Peter Schlesinger

Study site is the Trifinio Region of Central America. Forest PAs (12) in Trifinio are classified into two management restriction levels following the IUCN classification system—strict (Categories I-II) and multiple-use (Categories III-IV). PA classifications based on level of decentralization and management capacity is based on interview primary data. Medium resolution remotely sensed Landsat images (30 m pixels) were acquired to develop least-cloud, NDVI values over 30 years in seven epochs, spaced at approximately five-year intervals, starting in 1986 and ending in 2016, during the dry season (between January and March). Two dependent variables were developed with these data: (1) mean NDVI value for 1986-2016, averaging NDVI values across all seven epochs, and (2) annual change of NDVI values between 1986-2016. A number of spatial variables expected to be correlated with designation as a PA and greenness values were sampled for each pixel in the data set. These include baseline ND...

Study site is the Trifinio Region of Central America. Forest PAs (12) in Trifinio are classified into two management restriction levels following the IUCN classification system—strict (Categories I-II) and multiple-use (Categories III-IV). PA classifications based on level of decentralization and management capacity is based on interview primary data. Medium resolution remotely sensed Landsat images (30 m pixels) were acquired to develop least-cloud, NDVI values over 30 years in seven epochs, spaced at approximately five-year intervals, starting in 1986 and ending in 2016, during the dry season (between January and March). Two dependent variables were developed with these data: (1) mean NDVI value for 1986-2016, averaging NDVI values across all seven epochs, and (2) annual change of NDVI values between 1986-2016. A number of spatial variables expected to be correlated with designation as a PA and greenness values were sampled for each pixel in the data set. These include baseline ND...

A simple GIS soil-water balance model for the Amazon Basin, called RisQue (Risco de Queimadasa -- Fire Risk), was used to conduct an analysis of spatial and temporal patterns of drought in moist tropical forests and the complex relationships between patterns of drought and forest fire regimes from 1995 through 2001. The provided data products are the model output estimates of maximum plant-available soil water (PAWmax) at 10 m depth at 8 km resolution and model data inputs of monthly precipitation and evapotranspiration. RisQue estimates PAWmax at 10 m depth starting with a map of PAWmax (1-2 m depth) developed using 1,565 RADAMBRASIL soil texture profiles and empirical relationships between soil texture and critical soil water parameters and then interpolated to 8 km resolution. In RisQue, plant-available soil water (PAW) is depleted by monthly evapotranspiration estimated using the Penman Monteith equation and satellite-derived radiation and recharged by monthly precipitation. There are three data files with this data set, two *.zip, and one GeoTIFF image (.tif). The *.zip files expand to 83 *.asc files of evapotranspiration and 89 *.asc files for precipitation data. The image (.tif) is a map of maximum percent available water at 10 m depth. All the files in this data set are in standard arc/info asciigrid format at 8 km resolution.

This data set contains moisture content measurements for fuel sticks located in the forest understory of the rainfall exclusion experimental site, Tapajos National Forest, Para, Brazil. The mean and standard errors are reported for control and treatment plot measurements. The measurements were taken on various dates and times of day between 1998 and 2000 during the dry season. The rainfall exclusion treatment began in late January 2000 and continued through December 2004. About 60% of throughfall (equivalent to approximately half the rainfall) was diverted from a 1-hectare plot (i.e., dry) using plastic panels installed in the understory. The comparable 1-hectare control plot (i.e., wet) was unaltered. The purpose was to observe the potential effects of severe water stress on a humid Amazonian forest (Nepstad et al., 2002). There is one comma-separated (.csv) data file with this data set.

This data set provides the results of the two modeled scenarios for future patterns of deforestation across the Amazon Basin from 2002 to 2050. This larger defined Amazon Basin (PanAmazon area) includes the Amazon River watershed, the Legal Amazon in Brazil, and the Guiana region. The model SimAmazonia was used to simulate monthly deforestation in the Amazon Basin from 2002 to 2050 for two scenarios: (1) a "Business-as-Usual" scenario, which considered the deforestation trends across the basin and projected the rates by using historical images and their variations from 1997 to 2002 and then added to that the effect of paving a set of major roads, and (2) a "Governance" scenario, that also considered the current deforestation trends, but assumed a 50% limit imposed for deforested land within each basin's subregion, and that existing and proposed Protected Areas (PAs), play a decisive role in limiting deforestation as well (Soares et al., 2006). The provided da...

A AMAZÔNIA está entrando em uma era de rápidas mudanças impulsionadas pela previsão de asfaltamento de rodovias que estimularão a expansão da fronteira agrícola e de exploração madeireira. O declínio do custo de transporte tem importantes implicações para a biodiversidade, emissão de gases que contribuem para o efeito estufa e prosperidade da sociedade da Amazônia a longo prazo. Para analisar esse contexto, foi desenvolvido um modelo de simulação de desmatamento na bacia Amazônica, sensível a diferentes cenários de políticas públicas frente à expansão da infra-estrutura de transporte pela região. Resultados do modelo indicam que, dentro de um cenário pessimista, o desmatamento projetado pode eliminar, até meados deste século, 40% dos atuais 5,4 milhões de km² de florestas da Amazônia, liberando o equivalente a 32 Pg (10(9) toneladas) de carbono para atmosfera. A modelagem de cenários alternativos aponta que a expansão de uma rede de áreas protegidas, efetivamente implementadas, pode...

We have developed a map of the land cover of South America based largely on NOAA AVHRR LAC (Local Area Coverage) 1-km resolution data. Areas of South America for which there were no LAC data due to cloud cover or other reasons were supplemented with a 3-year time series of 15-km resolution NOAA AVHRR GVI data. Thirty-nine classes were labeled using existing vegetation maps, from phenology, and by visual interpretation of imagery. We estimated that nine percent of the closed tropical moist forests of South America had been cleared in the decade prior to 1988, the date of the imagery; 22 percent of the original closed forests, which include tropical moist forests as a major component, of South America had been cleared or degraded: 18 percent of the original woodland and 25 percent of the grasslands had been cleared or degraded.

We have developed several impermeable surface maps for all Cape Cod (Barnstable County) to help assess the contribution of paved and other impermeable surfaces to declines in local and regional water quality. These maps have been assembled with the cooperation of many town planning departments, the Cape Cod National Seashore, Mass Military Reservation, the Cape Cod Commission, the Association to Preserve Cape Cod (APCC) and from IKONOS data. The map are being used to predict where new impermeable surface will occur, define current and future hotspots of non-point pollution, and to map the relationships of impermeable surfaces to the zones of contribution (ZOC) of municipal wells. The maps are also used to define the percentage of impermeable surfaces in buffer zones around ponds and estuaries. Combining these data with census data on housing and population density allows us to define the importance of impervious surface as minor or major factors in water pollution.

Protected areas (PAs) are a prominent approach to maintaining and enhancing biodiversity and ecosystem services. A critical question for safeguarding these resources is how PA governance processes and management structures influence their effectiveness. We conduct an impact evaluation of 12 PAs in three Central American countries to assess how processes in management restrictions, management capacity, and decentralization affect the annual change in the satellite-derived Normalized Difference Vegetation Index (NDVI). NDVI varies with greenness that relates to plant production, biomass, and important ecosystem functions related to biodiversity and ecosystem services such as water quality and carbon storage. Any loss of vegetation cover in the form of deforestation or degradation would show up as a decrease in NDVI values over time and gains in vegetation cover and regeneration as an increase in NDVI values. Management restriction categories are based on international classifications ...

The forests of Russia cover a larger area and hold more carbon than the forests of any other nation and thus have the  potential for a major role in global warming. Despite a systematic inventory of these forests, however, estimates of total carbon stocks vary, and spatial variations in the stocks within large aggregated units of land are unknown, thus hampering measurement of sources and sinks of carbon. We mapped the distribution of living forest biomass for the year 2000 by developing a relationship between ground measurements of wood volume at 12 sites throughout the Russian Federation and data from the MODIS satellite bidirectional reflectance distribution function (BRDF) product (MOD43B4). Based on the results of regression-tree analyses, we used the MOD43B4 product to assign biomass values to individual 500 m × 500 m cells in areas identified as forest by two satellite-basedmaps of land cover. According to the analysis, the total living biomass varied between 46 and 67 Pg, largely because of different estimates of forest area. Although optical data are limited in distinguishing differences in biomass in closed canopy forests, the estimates of total living biomass obtained here varied more in response to different definitions of forest than to saturation of the optical sensing of biomass.

We have developed a map of the land cover of South America based largely on NOAA AVHRR (Local Area Coverage) 1-km resolution data. Areas of South America for which there were no LAC data due to cloud cover or other reasons were supplemented with a S-year time series of l-km resolution NOAAA AVHRR VI data. Thirty-mine classes were labeled using existing vegetation maps, from phenology, and by visual interpretation of imagery. We estimated that nine percent of the closed tropical moist forests of South America had been cleared in the decade prior to 1988, the date of the imagery; 22 percent of the original closed forests, which include tropical moist forests as a major component, of South America had been cleared or degraded; 18 percent of the original woodland and 25 percent of the grasslands had been cleared or degraded.

Soybean production is one of the main economic forces driving the expansion of the agricultural frontier in the Brazilian Amazon. To assess the potential for expansion we estimate a model of soybean yield that integrates the major climatic, edaphic, and economic determinants in the Amazon Basin. Yield is modeled as a function of yield as simulated by a crop physiology model that captures the effects of climate and physical attributes on the development of soybean plant; fertilizer applications; and economic/spatial parameters
such as credit, transports costs and latitude. Current values of these determinants indicate that roughly 20% of Amazon Region or ∼1,000,000 km2 (excluding protected areas) can
generate yields greater than 2000 kg/ha. Soybean production may be possible over a wider area of Amazon, but realizing this potential requires improvements in economic determinants such as the transportation infrastructure.