Carbon Balance and Management

Carbon Balance and Management

Research

BioMed Central

Open Access

Predicting the deforestation-trend under different carbon-prices

GeorgEKindermann*1,2, MichaelObersteiner1,3, EwaldRametsteiner1,2 and

IanMcCallum1

Address: 1International Institute for Applied Systems Analysis (IIASA), Laxenburg, Austria, 2University of Natural Resources and Applied Life Sciences (BOKU), Vienna, Austria and 3Institute for Advanced Studies (IHS), Vienna, Austria

Email: GeorgEKindermann*-kinder@iiasa.ac.at; MichaelObersteiner-oberstei@iiasa.ac.at; EwaldRametsteiner-ramet@iiasa.ac.at; IanMcCallum-mccallum@iiasa.ac.at* Corresponding author

Published: 06 December 2006

Carbon Balance and Management 2006, 1:15

doi:10.1186/1750-0680-1-15

This article is available from: /content/1/1/15

Received: 13 October 2006Accepted: 06 December 2006

© 2006 Kindermann et al; licensee BioMed Central Ltd.

which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Abstract

Background: Global carbon stocks in forest biomass are decreasing by 1.1 Gt of carbon annually,owing to continued deforestation and forest degradation. Deforestation emissions are partly offsetby forest expansion and increases in growing stock primarily in the extra-tropical north. Innovativefinancial mechanisms would be required to help reducing deforestation. Using a spatially explicitintegrated biophysical and socio-economic land use model we estimated the impact of carbon priceincentive schemes and payment modalities on deforestation. One payment modality is adding costsfor carbon emission, the other is to pay incentives for keeping the forest carbon stock intact.Results: Baseline scenario calculations show that close to 200 mil ha or around 5% of todays forestarea will be lost between 2006 and 2025, resulting in a release of additional 17.5 GtC. Today'sforest cover will shrink by around 500 million hectares, which is 1/8 of the current forest cover,within the next 100 years. The accumulated carbon release during the next 100 years amounts to45 GtC, which is 15% of the total carbon stored in forests today. Incentives of 6 US$/tC forvulnerable standing biomass payed every 5 year will bring deforestation down by 50%. This willcause costs of 34 billion US$/year. On the other hand a carbon tax of 12 $/tC harvested forestbiomass will also cut deforestation by half. The tax income will, if enforced, decrease from 6 billionUS$ in 2005 to 4.3 billion US$ in 2025 and 0.7 billion US$ in 2100 due to decreasing deforestationspeed.

Conclusion: Avoiding deforestation requires financial mechanisms that make retention of forestseconomically competitive with the currently often preferred option to seek profits from other landuses. Incentive payments need to be at a very high level to be effective against deforestation. Taxeson the other hand will extract budgetary revenues from the regions which are already poor. Acombination of incentives and taxes could turn out to be a viable solution for this problem.Increasing the value of forest land and thereby make it less easily prone to deforestation would actas a strong incentive to increase productivity of agricultural and fuelwood production, which couldbe supported by revenues generated by the deforestation tax.

Background

Deforestation is considered the second largest source ofgreenhouse gas (GHG) emissions amounting to an esti-mated 2 gigatonnes of carbon (GtC) per annum over thelast decade [1]. It is a persistent problem. The UN Foodand Agriculture Organization, in its recently released mostcomprehensive assessment of forests ever, puts deforesta-tion at about 12.9 mil. ha per year [2]. At the same time,forest planting, landscape restoration and natural expan-sion of forests reduce the net loss of forest area. Netchange in forest area in the period 2000–2005 is esti-mated at -7.3 million hectares per year [2]. This reducesthe annual GHG emissions to an estimated 1.1 GtC. Incomparison, 7.3 GtC were emitted in 2003 by using fossilenergy sources [3].

Deforestation has been difficult to tackle by governments,as its drivers are complex and many land uses yield higherrevenues than those from forested land. Some see climatepolicy as a new opportunity to effectively reduce a majorsource of greenhouse gases and biodiversity loss as well asto increase incomes of many people in rural areas whoselivelihood depends on forests. The implementation ofmeasures avoiding deforestation would require innova-tive financial mechanisms in the context of global climatepolicies. In this paper we study the potential magnitude ofeffects of different financial mechanisms to help reducedeforestation, using a modeling approach.

To estimate the impact of financial incentives, to reducedeforestation and assuming profit maximizing behavior,we calculate differences in net present value of differentland uses using a spatially explicit integrated biophysicaland socio-economic land use model. Key model parame-ters, such as agricultural land use and production, popu-lation growth, deforestation and forest productconsumption rates were calibrated against historical nd use changes are simulated in the model as a decisionbased on a difference between net present value of incomefrom production on agricultural land versus net presentvalue of income from forest products. Assuming fixedtechnology, the model calculates for each 0.5° grid cellthe net present value difference between agricultural andforest land-uses in one-year time steps. When carbon mar-ket prices, transferred through a financial mechanism, bal-ance out differences between the net present value ofagricultural land and forest-related income, it is assumed,consistent with profit maximising behavior, that deforest-ation is avoided.

The net present value difference of forest versus other landuses can be balanced out through two mechanisms. Oneis to reduce the difference by adding costs to conversionthrough taxing emissions from deforestation, e. g.through a land clearance tax and wood sales taxes. The

other is to enhance the value of the existing forest byfinancial support when keeping the forest carbon stock, tobe paid in certain time intervals. In both cases the value offorest carbon stock would be pegged to carbon marketprices. The modeling results for different hypothetical taxor subsidy levels show the potential magnitude of avoideddeforestation through financial incentive or disincentivemechanisms. The model results are annual, spatiallyexplicit estimates of the forest area and biomass develop-ment from 2000 to 2100, with particular focus on theperiod 2006 to 2025.

Results and discussion

Baseline deforestation 2000–2100 and effects of financial mechanisms aiming at cutting emissions in half

Baseline scenario calculations (i.e. a carbon price of 0US$/tC is assumed) show that close to 200 mil ha oraround 5% of todays forest area will be lost between 2006and 2025, resulting in a release of additional 17.5 GtC tothe atmospheric carbon pool. The baseline deforestationspeed is decreasing over time, which is caused by adecreasing forest area in regions with hight deforestationpressure. In the year 2025 the annual deforested areadecreases to 8.2 million hectares, compared to 12.9 mil-lion hectares in 2005. By the year 2100 deforestation ratesdecline to some 1.1 million hectares. According to thebase line scenario, today's forest cover will shrink byaround 500 million hectares or by more than 1/8 withinthe next 100 years (figure 1).

Carbon emissions from deforestation in 2005 is 1.1 GtC/year and decreases to 0.68 GtC/year in 2025 and further

Figure 1

Deforested Area until 2100. Deforested Area under alternative assumptions. Incentives... Periodic payments for standing biomass, Tax... Payments for harvesting wood, Burn... felled wood is burned immediately, Sell... harvested wood is soled, Burn/Sell... share of the wood will be burned the other part soled.

to 0.09 GtC/year in 2100. The accumulated carbon releaseduring the next 100 years amounts to 45 GtC which is15% of the total carbon stored in forests today. To bringdeforestation down by 50%, incentives of 6 US$/tC/5 yearor a land clearance tax of between 9 US$/tC and 25 US$/tC would be necessary, depending whether the harvestedwood is burned on the spot (e. g. slash-and-burn agricul-ture) or sold. In the latter case, a higher carbon tax of upto 25 US$/tC is necessary to effectively reduce incentivesto deforest, to a degree that cuts overall global deforesta-tion by 50%. If the wood is further used and convertedinto products, only 18% of the biomass could be saved bya carbon price of 9 US$/tC, caused by the compensatingeffect of an income by selling wood and a longer time-period for releasing carbon. On the other hand, if the car-bon price is 25$/tC and the wood is assumed to be slashburned, the reduction of deforestation calculated to be91% (figure 1 and 2). On a first sight it seems, that incen-tive payments might be more effective, than taxation.However, incentives payment contracts have to berenewed every 5 year for the actual standing biomass andthe change of biomass has to be known to detect a breachof the contract, while a deforestation tax will be payedonce for the harvested biomass once detected by targetedearth observation systems (see figure 3 and 4). In the lat-ter, transactions costs for implementing avoided deforest-ation are small.

The assumption, that either only slash burn or all woodwill be sold is unrealistic. Thus, a scenario where LatinAmerica has 90% slash burn and 10% selling, Africa 50%

Figure 2

Released Carbon from Deforestation until 2100. Released carbon from deforestation under alternative

assumptions. Incentives... Periodic payments for standing bio-mass, Tax... Payments for harvesting wood, Burn... felled wood is burned immediately, Sell... harvested wood is soled, Burn/Sell... share of the wood will be burned the other part soled.

Figure 3Avoided Carbon releases under different Carbon prices during the next 100 years. Incentives... Periodic payments for standing biomass, Tax... Payments for harvesting wood, Burn... felled wood is burned immediately, Sell... har-vested wood is soled, Burn/Sell... share of the wood will be burned the other part soled.

slash burned and 50% selling and in the remaining area10% slash burned and 90% selling, was examined. Undersuch scenario assumptions a carbon tax of 12 $/tC will cutdeforestation in half. Also the assumption, that a carbonprice will stay constant over time may not be close to real-ity but it can be used to see the long-term influence of agiven carbon price.

We differentiate between the following cases:Baseline: Introducing no carbon price.

Incentives: Introducing a carbon price which will bepayed periodic for the carbon stored in the standing forestbiomass.

All: Payments are done, without considering the effec-tiveness of the payment, in all regions.

Region: Payments are done in regions where the pay-ments protect forest against deforestation.

Affected: Payments are done for forests where the pay-ments protect them against deforestation.

Tax: Introducing a carbon price which has to be paid forreleasing the stored carbon to the atmosphere.

Burn: All wood will be burned immediately.Sell: All harvested wood will be sold.

Figure 4Saved Forest Area under different Carbon prices during the next 100 years. Incentives... Periodic payments for standing biomass, Tax... Payments for harvesting wood, Burn... felled wood is burned immediately, Sell... harvested wood is soled, Burn/Sell... share of the wood will be burned

the other part soled.

Burn/Sell: A share of the wood will be burned and theother part sold.

Costs and revenues under different carbon prices

The effectiveness of introducing a carbon price to influ-ence deforestation decisions depends largely on the levelsset for carbon prices, apart from considerations of politi-cal feasibility and implementability. Low prices have littleimpact on deforestation rates. During the 21st century car-bon tax schemes of 9 US$/tC for slash burn and 25 US$/tC for situations when removed wood enters a harvestedwood products pool (HWP) would generate some 2 to 5.7billion US$/year respectively when emissions from defor-estation are to be cut in half. For the variant of 12 US$/tC,with regionally differentiated slash burn and HWPassumptions, the average annual income for the next 100years are calculated to be around 2.7 billion US$. Thesetax revenues decrease dramatically over time mainly dueto the declining baseline deforestation rate. Tax revenuesare computed to be 6 billion US$ in 2005, 4.3 billion US$in 2025 and 0.7 billion US$ in 2100. This indicates themagnitudes and their temporal change of funds generatedfrom a deforestation tax scheme aiming at a 50% emissionreduction (figure 5 and 7).

In the alternative incentive scheme, the amount of fundsnecessary, is depending on the strategy of payments,either increasing, staying constant or decreasing over time.If incentives are paid only for those forest areas that areabout to be deforested, and with a global target of cuttingdeforestation by 50%, a minimum payment of 6 US$/tC/

Figure 5

Income under different Carbon Prices. Tax... Payments for harvesting wood, Burn... felled wood is burned immedi-ately, Sell... harvested wood is soled, Burn/Sell... share of the wood will be burned the other part soled.

5 year or 0.24 billion US$ in 2006 would be required. Thisamount rises to some 1.2 billion US$ in 2010, 4.1 billionUS$ in 2025 and 10 billion US$ in 2100 caused by theincreasing area of saved forest area. As precise informationof forests about to be deforested is absent, incentive pay-ment schemes would have to focus on regions underdeforestation pressure. Given that incentives are onlyspent on regions of 0.5° × 0.5° where they can effectivelyreduce deforestation in an amount that they will balanceout the income difference between forests and alternativeland use up the 6 US$/tC/5 year, this would come at a costof 34 billion US$/year (figure 6 and 8). It should be notedthat the tax applies only on places currently deforestedwhile the subsidy applies to larger areas depending onhow far it is in practice possible to restrict the subsidy tovulnerable areas. All figures above are intentionally free oftransaction costs. Transaction costs would inter aliainclude expenditure for protecting the forests against ille-gal logging by force and expenditures monitoring smallscale forest degregation. Governance issues such as cor-ruption and risk adjustment, depending on the countryare, however, considered in the analysis to the extent pos-sible.

Regional effects of carbon prices on deforestation

Sources of deforestation in the model are expansion ofagriculture and buildup areas as well as from unsustaina-ble timber harvesting operations impairing sufficientreforestation. Deforestation results from many pressures,both local and international. While the more direct causesare rather well established as being agricultural expansion,infrastructure extension and wood extraction, indirectdrivers of deforestation are made up of a complex web of

Figure 6

Expenditure under different Carbon Prices. Incentives... Periodic payments for standing biomass, All... Payments are done, without considering the effectiveness of the payment, in all regions, Region... Payments are done in regions where the payments protect forest against deforestation, Affected... Payments are done for forests where the payments protect

them against deforestation.

interlinked and place-specific factors. There is large spa-tially differentiated heterogeneity of deforestation pres-sures. Within a forest-agriculture mosaic, forests are underhigh deforestation pressure unless they are on sites whichare less suitable for agriculture (swamp, slope, altitude).Closed forests at the frontier to agriculture land are alsounder a high deforestation pressure while forest beyond

Figure 7

Cash flow until 2100 for different Carbon Prices. Incentives... Periodic payments for standing biomass, Tax... Payments for harvesting wood, Affected... Payments are done for forests where the payments protect them against defor-estation, Burn... felled wood is burned immediately, Sell... har-vested wood is soled, Burn/Sell... share of the wood will be burned the other part soled.

this frontier are under low pressure as long as they arebadly attainable. The model was build to capture suchheterogeneity in deforestation pressures.

Figure 9 shows that the model predicts deforestation tocontinue at the frontier to agricultural land and in areaswhich are easly accessible. Trans-frontier forests are alsopredicted to be deforested due to their relative accessibil-ity and agricultural suitablility. Forests in mosaic landscontinue to be under strong pressure. Figure 10 illustratesthe geography of carbon saved at a carbon tax of 12 US$/tC compared to biomass lost through deforestation.Under this scenario deforestation is maily occurring inclusters, which are sometimes surrounded by forests (e.g.Central Africa) or are concentrated along a line (Amazon).The geography of the remaining deforestation patternindicates that large areas are prevented from deforestationat the frontier by the 12 US$/tC tax. The remaining emis-sions from deforestation are explained mainly by theiraccessibility and favourable agricultural suitability.

Conclusion

Avoiding deforestation requires financial mechanismsthat make retention of forests economically competitivewith the currently often preferred option to seek profitsfrom other land uses. According to the model calcula-tions, even relatively low carbon incentives of around 6 $/tC/5 year, paid for forest carbon stock retention or carbontaxes of 12 $/tC would suffice to effectively cut emissionsfrom deforestation by half. Taxes revenues would bringabout annual income of US$6 bn in 2005 to US$0.7 bn in2100. The financial means required for incentives are esti-mated to range from US$3 bn to US$ 200 bn per year,depending on the design of the avoided deforestation pol-icy. Our scenario, where incentives are payed in regionswhere deforestation will appear and the payment has aneffect, estimates the necessary funds to cut emissions fromdeforestation in half in the magnitude of some US$ 33 bnper year, without including costs for transaction, observa-tion and illegal logging protection. Increasing the value offorest land and thereby make it less easily prone to defor-estation would act as a strong incentive to increase pro-ductivity of agricultural and fuelwood production.

Methods

The model is based mainly on the global afforestationmodel of [4] and calculates the net present value of for-estry with equation (1 – 16) and the net present value ofagriculture with equation (17 – 20). Main drivers for thenet present value of forestry are income from carbonsequestration, wood increment, rotation period length,discount rates, planting costs and wood prices. Main driv-ers for the net present value of agriculture on current forestland are population density, agricultural suitability andrisk adjusted discount rates.

All symbols in the following equations are explained inthe section "Abbreviations".

Net present value of forestry

The net present value of forestry is determined by theplanting costs, the harvestable wood volume, the wood-price and benefits from carbon sequestration.

Figure 8Expenditure until 2100 for different Incentive pay-ment Strategies. Incentives... Periodic payments for stand-ing biomass, All... Payments are done, without considering the effectiveness of the payment, in all regions, Region... Payments are done in regions where the payments protect forest against deforestation, Affected... Payments are done for for-ests where the payments protect them against deforestation.

For existing forests which are assumed to be under activemanagment the net present value of forestry given multi-ple rotations (Fi) over the simulation horizon is calculatedfrom the net present value for one rotation (fi) (equation1). This is calculated by taking into account the plantingcosts (cpi) at the begin of the rotation period and theincome from selling the harvested wood (pwi·Vi) at theend of the rotation period. Also the benefits from carbonsequestration are included denoted as (Bi).

The planting costs (eq. 3) are calculated by multiplyingthe planting costs of the reference country (cpref) with aprice index (pxi) and a factor which describes the share ofnatural regeneration (pri). The ratio of plantation to natu-ral regeneration is assumed to increase with increasingyield for the respective forests (eq. 4). The price index (eq.5) is calculated using the purchasing power parity of therespective countries. The stumpage wood price (eq. 6) iscalculated from the harvest cost free income range ofwood in the reference country. This price is at the lowerbound when the population density is low and the forestshare is high and at the higher bound when the popula-tion density is high and the forest share is low. The priceis also multiplied with a price index converting the pricerange from the reference country to the examined country.The population-density and forest-share was standardized

These two values are compared against each other anddeforestation is subsequently predicted to occur when theagricultural value exceeds the forest value by a certainmargin. When the model comes to the result, that defor-estation occurs, the speed of deforestation was constraintby estimates given by equation (24). The speed of defor-estation is a function of sub-grid forest share, agriculturalsuitability, population density and economic wealth ofthe country.

Figure 9

Removed Biomass without a carbon price. Green areas show grids where nowadays forests can be found. Red areas

indicate grids where deforestation will occur in a scenario without carbon prices.

Figure 10

Saved Biomass by 12$/tC (Burn Sell). Light green show grids where nowadays forests can be found. Dark green areas indicate grids where forest biomass can be saved by introducing a carbon price of 12$/tC compared to the baseline scenario.

Red ares indicate grids where there will still be deforestation.

between 1 and 10 by using equation (7) and equation (8)respectively.

The harvested volume (Vi) is calculated by multiplying themean annual increment (MAIi) with the rotation periodlength (Ri) accounting for harvesting losses (eq. 9).The rotation period length (eq. 10) depends on the yield.Fast growing stands have a short and slow growing sites along rotation length. In this study the rotation length is inthe range between 5 and 140 years.

The mean annual increment (eq. 11) is calculated by mul-tiplying the estimated carbon uptake (ωi) and a transfor-mation factor which brings the carbon weight to a woodvolume (C2Wi). The carbon uptake (ωi) is calculated bymultiplying the net primary production (NPPi) with a fac-tor describing the share of carbon uptake from the net pri-mary production (eq. 12).

The benefits of carbon sequestration (eq. 13) are calcu-lated by discounting the annual income from additionalcarbon sequestration and subtracting the expensesincurred from harvesting operations and silvicultural pro-duction. At the end of a rotation period the harvested car-bon is still stored in harvested wood products and willcome back to atmosphere with a delay. This is consideredin the factor (θi) which shares the harvested wood volumeto short and long living products(eq. 14).

The effective carbon price represents the benefit whichwill directly go to the forest owner. In equation (16) a fac-tor describing the percentage of the transaction cost free

carbon price is used. A factor leaki is calculated as the aver-age of the percentile rank from "political stability", "gov-ernment effectiveness" and "control of corruption" [5].

Fi=fi [1 (1+r) Ri] 1

fi = -cpi + pwi·Vi + Bi (2)cpi = cpref·pri·pxi (3)

(1)

MAIi<3 0

pri= (

MAIi 3)/63≤MAIi≤9

1MAIi>9 (4)

SNFs = 1 + (1 - Fs) * 9 (8)(9)Vi = MAIi·Ri·(1 - HLi)

MAIi = ωi·C2W (11)

ωi = NPPi·CU (12)

Bi=epci ωi (1 bi) {r 1 [1 (1+r) Ri] Ri (1 θi) (1+r) Ri}

(13)

fracslp = 1 - fracllp (15)epci = pci·leaki (16)

Net present value of agriculture

The net present value of agriculture (Ai) is calculated witha two-factor Cobb-Douglas production function (equa-tion 17). It depends on the agriculture suitability and thepopulation density. A high agriculture suitability and ahigh population density causes high agricultural values.The value ranges between a given minimum and a maxi-mum land price. The parameters ai and γi determine therelative importance of the agriculture suitability and thepopulation density and νi determines the price level forland. The agriculture suitability and the population den-sity are normalized between 1 and 10.

Ai=νi SAgSiαi SPdiγ

i

(17)

SAgSi=

10

AgSi≥0.5 1+9 AgS/0.5AgSi<0.5

(18)γi = αi (20)

Decision of deforestation

The deforestation decision is expressed by equation (21).It compares the agricultural and forestry net present valuescorrected by values for deforestation and carbon seques-tration. For the deforestation decision the amount ofremoved biomass from the forest is an important variable.The agricultural value needed for deforestation increaseswith the amount of timber sales and its concomitant flowto the HWP pool. On the other hand the agriculture valuewill be decreased by the amount of released carbon to theatmosphere. This mechanism is expressed by a deforesta-tion value (DVi, eq. 22). The model also allows for com-

pensation of ancillary benefits from forests. This

additional income is modeled either as a periodicalincome or a one time payment and will increase the for-estry value by (IPdiscounted, which has been done in equation (23).

i). If it is a periodic payment it has to be YesAi+DVi>Fi Hi+IPi

Defor=

∧not Protected A(21)

i+DVi≤Fi Hi+IPi

No∨Protected

handled. Two mechanisms are realized in equation (21).One is to pay the forest owner to avert from the deforest-ation, the other is to introduce a carbon price that the for-est owner gets money by storing carbon and paying forreleasing it. The introduction of a carbon price focuses themoney transfer to the regions where a change in biomasstakes place. Payments to avoid emissions from deforesta-tion can be transfered to cover all of the globe's forests,target to large "deforestation regions" or individual grids.

Deforestation rate

Once the principle deforestation decision has been madefor a particular grid cell (i.e. the indicator variable Defori =1) the actual area to be deforested within the respectivegrid is to be determined. This is done by the auxiliaryequation (24 – 25) computing the decrease in forestshare. We model the deforestation rate within a particulargrid as a function of its share of forest cover, agriculturalsuitability, population density and gross domestic prod-uct. The coefficients c1 to c6 were estimated with a general-ized linear model of the quasibinomial family with a logitlink. Values significant at a level of 5% were taken and areshown in table 1. The parameters of the regression modelwere estimated using R [6]. The value of c0 was determinedupon conjecture and directly influences the maximumpossible deforestation rate. For our scenarios the maxi-mum possible deforestation is set to 5% of the total landarea per year. That means, a 0.5° × 0.5° grid coveredtotally with forests can not be deforested in a shorter timeperiod than 20 years.

0Defor=No

Fdeci=

Fsi

Ftdeci>Fsi∧Defor=Yes(24)

Ftdec

iFtdeci≤Fsi∧Defoor=Yes

Table 1: Coefficients for equation (25) – Deforestation speed

CoefEstimateStd. Errorc00.05

-c1-1.799e+004.874e-01c2-2.200e-019.346e-02c3-1.663e-015.154e-02c44.029e-021.712e-02c5-5.305e-041.669e-04c6

-1.282e-04

3.372e-05

Ftdec 0

Fsi=

i=0∨AgSi=0 xi

Fsi>0∧AgSi>0

(25)

The deforestation rates (Ftdec) were taken from [2], wherethe forest area from 1990, 2000 and 2005 for each countrywas given. For the estimation of the model parameters thearea difference between 1990 and 2005 was used to inferthe deforestation rate. All values which showed anincrease of the forest area have been set to 0, because themodel should only predict the deforestation. Countrieswith an increasing forest area have a deforestation rate of0. It should be mentioned that the change rate is based onthe total land area in the grid i and not on the current for-est area.

By using c2/Fs the model can only be used on grid's wherethere is some share of forest. This makes sense, because onplaces where there is no forest, no deforestation canappear. The model will only be usable on grids where for-ests occur. Therefore, for parameterization, the averageagricultural suitability and the population density of acountry are also only taken from grids which indicate for-est cover.

Development of forest share

After calculating the deforestation rate, the forest sharehas to be updated each year with equation (27) assuringthat the forest share stays within the permissible range of0–1.

Fs fsxi,year≤1 (Buli+Crli)i,year=

fsxi,year

1 (Buli+Crli)fsxi,year>1 (Buli+Crli)(27)

fsxi, year = Fsi, year - 1 - Fi, dec (28)

Aboveground carbon in forest biomass

The model describes the area covered by forests on a cer-tain grid. It can also describe the forest biomass if the aver-

Pr(> |t|)-0.000310***0.019865*0.001529**0.019852*0.001789**0.000206

***

age biomass on a grid is known and the assumption wasmade, that the biomass in forests on the grid is propor-tional to the forest area.

For this reason a global carbon map of aboveground car-bon in forest biomass, was created, based on country val-ues from [2]. By dividing the given total carbon, for eachcountry, with the forest area of the country, the averagebiomass per hectare can be calculated. Now the assump-tion was made, that the stocking biomass per hectare onsites with a higher productivity is higher than on sites witha low productivity. Not for every country with forests [2]gives values of the stocking biomass. So a regression,describing the relation between tC/ha and NPP, was cal-culated and the biomass of grids of missing countries havebeen estimated to obtain a complete global forest biomassmap.

Simulations

In the simulations the effect of different carbon-pricesand/or incentives, for keeping forest, have been tested.The simulation period started in the year 2000 and endsin 2100. The decision, whether deforestation takes placeor not and how fast it goes on, was done in one year timesteps. Scenario drivers, available on coarser time resolu-tion (e.g. population density), have been interpolated lin-early between the given years.

Outputs of the simulations are trajectoria of forest cover,

changes in carbon stocks of forests, and financialresources required to cut emissions from deforestationunder varying scenario assumptions.

Data

The model uses several sources of input data some availa-ble for each grid, some by country aggregates and othersare global. The data supporting the values in table 2 areknown for each grid. Some of the values are also availablefor time series.

Beside the datasets, available at grid level, the purchasingpower parity PPP [7] from 1975–2003, the discount rates[8] for 2004, the corruption in 2005 [5] and the fraction

Table 2: Spatial dataset available on a 0.5° × 0.5° grid

Value

Land areaCountryNPP

Population densityPopulation densityGDPBuildupCropProtected

Agriculture suitabilityBiomassForest area

Year20002000-1990 – 20151990 – 21001990 – 21002010 – 20802010 – 2080

2004200220052000

Source[11][12][10][13][14][14][15][15][16][17]Self[11]

of long living products for the time span 2000–2005 [2]are available for each country (table 3).

The values of table 4 are used globally. Monetary valuesare transformed for each country with their price index.Brazil was taken as the price-reference country asdescribed in [8] and [9].

In figure 11 the net primary productivity taken from [10]is shown. The values range up to 0.75 gC/m2/year. Thehighest productivity is near the equator.

In figure 12 the population density in 2000 and in figure13 in the year 2100 is shown. It can be seen, that the high-est population densities are reached in India and in south-east Asia. The densities are also quite high in Europe andLittle Asia, Central Africa and the coasts of America. Themap of 2100 shows an increase in India and in south-eastAsia.

Figure 14 shows a map of the current forest, crop andbuildup land cover. Large regions are covered by forests.Adjacent to the forests, large areas, used for crop produc-tion, can be seen.

In figure 15 the suitability for agriculture is shown. Mostof the high suitable land is used today for crop production(see figure 14).

Figure 16 shows the carbon in forests. It can be seen, thatthe highest densities are located near the tropical belt.One reason for this is, that the biomass in tropical forestsis high. Note that this picture shows the tons of carbon pergrid and the grid size is 0.5° × 0.5° so the grid has it's larg-est size near the equator.

Figure 17 shows the purchasing power parity which wasused to calculate a price-index. It can be seen that thepoorest countries are in Africa and the richest in NorthAmerica, Europe, Australia and Japan.

Figure 18 shows the discount-rates given in [8]. Here alsothe richest countries have the lowest discount rates.Figure 19 shows the effectiveness of the carbon incentives.In low risk countries nearly all of the spent money will beused for maintaining forest sinks in risky countries not allof the money will come to the desired sink.

Figure 20 shows the proportion of harvested wood enter-ing the long living products pool [2].

Abbreviations

αi: Importance of agriculture

γi: Importance of population

νi: Land price level = minimum land price of reference

country × price index (pxi) [$/ha]

Table 3: Country level values

Discount rate

Fraction of long living productsCorruptionPPP

Source[8][2][5][7]

Table 4: Global values

Baseline

Decay rate longDecay rate shortFactor carbon uptake

Frequency of incentives paymenttC to m3

Harvest lossesHurdle

Maximum rotation intervalMinimum rotation intervalPlanting costsCarbon price

Carbon price incentivesMinimum Land priceMaximum Land priceMinimum wood priceMaximum wood price0.1ln(2)/200.50.55 years40.31.5

140 years5 years800 $/ha0–50 $/tC0–50 $/tC200 $/ha900 $/ha5$/ha35$/ha

ωi Carbon uptake per year [tC/year/ha]θi : Fraction of carbon benefits in products [1]

Ai: Net present value of agriculture [$/ha]AgSi: Agricultural suitability [0–1]

bi: Baseline, how much carbon uptake will be if there is noforest, e.g. 0.1 [1]

BMPi: Biomass in Products [tC/ha]

BMi: Aboveground living wood biomass [tC/ha]

Bi: Present value of carbon benefits [$/ha]Bul: Share of buildup land [1]

C2W: Conversion factor form 1t Carbon to 1m3 wood[m3/tC]

cpi: Planting costs [$/ha]

cpref: Planting costs reference country [$/ha]

CU: Carbon uptake, share of NPP stored in wood [1]Crl: Share of crop land [1]

Figure 11

Net Primary Production (NPP). Areas with a high increment have a high net primary productivity and are indicated by

dark green. Sites with low productivity are indicated by light green.

Figure 12

Population density in Year 2000. Grids with few people are given in white. A rising population density is marked by grey up to high population densities (≥1000 people/km2

) which are indicated by black.

decllp: Decay rate of long living products e.g. 0.03 [1]decslp: Decay rate of short living products e.g. 0.5 [1]DVi: Deforestation Value [$/ha]epci: Effectiv carbon price [$/tC]

fi: Net present value of forestry for one rotation period [$/ha]

Fi: Net present value of forestry [$/ha]Fs: Actual share of forest [0–1]Fdec: Decrease of the forest share

fri: Frequency of incentives money payment [Years]fracllp: Fraction of long living products e.g. 0.5 [0–1]

Figure 13

Population density in Year 2100. Grids with few people are given in white. A rising population density is marked by grey up to high population densities (≥1000 people/km2

) which are indicated by black.

Figure 14

Forest, Crop and Buildup Land cover

. Forests are shown in green, crop in red and buildup land in grey.

fracsb: Fraction of slash burned area e.g. 0.9 [0–1]fracslp: Fraction of short living products e.g. 0.5 [0–1]Fs: Forest area share [0–1]

Fsyear: Forest share of a certain year [1]

fsxyear: Theoretical forest share of a certain year [1]Ftdec : Theoretical decrease of the forest shareGDP: Gross domestic product [$1995/Person]

Hi: Hurdle e.g. 1.5 [1]

HLi: Harvesting losses e.g. 0.2 [1]i: Grid number

leaki: Factor of money which will in real reach the forest[1]

IPi: Incentive payment [$/ha]

MAIi: Mean annual wood volume increment [m3/ha]NPPi: Net primary production [tC/ha/year]

Figure 15

Agriculture suitability

. High suitability for agriculture is marked in dark red. White areas are not suitable for agriculture.

Figure 16

Carbon in Forest biomass. Regions with no carbon in forests are white. Regions with high values of carbon in forests are

dark green.

pci: Carbon price [$/tC]

pcai: Incentives carbon price [$/tC/fri]Pdi: Population density [People/km2]

PLmax: Maximal land price of reference country × priceindex (pxi) [$/ha]

PLmin: Minimal land price of reference country × priceindex (pxi) [$/ha]

PPPi: Purchasing power parity [$]

PPPref: Purchasing power parity of reference country [$]pri: Ratio of area planted [0–1]pwi: Stumpage wood price [$/m3]

pwmax: Maximum revenue of wood, e.g. 35$/fm [$/fm]Pwmin: Minimum revenue of wood, e.g. 5$/fm [$/fm]

Figure 17

Purchasing Power Parity (PPP). Countries with a low purchasing power parity are marked in red, moderate is in green,

high values in blue and very high in magenta.

Figure 18

Discount Rate. Countries with a low discount rate are marked in dark green, moderate countries in yellow and countries

with a high rate in red.

pxi: Price index [1]r: Discount rate [e.g. 0.05]Ri: Rotation interval length [years]

SAgSi: Standardized agricultural suitability [1-10]SFs: Standardized not forest area share [1-10]

Vi: Harvest wood volume [m3]

xi: Theoretical decrease of the forest share if Fsi > 0 ∧ AgSi> 0

Competing interests

The author(s) declare that they have no competing inter-ests.

Authors' contributions

SPd: Standardized population density [1-10]

Georg Kindermann has developed the deforestation ratemodel, implemented the whole model, collected somedata sources and organized them to be used for the imple-

Figure 19

Effectiveness (Corruption). Countries with high values of corruption are marked in red, moderate countries in yellow and

low values in green.

Figure 20

Share of long living products. Countries which use their wood mainly for fuel-wood are marked in blue, those who use it

for sawn-wood are in green.

mentation, runs the simulations, created figures andtables and wrote a first draft of the paper.

Michael Obersteiner developed the core model describingthe forest value, agricultural value and decision of defor-estation, worked on the paper, introduced the maximumtax income and contributed to the payment possibilities.Ewald Rametsteiner contributed to the carbon price andincentives model and their practical implementation,worked on the paper and brought in many backgroundinformations.

Ian McCallum collected and organized the data sourceand produced some figures of the paper.

2.3.

4.5.6.7.8.

Acknowledgements

We acknowledge the support by the Greenhouse Gas Initiative (GGI) project, an institute-wide collaborative effort within IIASA. The interdisci-plinary research effort within GGI links all the major research programs of IIASA that deal with research areas related to climate change, including population, energy, technology, and forestry, as well as LUCs and agricul-ture. GGI's research includes both basic and applied, policy-relevant

research that aims to assess conditions, uncertainties, impacts, and policy frameworks for addressing climate stabilization, from both near-term and long-term perspectives. Support from the EU FP 6 Project Integrated Sink Enhancement Assessment (INSEA, SSPI-CT-2003/503614 with DG RTD) is gratefully acknowledged.

We graceful thank the reviewers for their reports.

9.

10.11.12.

References

1.

IPCC International Panel on Climate Change, ed Robert T Watsonand Core Writing Team: Climate Change 2001: Synthesis Report. A Con-tribution of Working Groups I, II, and III to the Third Assessment Report of

the Integovernmental Panel on Climate Changebridge, United Kingdom, and New York, NY, USA

FAO: Global Forest Resources Assessment 2005, Progress towards sustain-able forest management Volume 147. FAO Forestry Paper. Rome: oodand Agriculture Organization of the United Nations; 2005.

Marland G, Boden T, Andres R: Global, Regional, and NationalCO2 Emissions. In Trends: A Compendium of Data on Global ChangeCarbon Dioxide Information Analysis Center, Oak Ridge NationalLaboratory, U.S. Department of Energy, Oak Ridge, Tenn., U.S.A.;2006.

Benítez PC, Obersteiner M: Site identification for carbonsequestration in Latin America: A grid–based economicapproach. Forest Policy and Economics 2006, 8:636-651.

Kaufmann D, Kraay A, Mastruzzi M: Governance Matters IV: GovernanceIndicators for 1996–2004, World Bank Policy Research Working PaperSeries No. 3630. World BankR Development Core Team: R: A language and environment for statisti-cal computingtistical Computing, Vienna, Austria [ISBN 3-900051-07-0].World Bank: World Development Indicators. World Bank 2005.

Benítez P, McCallum I, Obersteiner M, Yamagata Y: Global Supplyfor Carbon Sequestration: Identifying Least-Cost Afforesta-tion Sites Under Country Risk Consideration. Tech. rep.,International Institute for Applied System Analysis; 2004.

Obersteiner M, Alexandrov G, Benítez PC, McCallum I, Kraxner F,Riahi K, Rokityanskiy D, Yamagata Y: Global Supply of Biomassfor Energy and Carbon Sequestration from Afforestation/Reforestation Activities. Mitigation and Adaptation Strategies forGlobal Change 2006:1381-2386.

Alexandrov GA, Yamagata Y, Oikawa T: Towards a Model for Pro-jecting Net Ecosystem Production of the World Forests. Eco-logical Modelling 1999, 123:183-191.

JRC: The Global Land Cover Map for the Year 2000. GLC2000 databaseResearch Centre

CIESIN: Center for International Earth Science InformationNetwork (CIESIN), Columbia University; and Centro Inter-nacional de Agricultura Tropical (CIAT). Gridded Popula-tion of the World Version 3 (GPWv3): National Boundaries.Palisades, NY: Socioeconomic Data and Applications Center(SEDAC), Columbia University.

Carbon Balance and Management 2006, 1:1513.

CIESIN: Network, Columbia University; and Centro Internacional deCenter for International Earth Science InformationAgricultura Tropical (CIAT). Gridded Population of theWorld Version 3 (GPWv3): Population Density Grids. Pali-sades, NY: Socioeconomic Data and Applications Center(SEDAC), Columbia University.14.

Grubler A, Nakicenovic N, Riahi K, Wagner F, Fischer G, Keppo I,Obersteiner M, O'Neill B, Rao S, Tubiello F: ment of uncertainties in greenhouse gas emissions and theirIntegrated assess-mitigation: Introduction and overview.and Social Change 15.Tubiello FN, Fischer G in press.

Technological Forecastingagriculture: Global and regional effects of mitigation, 2000–: Reducing climate change impacts on16.

2080.WDPA Consortium: Technological Forecasting and Social Change in press.

Copyright World Conservation Union (IUCN) and UNEP-World Database on Protected Areas.World Conservation Monitoring Centre (UNEP-WCMC).17.

2004.

Ramankutty N, Foley JA, Norman J, McSweeney K: tribution of cultivable lands: current patterns and sensitivityThe global dis-to possible climate change.11(5):377-392.

Global Ecology & Biogeography 2002,/content/1/1/15

本文来源：https://www.bwwdw.com/article/dmmj.html