We calculated the extent ofpeat emissions by intersecting the map of forestthreatened by oil palm and timber with a map ofcarbon stored in peat swamps (Minnemeyer et al 2009figure 1(d ))

212 Reduced-impact loggingEmploying RIL techniques to logging operations savesan additional 19 of the pre-harvest biomasscompared to conventional logging (CL Healey et al2000 Pinard and Cropper 2000 Putz et al 2008) Weestimated the emissions that could be reduced fromminimizing forest degradation during log-harvestingunder RIL practices by multiplying the 30 yr carbonbenefit of RIL (19) by the carbon stored in eachexisting logging concession (Minnemeyer et al 2009figure 2(c)) Selective logging of forests can beconducted without major disturbances to peathydrology (FAO 2014) and therefore we did notaccount for emissions from peat drainage in loggingpermits

213 Illegal deforestation within protected areasFor each terrestrial PA (IUCN and UNEP-WCMC2016 figure 2(d)) we projected 30 years of future

(A) (B)

(C) (D)

Forest cover

Carbon (tC per ha) Peat (tC per ha)2281000

Forest loss

80004000100

N

0 500 1000 2000Kilometers

Figure 1 Forest and carbon data used in the spatial analysis (a) 2010 forest cover (Miettinen et al 2012b) (b) forest loss between 2000and 2010 (c) terrestrial above-ground carbon for the period 2007ndash2008 (Baccini et al 2012) and (d) carbon stored in peat swamps for2002 (Minnemeyer et al 2009)


3

emissions from illegal activities using a linearextrapolation of deforestation observed over theperiod 2000ndash2010 (Miettinen et al 2012b) It iscommon in Indonesia to plant cacao oil palm rubberand coffee (hereafter lsquomixed-cropsrsquo) in PAs followingdeforestation (Swallow et al 2007) We estimated thecarbon lost by converting natural forests to mixed-crops (accounting for carbon stocks in replacementvegetation) and multiplied it by the deforestation rateto project the carbon emissions from illegal deforesta-tion We assumed that half of the deforestationactivities that occur on peat soils in PAs requiredrainage while the other half do not (FAO 2014) Wecalculated the extent of peat emissions by intersectingthe map of forest threatened by agriculture in PAs witha map of peat soil and multiplied this by 05 and by thedeforestation rate

214 Reforesting abandoned landWe overlayed a map of biomes (Olson et al 2001) withour 2010 forest cover map to find sites where forestspreviously existed but had been cleared We dis-regarded lsquoafforestationrsquo activities (planting forests inhistorically non-forest locations) We classified theWorld Wildlife Fund for Nature (WWF) biomeslsquotropical and subtropical moist broadleaf forestsrsquolsquotropical and subtropical dry broadleaf forestsrsquo andlsquomangrovesrsquo as lsquoforestrsquo We then refined our area tolsquohighly degraded landrsquo which we identified as areaswith less than 35 tCmiddotha1 (Baccini et al 2012) which isa recommended practice for identifying degradedforest lands in Indonesia (Gingold et al 2012) Weexcluded areas that overlapped with oil palm timberor logging concessions and all areas classified as lsquoAPLrsquo

which is outside of the national forest estate(Minnemeyer et al 2009) We created 2214 lsquohypotheti-cal management unitsrsquo for reforestation in areas of 900ha in size to compare against permits and PAs Weestimated the potential carbon benefit of reforestationprojects in Indonesia based on the 30 yr sequestrationrate (appendix S1) of regenerating tropical forests

22 Cost of reducing emissionsThe financial costs of employing each REDD+ strategyas calculated by Graham et al (2016 see appendix S1for details) included opportunity management andtransaction costs Most costs were presented as netpresent values which are the discounted value of thesum of projected future cash flows expected under thebusiness as usual scenario (Stone 1988) that wereextrapolated over 30 years at a discount rate of10 pa In this paper we modified the average perhectare costs based on spatially-explicit site character-istics Spatially-explicit opportunity costs of oil palmwere estimated by overlaying a suitability map for oilpalm (FAO 2012) to determine where oil palm isprofitable Opportunity costs of land that is unsuitablefor oil palm are restricted to the profits from timberextraction Conversely sites that have high suitabilityfor oil palm will generate larger revenues from itsproduction and sale as well as from timber extractionthan sites that have low or no suitability Depending ona plantationrsquos suitability we applied different costs topermits (see appendix S1) Costs for oil palm timberPAs and logging permits were calculated based on theforested part of the permit only

We calculated the cost of reducing emissions($middottC1) by dividing the total cost by the total carbon

(A) (B)

(C) (D)

Timber concessions

Logging concessions

Oil palm concessions

Protected areas

N


Figure 2 Cadastral data layers used in the spatial analysis (a) timber concessions (Minnemeyer et al 2009) (b) oil palm concessions(Greenpeace 2011) (c) logging concessions (Minnemeyer et al 2009) and (d) protected areas (IUCN and UNEP-WCMC 2016)


4

benefit for each permit PA and reforestation siteusing the formula below

Cost of reducing emissions eth$middottc1THORN frac14Total costeth$THORN

Total carbon benef itethtCTHORN

3 Results

Across Indonesia sim853 Mha of forest cover remainedas of 2010 of which logging concessions representedthe largest area (sim178 Mha 21 table 1) followedby PAs (sim138 Mha 16) oil palm concessions(sim300 Mha 4) and timber concessions (sim205Mha 2) Sites suitable for reforestation coveredsim500 Mha of degraded land We estimated themaximum potential 30 yr carbon benefit of employingfive REDD+ strategies (1) reforesting degraded landcould sequester 965MtC (2) limiting the expansion ofoil palm into forests could reduce 836 MtC (3)limiting the expansion of timber plantations intoforests could reduce 831 MtC(4) employing RILtechniques in logging concessions could reduce 638MtC and (5) halting illegal forest loss in PAs couldreduce 414 MtC On an annual basis the combinedcarbon benefit of applying these strategies acrossIndonesia is 123 MtC (3684 MtC over 30 years) at acost of $19 billion or $157 tC1

On average reforestation is cheaper than the otherstrategies assessed in terms of cost-effectiveness forreducing emissions ($9middottC1) but has no variance incosts due to the flat carbon sequestration rate appliedhere (figure 3) Oil palm and timber concessions andPAs had some of the cheapest (lt$7middottC1) and themost expensive sites (gt$200middottC1) for reducingemissions and the most variation (table 1) indicatingsite-specific factors strongly influence the cost ofreducing emissions at each permit or PA Cost-effective locations for reducing emissions from timberplantations (figure 4(a)) are where carbon-rich forests(eg peat forests) remain while expensive locations

have remaining forests of low quality Approximately40 of forested timber plantations in Indonesiaoverlapped with peat soils predominantly in easternSumatra storing on average twenty times morecarbon and making these permits four times cheaperfor reducing emissions than forests on mineral soils

To reduce emissions from oil palm cost-effectivelocations are mainly in Borneo (figure 4(b)) whereremaining forests occur on peat deposits (31 ofpermits with forest) or where land has climatic andedaphic conditions that is not highly suitable forcultivating oil palm (85 of permits with forest) Thecost of reducing emissions in oil palm permits withlow or no suitability (sim$39middottC1) is seven timescheaper than permits with high suitability(sim$265middottC1) Across Indonesia logging concessionsconsistently provide low-cost options for reducingemissions from forest degradation through oppor-tunities for employing RIL practices (figure 4(c))Cost-effective opportunities to reduce illegal forestcarbon loss in PAs occur on all islands (figure 4(d))and are characterized by high deforestation rates(gt3 pa between 2000 and 2010) and dense carbonstores (gt500 tCmiddotha1)

Table 1 Summary information on the total area (ha) cost (US$) and carbon benefit (C) of the following REDD+ strategies targetingdeforestation within timber and oil palm concessions halting illegal forest clearing in protected areas reforesting degraded land andemploying reduced-impact logging techniques at logging concessions Total figures are for all of Indonesia and means are the averageacross all permits protected areas or reforestation sites The cost of reducing emissions ($ tC1) at each site is displayed in figure 4Reforestation has no forest area because the target area for forest restoration is where forest has been cleared and no variance becauseof the flat rate of carbon accrual used

REDD+ strategy (a) Timber (b) Palm oil (c) RIL (d) Protected areas (e) Reforestation

Number of sites 429 1845 557 289 2214

Total area (ha) 8586 711 15 200 084 29 575 904 18 425 301 5002 200

Total forested area (ha) 2053 338 3003 896 17 775 332 13 831 004 mdash

Average forest area (ha) 8181 3530 33 922 62 584 mdash

Total cost (US$ millions) 8978 18 028 14 791 7306 8717

Total carbon emissions (tC millions) 831 836 638 414 965

Mean carbon benefit (tCmiddotha1) including peat 308 234 35 54 193

Mean cost (and range) of reducing emissions (USmiddottC1)5636 7314 2377 3927 903

(5ndash972) (6ndash8272) (21ndash30) (2ndash1725) 903ndash903

150

100

150

0Oil

palmProtected

areasRIL Reforestation Timber

Strategy

Cos

t of r

educ

ing

emis

sion

s ($

tC-1

)

Figure 3 The median cost of reducing emissions for eachstrategy (US$middottC1) as shown by the horizontal line in thebox The inter-quartile range shown by the middle boxrepresents 50 of the estimates The whiskers represent thefull range of estimates excluding outliers


5

We found that different REDD+ strategies areeffective at varying budgets and emissions reductiontargets (figures 5(a) and (b)) and that a combination ofstrategies should be employed to reduce emissionscost-effectively across Indonesia For example toachieve a low emissions reduction target of 25(920 MtC) through REDD+ funding should beallocated between PAs timber and oil palm con-cessions which incurs a total cost of $51 billion(table 2) The least costly approach to reduce 50 offorest carbon emissions (1842 MtC) includes thesethree strategies as well as reforesting degraded whichincurs a combined cost of $129 billion A reduction of75 of emissions (2746 MtC) can be achieved at atotal cost of $257 billion by employing a combinationof all strategies targeting deforestation within timberand oil palm concessions investing in better managedPAs employing RIL techniques in logging concessionsand by promoting reforestation Reducing 100 ofemissions from these strategies (3684MtC) costs $578billion The findings of the spatial-targeting approachshow that even the strategies that were most expensiveon average (limiting oil palm and timber expansioninto forests) provided some of the cheapest locationsfor reducing emissions while the cheapest strategieson average (reforestation and RIL) were not ascompetitive for meeting low emissions targets (ie hadfew very low-cost opportunities)

The results from the sensitivity analysis showedthat using surrogate forest cover and carbon datasetsor both combined caused quantitative variances in theproportion of strategies employed to meet emissions

reduction targets but did not change which strategieswere employed (appendix S1 table S3) Using asurrogate forest cover map resulted in the average costof reducing emissions to increase for PAs and oil palmand timber concessions (appendix S1 table S4)however using a surrogate carbon map caused the costof reducing emissions to decrease for all strategiesexcept for reforestation which did not change or RILwhich did not change by more than $1middottC1

4 Discussion

This paper reports on the cost-effective allocation ofREDD+ resources in Indonesia using a spatially-targeted approach The maximum potential carbonbenefit of applying five REDD+ strategies at allpotential locations is 123 MtCmiddotyr1 This is 17 morethan the 105 MtCmiddotyr1 estimated from deforestationfor 2000ndash2005 reported by Harris et al (2012)however our approach differed by accounting forcarbon losses from degradation (logging) and carbongains from reforestation and replacement vegetation(where cleared forests were expected to be replaced byother crops) The prevention of emissions of this scalewould involve employing RIL techniques at all loggingconcessions stopping further deforestation within allPAs and oil palm and timber permits and reforestingall degraded land that has been cleared of forest butwas not listed as lsquonon-forest estatersquo Clearly this is ahighly ambitious scenario and unlikely to beimplemented in the near term A more realistic

(A) Timber concessions (B) Oil palm concessions

(C) Logging concessions (D) Protected areas

Cost per tC reduced

0 500 1000 2000Kilometrers

lt 25

25 -

50

50 -

75

75 -

100

gt 10

0Figure 4 The cost of reducing carbon emissions within (a) timber concessions (b) oil palm concessions (c) logging concessions and(d) protected areas in Indonesia Costs are per tonne of carbon reduced (US$middottC1) for the forested part of the permit or protectedarea Only permits and protected areas with forest cover are included in these figures To improve visibility the whole permit orprotected area has been displayed on the map regardless of where the remaining forest exists Reforestation sites are not shown here asthey have a fixed cost for all areas


6

emissions reduction target for Indonesia in the rangeof 25ndash50 would reduce 920ndash1842 MtC respec-tively over 30 years (31ndash61 MtCmiddotyr1) Whencompared to the average cost estimates from Grahamet al (2016) that did not consider spatial heterogeneitythe inclusion of spatially-discrete cost-benefit esti-mates caused large changes in the average cost ofreducing emissions for the timber oil palm and PAstrategies This is partly because for these threestrategies carbon stored in natural forests is lost whencleared and converted to agriculture whereas the RIL

strategy assesses the proportional carbon benefit fromreduced degradation and the reforestation strategyuses a flat rate of carbon accrual Our results highlightthat at lower emissions targets it is crucial to choosethe most cost-effective strategies in the most cost-effective locations as costs and benefits of REDD+vary spatially in Indonesia

Our spatial analysis revealed that because of thevariability in cost-effectiveness low-cost opportunitiesexist for all of the strategies we reviewed depending onemissions target and budget To reduce the first 25 of

(A)

(B)

50000

40000

30000

20000

10000

0

3500

3000

2500

1500

2000

1000

500

025 50 75 100

Emissions reduction target ()

Em

issi

ons

redu

ced

(tC m

illio

ns)

Cos

t (U

S m

illio

ns)

Timber Oil palm RIL Protected areas Reforestation

Figure 5 Accumulation curves showing the proportion of each REDD+ strategy employed to reduce emissions at the lowest cost Thex-axis represents the emissions reduction target and the y-axis represents (a) the cumulative cost (in US$ millions) and (b) thecumulative emissions reduced (tC millions) Strategies are prioritized by the cost of reducing one tonne of carbon from lowest tohighest Dashed lines display (a) the costs of achieving two emissions reduction targets and (b) the carbon emissions reduced Forexample spending $12 942 million will reduce 1842 MtC (50 of emissions) and spending $25 678 million will reduce 2746 MtC(75 of emissions) RIL = reduced-impact logging

Table 2 The cost (US$ millions) of reducing 25 50 75 and 100 of carbon emissions from five REDD+ strategies The mix ofstrategies that contributes to achieving the emissions target is prioritized by the cost of reducing one tonne of carbon at each site(concession protected area or reforestation site) from lowest to highest

Emissions reduction target

25 50 75 100

Cost (millions) of achieving emissions reduction targets

(a) Timber 3377 4137 4913 8978

(b) Oil palm 1219 2462 4757 18 028

(c) RIL mdash mdash 4917 14 791

(d) Protected areas 467 763 2374 7306

(e) Reforestation mdash 5579 8717 8717

(f) All strategies 5063 12 942 25 678 57 820

Average cost per tonne of avoided emissions ($ tC1) 550 703 935 1570


7

emissions through REDD+ only three strategiesoffered very low-cost opportunitiesmdashreducing defor-estation from oil palm timber and PAs A factordriving this result is that sim82 of oil palm permitshave been granted on land with partial suitability and3 on land that has no agricultural potential for oilpalm mostly in Borneo resulting in costs that areseven times cheaper than sites with high potential ForPAs priority areas for REDD+ projects are spreadacross all major Indonesian islands and are drivenby high deforestation rates coupled with densecarbon stores A significant opportunity for carbonmitigation and biodiversity conservation lies in abatingthe high level of illegal forest loss (Spracklen et al 2015)and the carbon emissions predicted to occur in thefuture (414MtC) if the current pace (sim2pa) of illegaldeforestation in Indonesia continues Within individ-ualPAs the allocationof resources shouldbeprioritizedby accessibility factors as some areas within parks areprotected lsquode factorsquodue to inaccessibility while lowlandforests that are close to roads or urban areas are exposedto greater risk of forest conversion (Gaveau et al 2009Laurance et al 2012)

At a 50 emissions target reforesting degradedland becomes the most important strategy alongsidelowering forest carbon loss in PAs and oil palm andtimber concessions Employing RIL in loggingconcessions is not cost-effective until targeting a75 emissions reduction Although some strategiesare more expensive on average (eg limiting timberand oil palm expansion) these strategies are still veryimportant for achieving even the lowest of emissionsreduction targets (25ndash50) through REDD+ whenspatially-explicit costs and benefits are consideredConversely some strategies with low average costs (egreforestation and RIL) are less important for meetinglow emissions targets highlighting the importance ofspatial-targeting when prioritizing the allocation ofREDD+ resources

The most widespread spatial pattern observed inour analysis was the importance of protecting forestson lowland peat swamps which cover peat deposits ofup to 10 metres in depth (Page et al 1999) Peatlands inBorneo have been declining by 29 pa and by 46pa in Sumatra over the last two decades (Miettinenet al 2012a) presenting an increased challenge forIndonesia to meet their climate mitigation targets asonce cleared peatlands are highly fire-prone (IPCC2007) and their emissions have contributed substan-tially to the high level of national emissions (Bacciniet al 2012) Approximately 21 of PAs 40 of timberpermits and 31 of oil palm permits with remnantforest cover in Sumatra Borneo and Papua occur onpeatlands (mainly in eastern Sumatra and southernBorneo) representing high priority areas for forestprotection through REDD+ In terms of size peatforests account for 9 of forested area in PAs 26 offorested area in oil palm concessions and 62 offorested area in timber plantations

Our paper has focused on carbon and financialelements of REDD+ however other social andecological dimensions of these strategies are alsoimportant determinants of which strategies shouldbe employed and where While scholars are debatingthe non-carbon benefits and risks little attentionhas been directed to how the outcomes vary betweenproject type For example projects that focus onavoided deforestation have the greatest opportunityfor delivering biodiversity co-benefits (Stickler et al2009) Conversely projects tacking illegal deforesta-tion in PAs have high social risks to forest-dependentcommunities whereby communities can be displacedor deprived of access to livelihood resources (Brock-ington et al 2006) yet they can also create employmentopportunities for communities associated with imple-mentation (Mustalahti et al 2012) and can lead toenhancements in ecosystem service function (Mullan2014) Biodiversity benefits from reforestation can belarge where regrowth is promoted on degraded forestbut one of the most serious risks to biodiversity isafforestation which could lead to carbon-richplantation forests being valued over biodiverse low-carbon grasslands (Veldman et al 2015) Loggingconcessions provide a significant opportunity toachieve biodiversity benefits in tropical Asia (Fisheret al 2011b Gaveau et al 2013 Abood et al 2014)because they contain more forests (sim178 Mha 21)than PAs (sim138 Mha 16) in Indonesia and areadvocated alongside PAs for their role in biodiversityconservation (Fisher et al 2011b Gaveau et al 2013)For example concessions that operate well-managedRIL policies and protect forests from agriculturalencroachment can maintain a comparable amount offorest cover as PAs (Putz et al 2012 Gaveau et al 2013)Also approximately 76 of carbon and 85ndash100 ofspecies of mammals birds invertebrates and plants areretained in once-logged forests (Edwards et al 2010Fisher et al 2011b Putz et al 2012) Directing REDD+finance towards logging operations could assist theindustry to expand RIL practices and achieve theseenvironmental benefits

While reducing greenhouse gas emissions cost-effectively was the original motivation for REDD+it is widely agreed that projects need to achievebroader social and environmental objectives such asenhancing the livelihoods of local people andconserving biodiversity (Vijge et al 2016) Theseare referred to as lsquonon-carbon outcomesrsquo (Agrawalet al 2011) The majority of projects in Indonesia areimplemented in highly biodiverse areas and show noconsistent spatial correlation with carbon stocks(Murray et al 2015) demonstrating that factorsother than carbon are driving REDD+ projectimplementation Although they are clearly impor-tant outcomes most nations are yet to developcapacities for monitoring non-carbon outcomes(Vijge et al 2016) though they should be considerednonetheless


8

This analysis could be enhanced with the additionof spatial information on the potential rate of carbonaccrual during forest regeneration Remote sensingforest cover data can confound natural forest withforest plantations resulting in overestimating forestedareas (Sexton et al 2016) To address this issue weimposed a minimum carbon requirement on forestcover which is an accepted approach to reduceambiguity in global forest classification (Sexton et al2016) In the supporting information (appendix S1)we discuss these issues and disclose the carbonthreshold applied for each strategy In this study wedid not assess emissions from the 57 of remainingforest cover that occurs outside of PAs or loggingtimber and oil palm concession areas Roughly 55 ofdeforestation in Indonesia is estimated to occuroutside concession areas driven by logging oil palmsmallholder agriculture rubber coffee mining urbandevelopment and fire (Abood et al 2014 Stibig et al2014) This analysis did not incorporate fluctuationsin opportunity costs in response to supply anddemand conditionsmdashan effect picked up in dynamicmodels (Wertz-Kanounnikoff 2008 Lu and Liu 2015)For example limiting production at an oil palmconcession that could have been profitable canincrease the opportunity costs at another location asdecreased land supply causes costs to rise Ourmeasurements do not include the recovery state offorest carbon stocks following deforestation anddegradation for rotational farming in PAs becausespatial data on the proportional area of rotationalfarming as well as the state of recovery is not availablefor all of Indonesia Future research should investigatespatial patterns of deforestation in Indonesian PAs andrates of carbon accrual in forest regrowth as thisinformation will more accurately inform spatial-targeting of REDD+ finance

By substituting the primary forest layer withsurrogate data we found the average cost of reducingemissions was much higher for timber and oil palmconcessions and PAs because the secondary forest mapconfounds plantation forests with natural forestscausing the projected carbon emissions to decreaseand the cost of reducing emissions to increase There aretwo reasons for this First natural forests that are clearedand replaced with plantation forests may be stillclassified as forests in this map and therefore the carbonemissions resulting from this type of deforestation maynot be included Second plantation forests with higherthan average carbon levels could bemistaken for naturalforests which would drag down the average carbonstored in natural forests at that site

5 Conclusions

The optimal allocation of REDD+ resources shouldconsider the spatial heterogeneity of landscapes anduse this information to apply spatially-targeted

strategies (Venter et al 2012) Our analysis demon-strates that when fine-scale variation in costs andcarbon benefits is considered there is no single-strategy for curbing future forest carbon loss cost-effectively at all potential REDD+ locations Ratheradopting a spatially-targeted approach to resourceallocation reduces carbon emissions most effectivelyThis approach involves identifying the cheapestlocations for reducing carbon emissions for eachREDD+ strategy and targeting these as priority areasfor investment Across Indonesia avoiding additionaldeforestation on peat soils and minimizing forestdegradation caused during log-harvesting (by employ-ing RIL) are highly cost-effective opportunities forreducing emissions This type of spatial analysis marksa crucial step forward in multi-disciplinary land-useplanning in Indonesia The outcomes of our analysiscan guide the implementation of national and regionalplans towards priority areas for combatting forestcarbon loss cost-effectively through REDD+

Acknowledgments

This research was supported by an AustralianPostgraduate Award James Cook University and aSkyrail Rainforest Foundation grant to VG TheAustralian Research Council supported SL with aFuture Fellowship and OV with a Discovery Grant andDECRA Fellowship We would like to thank the twoanonymous reviewers for their constructive feedbackon the draft manuscript Finally the authors wouldlike to thank Sassan Saatchi of the California Instituteof Technology for providing access to the national mapof forest carbon density for Indonesia (Saatchi et al2011) which is also available at httpcarbonjplnasagov

References

Abood S A Lee J S H Burivalova Z Garcia-Ulloa J and Koh L P2014 Relative contributions of the logging fiber oil palmand mining industries to forest loss in Indonesia ConservLett 8 58ndash67

Achard F et al 2014 Determination of tropical deforestation ratesand related carbon losses from 1990 to 2010 Glob ChangeBiol 20 2540ndash54

Agrawal A Nepstad D and Chhatre A 2011 Reducing emissionsfrom deforestation and forest degradation Annu Rev EnvResour 36 373ndash96

Alexander S et al 2011 Opportunities and challenges forecological restoration within REDD+ Restor Ecol 19683ndash9

Baccini A et al 2012 Estimated carbon dioxide emissions fromtropical deforestation improved by carbon-density maps(Paper and GeoLayer) Nat Clim Change 2 182ndash5

Brockington D Igoe J and Schmidt-Soltau K 2006 Conservationhuman rights and poverty reduction Conserv Biol 20250ndash2

Butler R A Koh L P and Ghazoul J 2009 REDD in the red palmoil could undermine carbon payment schemes ConservLett 2 67ndash73


9

den Besten J W Arts B and Verkooijen P 2014 The evolution ofREDD+ an analysis of discursive-institutional dynamicsEnviron Sci Policy 35 40ndash8

Edwards D P Larsen T H Docherty T D S Ansell F A Hsu WW Derheacute M A Hamer K C and Wilcove D S 2010Degraded lands worth protecting the biologicalimportance of Southeast Asiarsquos repeatedly logged forestsProc R Soc B 278 82ndash90

FAO 2012 Crop Suitability Index (Class) for Intermediate InputLevel Rain-fed Oil Palm (GeoLayer) (Rome Italy GAEZFood and Agriculture Organisation of the United Nations)

FAO 2014 Towards climate-responsible peatlands managementMitigation of Climate Change in Agriculture Series 9 ed RBiancalani and A Avagyan (Rome Food and AgicultureOrganization of the United Nations)

Fisher B Edwards D P Giam X and Wilcove D S 2011a Thehigh costs of conserving Southeast Asiarsquos lowlandrainforests Front Ecol Environ 9 329ndash34

Fisher B Edwards D P Larsen T H Ansell F A Hsu W WRoberts C S and Wilcove D S 2011b Cost-effectiveconservation calculating biodiversity and logging trade-offsin Southeast Asia Conserv Lett 4 443ndash50

Forest Trends Association 2016 REDDX Tracking forest finance(wwwREDDxforest-trendsorg) (Accessed 30 March 2016)

Gaveau D L Kshatriya M Sheil D Sloan S Molidena E WijayaA Wich S Ancrenaz M Hansen M and Broich M 2013Reconciling forest conservation and logging in IndonesianBorneo PLoS One 8 e69887

Gaveau D L Wich S Epting J Juhn D Kanninen M and Leader-Williams N 2009 The future of forests and orangutans(Pongo abelii) in Sumatra predicting impacts of oil palmplantations road construction and mechanisms forreducing carbon emissions from deforestation EnvironRes Lett 4 034013

Gibbs H K Brown S Niles J O and Foley J A 2007 Monitoringand estimating tropical forest carbon stocks makingREDD a reality Environ Res Lett 2 045023

Gingold B A Rosenbarger Y Muliastra I K D Stolle F Sudana IM Manessa M D M Murdimanto A Tiangga S BMadusari C C and Douard P 2012 How to IdentifyDegraded Land for Sustainable Palm Oil in IndonesiaWorking Paper (Washington DC World ResourcesInstitute and Sekala)

Graham V Laurance S G Grech A McGregor A and Venter O2016 A comparative assessment of the financial costs andcarbon benefits of REDD+ strategies in Southeast AsiaEnviron Res Lett 11 114022

Greenpeace 2011 Oil-palm concession map of Indonesia circa2008 (GeoLayer) Data compiled by GreenpeaceInternational from The Ministry of Forestry The NationalLand Agency The Agricultural Agency and RegionalPlanning Agencies pp 17ndash18 (wwwgreenpeaceorginternationalGlobalinternationalphotosforests2011appsectionsToyingWithExtinction_Fullpdf)

Hansen M C et al 2013 High-resolution global maps of 21st-century forest cover change Science 342 850ndash3

Harris N L Brown S Hagen S C Saatchi S S Petrova S SalasW Hansen M C Potapov P V and Lotsch A 2012 Baselinemap of carbon emissions from deforestation in tropicalregions Science 336 1573ndash6

Healey J R Price C and Tay J 2000 The cost of carbon retentionby reduced impact logging For Ecol Manag 139237ndash55

Houghton R A 2013 The emissions of carbon from deforestationand degradation in the tropics past trends and futurepotential Carbon Manag 4 539ndash46

Houghton R A Byers B and Nassikas A A 2015 A role fortropical forests in stabilizing atmospheric CO2 Nat ClimChange 5 1022ndash3

Howson P and Kindon S 2015 Analysing access to the localREDD+ benefits of Sungai Lamandau Central KalimantanIndonesia Asia Pac Viewp 56 96ndash110

IPCC 2007 Summary for policymakers Climate Change 2007Impacts Adaptation and Vulnerability Contribution ofWorking Group II to the Fourth Assessment Report of theIntergovernmental Panel on Climate Change ed M L Parry OF Canziani J P Palutikof P J van der Linden and C E Hanson(Cambridge UK Cambridge University Press) pp 7ndash22

Irawan S Tacconi L and Ring I 2011 Stakeholdersrsquo incentives forland-use change and REDD+ the case of Indonesia EcolEcon 87 75ndash83

IUCN and UNEP-WCMC 2016 The World Database on ProtectedAreas (WDPA) (GeoLayer) (wwwprotectedplanetnet)(Accessed 15 February 2016)

Koh L P and Ghazoul J 2010 Spatially explicit scenario analysisfor reconciling agricultural expansion forest protectionand carbon conservation in Indonesia Proc Natl Acad SciUSA 107 11140ndash4

Kotowska M M Leuschner C Triadiati T Meriem S and HertelD 2015 Quantifying above- and belowground biomasscarbon loss with forest conversion in tropical lowlands ofSumatra (Indonesia) Glob Change Biol 21 3620ndash34

Laurance W F et al 2012 Averting biodiversity collapse in tropicalforest protected areas Nature 489 290ndash4

Lu H and Liu G 2015 Opportunity costs of carbon emissionsstemming from changes in land use Sustainability 7 3665ndash82

Miettinen J Shi C and Liew S C 2012a Two decades ofdestruction in Southeast Asiarsquos peat swamp forests FrontEcol Environ 10 124ndash8

Miettinen J Shi C Tan W J and Liew S C 2012b 2010 landcover map of insular Southeast Asia in 250 m spatialresolution (GeoLayer) Remote Sens Lett 3 11ndash20

Minnemeyer S Boisrobert L Stolle F Muliastra K D Hansen MArunarwati B Prawijiwuri G Purwanto J and Awaliyan R2009 Interactive Atlas of Indonesiarsquos Forests (CD-ROM)(Washington DC World Resources Institute)

Mullan K 2014 The Value of Forest Ecosystem Services toDeveloping Economies CGD Working Paper (WashingtonDC Center for Global Development)

Murray J P Grenyer R Wunder S Raes N and Jones J P G 2015Spatial patterns of carbon biodiversity deforestationthreat and REDD+ projects in Indonesia Conserv Biol 291434ndash45

Mustalahti I Bolin A Boyd E and Paavola J 2012 Can REDD+Reconcile local priorities and needs with global mitigationbenefits Lessons from Angai forest TanzaniaEcol Soc 17 16

Olson D M et al 2001 Terrestrial ecoregions of the world a newmap of life on Earth Biosci Biotechnol Biochem 51 933ndash8

Page S E Rieley J O Shotyk W and Weiss D 1999Interdependence of peat and vegetation in a tropical peatswamp forest Phil Trans R Soc Lond B 354 1885ndash97

Pagiola S and Bosquet B 2009 Estimating the costs of REDD atthe country level MPRA paper No 18062 Forest CarbonPartnership Facility World Bank

Pan Y et al 2011 A large and persistent carbon sink in theworldrsquos forests Science 333 988ndash93

Pinard M A and Cropper W P 2000 Simulated effects of loggingon carbon storage in dipterocarp Forest J Appl Ecol 37267ndash83

Pinard M A and Putz F E 1996 Retaining forest biomass byreducing logging damage Biotropica 28 278ndash95

Putz F E et al 2012 Sustaining conservation values in selectivelylogged tropical forests the attained and the attainableConserv Lett 5 296ndash303

Putz F E Zuidema P A Pinard M A Boot R G Sayer J A Sheil DSist P Elias and Vanclay J K 2008 Improved tropical forestmanagement for carbon retention PLoS Biol 6 e1166

Ruslandi Venter O and Putz F E 2011 Overestimatingconservation costs in Southeast Asia Front Ecol Environ 9542ndash4

Saatchi S et al 2011 Benchmark map of forest carbon stocks intropical regions across three continents (GeoLayer) ProcNatl Acad Sci USA 108 9899ndash904


10

Sexton J O et al 2016 Conservation policy and the measurementof forests Nat Clim Change 6 192ndash6

Spracklen B D Kalamandeen M Galbraith D Gloor E andSpracklen D V 2015 A global analysis of deforestation inmoist tropical forest protected areas PLoS One 10 e0143886

Stibig H J Achard F Carboni S Rasi R and Miettinen J 2014Change in tropical forest cover of Southeast Asia from1990 to 2010 Biogeosciences 11 247ndash58

Stickler C M Nepstad D C Coe M T McGrath D G RodriguesH O Walker W S Soares-Filho B S and Davidson E A2009 The potential ecological costs and cobenefits ofREDD a critical review and case study from the Amazonregion Glob Change Biol 15 2803ndash24

Stone R 1988 Management of Engineering Projects (LondonMacmillan Education)

Swallow B et al 2007 Opportunities for avoided deforestation withsustainable benefits An interim report by the ASBPartnership for the Tropical Forest Margins NairobiKenya ASB Partnership for the Tropical Forest Margins

van Kooten G C Eagle A J Manley J and Smolak T 2004 Howcostly are carbon offsets A meta-analysis of carbon forestsinks Environ Sci Policy 7 239ndash51

Veldman J W Overbeck G E Negreiros D Mahy G Le Stradic SFernandes G W Durigan G Buisson E Putz F E andBond W J 2015 Where tree planting and forest expansionare bad for biodiversity and ecosystem services Bioscience65 1011ndash8

Venter O and Koh L P 2011 Reducing emissions from deforestationand forest degradation (REDD+) game changer or justanother quick fix Ann NY Acad Sci 1249 137ndash50

Venter O Meijaard E Possingham H Dennis R Sheil D Wich SHovani L and Wilson K 2009 Carbon payments as asafeguard for threatened tropical mammals Conserv Lett 2123ndash9

Venter O Possingham H P Hovani L Dewi S Griscom B PaoliG Wells P and Wilson K A 2012 Using systematicconservation planning to minimize REDD+ conflict withagriculture and logging in the tropics Conserv Lett 6 1ndash9

Vijge M J Brockhaus M Di Gregorio M and Muharrom E 2016Framing national REDD+ benefits monitoringgovernance and finance a comparative analysis of sevencountries Glob Environ Change 39 57ndash68

Wertz-Kanounnikoff S 2008 Estimating the costs of reducing forestemissions a review of methods CIFORWorking Paper


11

provide the economic incentives needed to conserveforests by linking financial rewards to emissionsreduced or carbon sequestered (Agrawal et al 2011)When REDD+ was first conceived in 2005 it sought toReduce Emissions from Deforestation (RED see denBesten et al 2014) at which point it was chieflyconcerned with limiting tropical deforestation Duringearly-stage discussions the scope of REDD+ wasbroadened to include reducing degradation (REDD)as well as conserving and sustainably managing forestsand enhancing forest carbon stocks (REDD+) Thisdevelopment opened up a range of new opportunitiesfor addressing forest carbon loss including activitiesthat sequester carbon such as reforestation and thatreduce degradation such as reduced-impact logging(RIL Putz et al 2008 Alexander et al 2011)

Through its range of strategies REDD+ has thepotential to reduce carbon dioxide (CO2) concen-trations in the atmosphere which will aid in thetransition to a low fossil fuel global economy(Houghton et al 2015) Since its inception REDD+has attracted over US$73 billion in fundingincluding pledges of over US$2 billion to Indonesiaalone (Forest Trends Association 2016) A key issuehindering the implementation of REDD+ is how wellcost-effective climate mitigation activities align withthe rights of local forest users with concerns raisedthat the priorities of international investors will beprivileged over those of local communities (Howsonand Kindon 2015) Additionally economic concernshave been centred on the unlikelihood that REDD+will generate sufficient finance to off-set lost revenuesfrom alternative land-use activities drawing compar-isons against the moderate level of funding directedtowards REDD+ relative to the high profits generatedfrom deforestation-dependent activities such astimber and oil palm production (Venter and Koh2011) The literature shows that projects aimed atlimiting deforestation from large-scale oil palmproduction are expensive due to the high forgonerevenues (ie opportunity cost) from convertingforest into oil palm (eg Butler et al 2009 Venter et al2009 Fisher et al 2011a Irawan et al 2011 Ruslandiet al 2011)

An alternative and potentially cheaper pathway forREDD+ to contribute towards carbonmitigation is viareforestation reducing illegal deforestation in pro-tected areas (PAs) and reducing forest degradationThe optimal approach to allocating REDD+ resourceswill be influenced by the spatial context in which eachproject is applied as costs and carbon benefits can varyspatially (Pagiola and Bosquet 2009) Site-specificfactors that influence costs and benefits includeterrain distance to markets and soil type (Gibbset al 2007 Pagiola and Bosquet 2009) Recent studiesundertaken in Indonesia highlight how applying aspatially-targeted approach to regional developmentplans can reduce the trade-offs of agricultural ortimber expansion and forest protection (Koh and

Ghazoul 2010 Venter et al 2012) A key questiontherefore is how does spatial variation influence theeffectiveness of REDD+ strategies to mitigate forest-based carbon emissions at low-cost across Indonesia

To address this question we used spatial analysesto assess the variation in costs and carbon benefits ofvarious REDD+ strategies in Indonesia and identifiedthe factors that drive cost-effectiveness We used mapsof carbon stocks forest cover peatlands and cropsuitability to estimate the potential for REDD+ to slowor reverse carbon emissions from oil palm timber andlogging permits PAs and on degraded land Weexplored the cost-effectiveness of REDD+ strategiesfor reducing one tonne of carbon and for achieving arange of emissions targets We compared the resultsfrom this spatial analysis to estimates from a cost-benefit analysis of REDD+ that used average costs andbenefits (Graham et al 2016) This paper is designed todeliver fine-scale information to policy makers onspatially-targeted opportunities for mitigating carbonemissions from deforestation and forest degradationin Indonesia

2 Materials and methods

In this paper we estimated the 30 yr carbon emissionsand financial costs from anticipated land conversionand determined the carbon sequestered from restoringland that is not slated for urban development oragriculture across Indonesia The term lsquopermitsrsquo refersto land use rights issued to companies for logging oilpalm or timber concessions We ranked all permitsPAs and reforestation sites by the cost of reducing onetonne of carbon (from low to high) to determine thecombination of strategies that achieve emissionstargets (25 50 75 100) most cost-effectivelyAll carbon values are in tonnes (1 tonne frac14 1 Mg) ofcarbon (C) Carbon dioxide (CO2) was converted tocarbon by dividing by 367 (van Kooten et al 2004)Biomass was converted to carbon by multiplying by0492 (Pinard and Putz 1996) All financial figures arein 2010 US dollars Here we present a summarizedversion of the steps involved in calculating the spatiallyexplicit emissions and costs individually for eachpermit PA or reforestation site See appendix S1(available at httpsstacksioporgERL12044017mmedia) in supporting information for details onthe input data and detailed methods

21 Estimating carbon benefits of REDD+ strategiesSpatial analysis was performed in ArcGIS v103 (ESRI2014) We used 250 m spatial resolution land covermaps for 2000 and 2010 that were produced usingModerate Resolution Imaging Spectroradiometer(MODIS) images and Daichi-Advanced Land Observ-ing Satellite data (Miettinen et al 2012b) We createdbinary maps of 2000 and 2010 natural forest cover(figure 1(a)) by classifying mangrove forest peat


2






(A) (B)

(C) (D)

Forest cover


Forest loss

80004000100

N




3






(A) (B)

(C) (D)

Timber concessions

Logging concessions


Protected areas

N




4




3 Results







Number of sites 429 1845 557 289 2214

Total area (ha) 8586 711 15 200 084 29 575 904 18 425 301 5002 200








150

100

150

0Oil

palmProtected


Strategy

Cos

t of r

educ

ing

emis

sion

s ($

tC-1

)



5




4 Discussion




Cost per tC reduced


lt 25

25 -

50

50 -

75

75 -

100

gt 10



6




(A)

(B)

50000

40000

30000

20000

10000

0

3500

3000

2500

1500

2000

1000

500

025 50 75 100


Em

issi

ons

redu

ced

(tC m

illio

ns)

Cos

t (U

S m

illio

ns)





25 50 75 100


(a) Timber 3377 4137 4913 8978

(b) Oil palm 1219 2462 4757 18 028




(f) All strategies 5063 12 942 25 678 57 820



7







8



5 Conclusions



Acknowledgments


References









9












































10















11






(A) (B)

(C) (D)

Forest cover


Forest loss

80004000100

N




3






(A) (B)

(C) (D)

Timber concessions

Logging concessions


Protected areas

N




4




3 Results







Number of sites 429 1845 557 289 2214

Total area (ha) 8586 711 15 200 084 29 575 904 18 425 301 5002 200








150

100

150

0Oil

palmProtected


Strategy

Cos

t of r

educ

ing

emis

sion

s ($

tC-1

)



5




4 Discussion




Cost per tC reduced


lt 25

25 -

50

50 -

75

75 -

100

gt 10



6




(A)

(B)

50000

40000

30000

20000

10000

0

3500

3000

2500

1500

2000

1000

500

025 50 75 100


Em

issi

ons

redu

ced

(tC m

illio

ns)

Cos

t (U

S m

illio

ns)





25 50 75 100


(a) Timber 3377 4137 4913 8978

(b) Oil palm 1219 2462 4757 18 028




(f) All strategies 5063 12 942 25 678 57 820



7







8



5 Conclusions



Acknowledgments


References









9












































10















11






(A) (B)

(C) (D)

Timber concessions

Logging concessions


Protected areas

N




4




3 Results







Number of sites 429 1845 557 289 2214

Total area (ha) 8586 711 15 200 084 29 575 904 18 425 301 5002 200








150

100

150

0Oil

palmProtected


Strategy

Cos

t of r

educ

ing

emis

sion

s ($

tC-1

)



5




4 Discussion




Cost per tC reduced


lt 25

25 -

50

50 -

75

75 -

100

gt 10



6




(A)

(B)

50000

40000

30000

20000

10000

0

3500

3000

2500

1500

2000

1000

500

025 50 75 100


Em

issi

ons

redu

ced

(tC m

illio

ns)

Cos

t (U

S m

illio

ns)





25 50 75 100


(a) Timber 3377 4137 4913 8978

(b) Oil palm 1219 2462 4757 18 028




(f) All strategies 5063 12 942 25 678 57 820



7







8



5 Conclusions



Acknowledgments


References









9












































10















11




3 Results







Number of sites 429 1845 557 289 2214

Total area (ha) 8586 711 15 200 084 29 575 904 18 425 301 5002 200








150

100

150

0Oil

palmProtected


Strategy

Cos

t of r

educ

ing

emis

sion

s ($

tC-1

)



5




4 Discussion




Cost per tC reduced


lt 25

25 -

50

50 -

75

75 -

100

gt 10



6




(A)

(B)

50000

40000

30000

20000

10000

0

3500

3000

2500

1500

2000

1000

500

025 50 75 100


Em

issi

ons

redu

ced

(tC m

illio

ns)

Cos

t (U

S m

illio

ns)





25 50 75 100


(a) Timber 3377 4137 4913 8978

(b) Oil palm 1219 2462 4757 18 028




(f) All strategies 5063 12 942 25 678 57 820



7







8



5 Conclusions



Acknowledgments


References









9












































10















11




4 Discussion




Cost per tC reduced


lt 25

25 -

50

50 -

75

75 -

100

gt 10



6




(A)

(B)

50000

40000

30000

20000

10000

0

3500

3000

2500

1500

2000

1000

500

025 50 75 100


Em

issi

ons

redu

ced

(tC m

illio

ns)

Cos

t (U

S m

illio

ns)





25 50 75 100


(a) Timber 3377 4137 4913 8978

(b) Oil palm 1219 2462 4757 18 028




(f) All strategies 5063 12 942 25 678 57 820



7







8



5 Conclusions



Acknowledgments


References









9












































10















11




(A)

(B)

50000

40000

30000

20000

10000

0

3500

3000

2500

1500

2000

1000

500

025 50 75 100


Em

issi

ons

redu

ced

(tC m

illio

ns)

Cos

t (U

S m

illio

ns)





25 50 75 100


(a) Timber 3377 4137 4913 8978

(b) Oil palm 1219 2462 4757 18 028




(f) All strategies 5063 12 942 25 678 57 820



7







8



5 Conclusions



Acknowledgments


References









9












































10















11







8



5 Conclusions



Acknowledgments


References









9












































10















11



5 Conclusions



Acknowledgments


References









9












































10















11












































10















11















11

LETTER Spatially explicit estimates of forest carbon ... · swamp forest, lowland forest, lower...

Documents

Transcript of LETTER Spatially explicit estimates of forest carbon ... · swamp forest, lowland forest, lower...