Filetype PDF | Diposting 29 Jun 2022 | 3 thn lalu
Authentication
digital resources
178x Tipe PDF Ukuran file 2.24 MB Source: 2008
Humid tropical forest clearing from 2000 to 2005
quantified by using multitemporal and multiresolution
remotely sensed data
† ‡ §
MatthewC.Hansen*, StephenV.Stehman , PeterV.Potapov*, ThomasR.Loveland* , John R. G. Townshend ,
§¶ †† §
Ruth S. DeFries , Kyle W. Pittman*, Belinda Arunarwati, Fred Stolle**, Marc K. Steininger , Mark Carroll ,
§
andCharleneDiMiceli
*South Dakota State University, Brookings, SD 57007; †State University of New York College of Environmental Science and Forestry, Syracuse, NY 13210;
‡United States Geological Survey, Sioux Falls, SD 57103; §University of Maryland, College Park, MD 20742; Indonesian Ministry of Forestry, Jalan Gatot
††
Subroto, Senayan, Jakarta, 10270 Indonesia; **World Resources Institute, Washington, DC 20002; and Conservation International, Washington, DC 20002
Contributed by Ruth S. DeFries, May 2, 2008 (sent for review February 21, 2008)
Forestcoverisanimportantinputvariableforassessingchangesto needforimprovedmonitoringprograms.Apracticalsolutionto
carbon stocks, climate and hydrological systems, biodiversity rich- examining trends in forest cover change at biome scales is to
ness,andothersustainabilitysciencedisciplines.Despiteincremen- employ remotely sensed data. Satellite-based monitoring of
tal improvementsinourabilitytoquantifyratesofforestclearing, forest clearing can be implemented consistently across large
thereisstill no definitive understandingonglobaltrends.Without regions at a fraction of the cost of obtaining extensive ground
timely and accurate forest monitoring methods, policy responses inventory data. Remotely sensed data enable the synoptic quan-
will be uninformed concerning the most basic facts of forest cover tification of forest cover and change, providing information on
change.Resultsofafeasibleandcost-effectivemonitoringstrategy where and how fast forest change is taking place. Various
arepresentedthatenabletimely,precise,andinternallyconsistent remote-sensing-basedmethodshavebeenprototypedwithinthis
estimatesofforestclearingwithinthehumidtropics.Aprobability- biome(5,7–11)andcombinedwithinformationoncarbonstocks
based sampling approach that synergistically employs low and to estimate carbon emissions (8, 12, 13). The method presented
high spatial resolution satellite datasets was used to quantify here advances the science of monitoring forest cover change by
humidtropicalforestclearingfrom2000to2005.Forestclearingis employing an internally consistent and efficient probability-
estimated to be 1.39% (SE 0.084%) of the total biome area. This based sampling approach that synergistically employs low- and
translates to an estimated forest area cleared of 27.2 million high-spatial-resolution satellite datasets. The results represent a
hectares (SE 2.28 million hectares), and represents a 2.36% reduc- synoptic update on rates of forest clearing within the humid
tion in area of humid tropical forest. Fifty-five percent of total tropics since 2000. For this study, forest clearing equals gross
biome clearing occurs within only 6% of the biome area, empha- forest cover loss during the study period without quantification
sizing the presence of forest clearing ‘‘hotspots.’’ Forest loss in ofcontemporaneousgainsinforestcoverduetoreforestationor
Brazil accountsfor47.8%oftotalbiomeclearing,nearlyfourtimes afforestation. The method presented could be implemented
that of the next highest country, Indonesia, which accounts for repeatedly for both forest cover loss and gain in establishing
12.8%. Over three-fifths of clearing occurs in Latin America and internally consistent biome-scale trends in both gross and net
over one-third in Asia. Africa contributes 5.4% to the estimated forest cover loss and/or gain.
loss of humid tropical forest cover, reflecting the absence of Moderate spatial resolution (250 m, 500 m, and 1 km) data
current agro-industrial scale clearing in humid tropical Africa. from the MODerate Resolution Imaging Spectroradiometer
(MODIS)areimagednearly daily at the global scale, providing
deforestation humid tropics remote sensing change detection the best possibility for cloud-free observations from a polar-
monitoring orbiting platform. However, MODIS data alone are inadequate
for accuratechangeareaestimationbecausemostforestclearing
uantifying rates of humid tropical forest cover clearing is occursatsub-MODISpixelscales.High-spatial-resolutionLand-
Qcritical for many areas of earth system and sustainability sat data (28.5 m), in contrast, do allow for more accurate
science, including improved carbon accounting, biogeochemical measurement of forest area cleared. However, because of infre-
cycle and climate change modeling, managementofforestryand quentrepeatcoverage,frequentcloudcover,anddatacosts,the SCIENCE
agricultural resources, and biodiversity monitoring. Concerning useofLandsatdataforbiome-scalemappingisoftenprecluded.
land cover dynamics, humid tropical forest clearing results in a Integrating both MODIS and Landsat data synergistically en- SUSTAINABILITY
large loss of carbon stock when compared with most other ables timely biome-scale forest change estimation.
change scenarios. The humid tropical forests are also the site of WeusedMODISdatatoidentify areas of likely forest cover
considerable economic development through direct forestry lossandtostratifythehumidtropicsintoregionsoflow,medium,
exploitation and frequent subsequent planned agro-industrial and high probability of forest clearing. A stratified random
activities. The result is that tropical forests and their removal sample of 183 18.5-km 18.5-km blocks taken within these
feature prominently in the global carbon budget (1). In addition,
the humid tropics include the most biodiverse of terrestrial
ecosystems(2),andthelossofhumidtropicalforestcoverresults Authorcontributions: M.C.H., S.V.S., T.R.L., J.R.G.T., and R.S.D. designed research; M.C.H.,
in a concomitant loss in biodiversity richness. S.V.S., P.V.P., T.R.L., K.W.P., and M.K.S. performed research; M.C.H., S.V.S., P.V.P., K.W.P.,
Assessing the dynamics of this biome is difficult because of its B.A., F.S., M.K.S., M.C., and C.D. analyzed data; and M.C.H., S.V. S., P.V.P., and R.S.D. wrote
the paper.
sheer size and varying level of development within and between Theauthors declare no conflict of interest.
countries. To date, there is no clear consensus on the trends in ¶
forest cover within the humid tropics. Grainger (3) illustrated To whomcorrespondence should be addressed at: 2181 Lefrak Hall, University of Mary-
land, College Park, MD 20742. E-mail: rdefries@mail.umd.edu.
this point mainly through the use of data from the Food and This article contains supporting information online at www.pnas.org/cgi/content/full/
Agriculture Organization of the United Nations Forest Re- 0804042105/DCSupplemental.
source Assessments (4–6) and consequently emphasized the ©2008byTheNationalAcademyofSciencesoftheUSA
www.pnas.orgcgidoi10.1073pnas.0804042105 PNAS July 8, 2008 vol. 105 no. 27 9439–9444
Fig.1. Forestclearingandforestcoverinthehumidtropicalforestbiome,2000–2005.Totalforestclearingoverthestudyperiodisestimatedtobe27.2million
hectares (SE 2.28 million hectares). Regional variation in clearing intensity is shown: Region 1 covers 6% of the biome and contains 55% of clearing; region 2
covers 44%ofthebiomeandcontains40%offorestclearing;andregion3covers50%ofthebiomeandcontains5%offorestclearing.Datafromthisfigure
are available at http://globalmonitoring.sdstate.edu/projects/gfm.
regions was interpreted for forest cover and forest clearing by 5%offorestclearing is found within a third region consisting of
using high-spatial-resolution Landsat imagery from 2000 and the remaining predominantly intact forest zones (35% of the
2005. Typically, Landsat imagery has been used to provide biome area) and areas largely deforested before 2000 (15% of
regionalforestareachangeestimatesbecauseitssufficientlyhigh the biome area).
spatial resolution enables the detection of most forest clearing Ourfindings emphasize the predominance of Brazil in humid
events (11, 14, 15). Consistent with this practice, our estimates tropical forest clearing (Table 1). By area, Brazil accounts for
of forest clearing are based on interpreting Landsat imagery for 47.8%ofallhumidtropicalforestclearing,nearlyfourtimesthat
the 183 sample blocks selected. Our sampling strategy differs ofthenexthighestcountry,Indonesia,whichaccountsfor12.8%
from previous efforts (5, 8) in that we took advantage of forest ofthetotal.Overthree-fifthsofclearingoccursinLatinAmerica
clearing information available from independent imagery, the and over one-third in Asia. Forest clearing as a percentage of
MODISchangeindicatormaps,todefinestrataandtoconstruct year-2000 forest cover for Brazil (3.6%) and Indonesia (3.4%)
regression estimators of forest clearing. exceedstherestofLatinAmerica(1.2%),therestofAsia(2.7%),
Results and Africa (0.8%). Beyond the arc of deforestation in Brazil,
Latin American hotspots include northern Guatemala, eastern
Our results reveal that rates of clearing in the biome remain Bolivia, and eastern Paraguay. As a percentage of year-2000
comparable with those observed in the 1990s (5, 8, 9). Forest forest cover, Paraguay features the highest areal proportion of
clearing is estimated to be 1.39% (SE 0.084%)ofthetotalbiome change hotspots, indicating an advanced, nearly complete forest
area. This translates to an estimated forest area cleared of 27.2 clearing dynamic. Indonesian island groups of Sumatra, Kali-
million hectares (SE 2.28 million hectares) and represents a mantan, Sulawesi, and Papua feature varying degrees of forest
2.36% reduction in year-2000 forest cover. Fig. 1 depicts the removal, with Sumatra the site of the most intense recent
spatial variation in gross forest cover loss from 2000 to 2005. The large-scale clearing and Papua a measurable but low level of
biome can be divided into three regions of forest clearing forest clearing. Riau province in Sumatra has the highest indi-
intensity.Thefirstregionconsistsofareaswith5%clearingper cated change within Indonesia. Hot spots of clearing are present
block and largely captures the current centers of agro-industrial in every state of Malaysia, and clearing in Cambodia along its
scale clearing in South America and Insular Southeast Asia. Of border with Thailand is among the highest of indicated change
the total biome area cleared, 55% occurs in this region that hot spots. Africa, although a center of widespread, low-intensity
constitutes only 6% of the biome area, illustrating the presence selectivelogging(16),contributesonly5.4%totheestimatedloss
of forest clearing ‘‘hotspots’’ (region 1 in Fig. 1). The second of humidtropicalforestcover.Thisresultreflectstheabsenceof
regionof0.7–5%clearingperblockconstitutes44%ofthebiome current agro-industrial scale clearing in humid tropical Africa.
area. This region consists of less spatially concentrated clearing Our results reveal a higher degree of regional variation in
andaccountsfor40%ofallclearingwithinthebiome.Theother forest clearing than currently portrayed by the only other source
9440 www.pnas.orgcgidoi10.1073pnas.0804042105 Hansenetal.
Table 1. Regional estimates of humid tropical forest area cleared
Percent contribution Within-region forest Within-region forest loss
Percent of of region to forest loss as percent of as percent of year 2000 Blocks
Region biomearea loss in the biome land area (SE) forest area sampled
Brazil 27.09 47.8 2.45 (0.14) 3.60% 53
Americas sans Brazil 21.27 12.6 0.82 (0.13) 1.23 10
Indonesia 9.16 12.8 1.95 (0.20) 3.36 77
Asia sans Indonesia 27.60 21.4 1.08 (0.33) 2.68 31
Africa 14.88 5.4 0.50 (0.13) 0.76 12
Pan-Americas 48.36 60.4 1.73 (0.10) 2.56 63
Pan-Asia 36.76 34.3 1.29 (0.25) 2.90 108
Biometotal 100 100 1.39 (0.084) 2.36 183
of information for the pan-tropics during the study period, the Considerable debate on the appropriate use of Landsat data
2005 Forest Resource Assessment (FRA) report from the Food for regional monitoring has concerned the alternative uses of
and Agriculture Organization of the United Nations (6). The exhaustive mapping versus sampling-based approaches (17–19).
FRA2005reporthighlightsAfricaandSouthAmericaashaving Datalimitations, namely cloud cover and costs of imagery, have
the highest rates of forest area loss, both in excess of 4 million been the principal arguments against exhaustive mapping. The
hectares per year. For those African countries predominantly challenge to a sampling approach is that change is typically rare
within the humid tropics, our humid-tropics-only estimate is less at the scale of a biome. Consequently, a critical requirement for
than one-third of the FRA estimate. For both this study and the obtaining precise sample-based estimates is to construct strata
FRA, Brazil and Indonesia are the countries featuring the thateffectivelyidentifyareasofintensiveforestclearing.Theuse
highest forest clearing rates. However, our results differ as to ofexpertopiniontodelineatebroadregionsofsuspectedchange
the relative magnitude of change. For Brazil and Indonesia, the has been used to achieve this end (8). In contrast, we imple-
FRA reports annual change in forest area from 2000 to 2005 mented a more spatially targeted approach to stratification,
equal to 3.10 and 1.87 million ha/yr, respectively (6). Our using MODISimagerytoflagareasoflikelyforestclearing.The
estimates of forest clearing for Brazil and Indonesia are 2.60 and MODIS imagery allowed assigning each 18.5-km 18.5-km
0.70 million ha/yr, respectively. The results for Indonesia rep- block in the biome individually to a stratum, thus improving on
resent a dramatic decrease from 1990 to 2000 clearing rates. the broader regional strata used previously (8). Furthermore,
MODISimagery allows for the identification of clearing on an
Discussion annual basis and therefore provides a more temporally resolved
Our strategy incorporating the MODIS-derived forest clearing view of change than possible with Landsat data alone.
information in both the sampling design (stratification) and An additional criticism of the sampling approach is the
estimation (regression estimator) components of the monitoring absence of a spatial representation of where in the biome forest
strategy yielded the requisite precision and cost efficiency clearing is occurring. We address this concern by applying the
desired for an operational monitoring protocol at the pan- stratum-specific regression models relating Landsat-derived
tropical scale. The standard error we obtained for the biome- clearing to MODIS-derived clearing at the support of the
wide estimated forest loss of the humid tropics was comparable 18.5-km18.5-kmblockstopredictclearingforeachblock(Fig.
with those reported by the Food and Agriculture Organization 1). Thisspatialdepictionofforestclearingtakesadvantageofthe
of the United Nations in 2000 (5) and Achard et al. (8), but we respective strengths of the complete coverage MODIS imagery
were able to achieve this level of precision with much smaller andthehighspatialresolutionoftheLandsatimagery.Themore
sample coverage. The total area of Landsat imagery sampled in frequent temporal coverage of the MODIS imagery alleviates
our study was 0.21% of the biome, whereas previous studies (5, the problem of cloud cover obscuring tropical areas during the
8) used samples covering 10% and 6.5% of the tropical domain. few available Landsat overpasses (20). Calibrating the MODIS-
derived clearing values based on the Landsat-derived clearing
Our sampling strategy thus yields precise estimates of forest observed on the sample blocks compensates for the inability of
clearing based on an areal sample coverage that could be the larger MODIS pixel size (500 m) to detect smaller areas of SCIENCE
sustainable from an effort and cost standpoint for future mon- clearing that are observable from the 28.5-m Landsat pixels. SUSTAINABILITY
itoring goals. Our approach is readily adaptable to other high- Although area estimates derived from coarser-resolution data
spatial-resolution sensors because the success of the strategy are commonly calibrated by using a nonrandom sample of
derives from advantageously incorporating the MODIS data in high-resolution data (21, 22), a strength of our approach is that
both the sampling design and analysis components. by implementing a probability sampling design to collect the
Table 2. Stratified sampling design
Humidtropics (excluding Indonesia) Indonesian humid tropics
MODISchange No. of blocks Percent of No. of blocks Percent of
(90%) Stratum no. sampled stratum sampled Stratum no. sampled stratum sampled
0–2% 1A 21 0.08 5A 8 0.51
1B 25 0.12 5B 33 1.17
2–9% 2 23 1.76 6 17 9.24
9% 3 32 8.10 7 18 26.09
— 4(certainty) 5 100 8(certainty) 1 100
Hansen et al. PNAS July 8, 2008 vol. 105 no. 27 9441
a Central African Republic (3°25’N, 15°37’E). Low change stratum. Forest loss 0.1%.
12/14/2000 1/10/2005 Classification results
b Brazil (11°25’S, 56°32’W). Medium change stratum. Forest loss 8.2%.
7/30/2001 7/9/2005 Classification results
Malaysia (4°15’N, 117°24’E). High change stratum. Forest loss 33.0%.
c
7/10/2001 2/27/2005 Classification results
Brazil (11°25’S, 55°51’W). Highest change stratum (certainty stratum). Forest loss 37.3%.
d
7/30/2001 7/9/2005 Classification results
Forest cover, 2000 Forest loss areas
Non-forest areas No data/Clouds
Fig. 2. ExamplesofLandsatsampleblockscharacterizedtoestimateforestcoverandchangefrom2000to2005.Eachblockcovers18.532kmpersideandhas
beenreprojectedintolocalUniversalTransverseMercatorcoordinates.Thestrataarecreatedbyusingthebiome-wideMODIS2000to2005forestclearingprobability
maps.(a)SampleblockfromtheMODISchangestrata1and5.(b)SampleblockfromMODISchangestrata2and6.(c)SampleblockfromMODISchangestrata3and
7. (d) Sample block from MODIS change certainty strata 4 and 8. All blocks used in this analysis can be viewed at http://globalmonitoring.sdstate.edu/projects/gfm.
sample of high-resolution data, we retain the rigorous design- data offer a way forward for monitoring forests in support of
basedinferenceframework(23)tosupportthestatisticalvalidity both basic earth science research and policy formulation and
of our estimates. Furthermore, by construction, the aggregate implementation. For example, these results could be combined
predictedchangeoveranydefinedsubregionofthebiome(Table withinformationoncarbonstockstosupportcarbonaccounting
1)equalstheestimatedforestcoverlossderivedfromthesample programs such as the ‘‘Reducing Emissions for Deforestation
blocks, thus ensuring internal consistency between the mapped and Degradation’’ (REDD) initiative (25). Such an approach
(Fig. 1) and estimated forest loss. could be implemented at both national and regional scales for
The results of this analysis highlight the need for internally the synoptic assessment of forest cover change and the moni-
consistent biome-scale monitoring to accurately depict relative toring of intra- or international displacement, or leakage, of
variations in forest clearing dynamics within and between coun- forest cover clearing.
tries. Results from national-scale studies that employ varying Although forest resources are a key component of economic
methods, definitions, and input data may result in incompatible developmentinthisbiome,forestgovernanceisgreatlyhindered
products that preclude regional syntheses (24). Biome-scale by a lack of timely information on change within the forest
forest cover and change estimates derived from remotely sensed domain. A monitoring strategy combining data from sensors at
9442 www.pnas.orgcgidoi10.1073pnas.0804042105 Hansenetal.
no reviews yet
Please Login to review.
