Country: Brazil

MAAP #108: Understanding the Amazon Fires with Satellites, part 2

Posted on by dcadmin
Base Map. Updated Amazon fire hotspots map, August 20-26, 2019. Red, Orange, and Yellow indicate the highest concentrations of fire, as detected by NASA satellites that detect fires at 375 meter resolution. Data. VIIRS/NASA, MAAP.

Here we present an updated analysis on the Amazon fires, as part of our ongoing coverage and building off what we reported in MAAP #107.

First, we show an updated Base Map of the “fire hotspots” across the Amazon, based on very recent fire alerts (August 20-26). Hotspots (shown in red, orange, and yellow) indicate the highest concentrations of fire as detected by NASA satellites.

Our key findings include:

– The major fires do NOT appear to be in the northern and central Brazilian Amazon characterized by tall moist forest (Rondônia, Acre, Amazonas, Pará states),* but in the drier southern Amazon of Brazil and Bolivia characterized by dry forest and shrubland (Mato Grosso and Santa Cruz).

– The most intense fires are actually to the south of the Amazon, along the border of Bolivia and Paraguay, in areas characterized by drier ecosystems.

– Most of the fires in the Brazilian Amazon appear to be associated with agricultural lands. Fires at the agriculture-forest boundary may be expanding plantations or escaping into forest, including indigenous territories and protected areas.

– The large number of agriculture-related fires in Brazil highlights a critical point: much of the eastern Amazon has been transformed into a massive agricultural landscape over the past several decades. The fires are a lagging indicator of massive previous deforestation.

– We continue to warn against using satellite-based fire detection data alone as a measure of impact to Amazonian forests. Many of the detected fires are in agricultural areas that were once forest, but don’t currently represent forest fires.

In conclusion, the classic image of wildfires scorching everything in their path are currently more accurate for the unique and biodiverse dry forests of the southern Amazon then the moist forests to the north. However, the numerous fires at the agriculture-moist forest boundary are both a threat and stark reminder of how much forest has been, and continues to be, lost by deforestation.

Next, we show a series of 11 satellite images that show what the fires look like in major hotspots and how they are impacting Amazonian forests. The location of each image corresponds to the letters (A-K) on the Base Map.

*If anyone has detailed information to the contrary, please send spatial coordinates to maap@amazonconservation.org

Zooms A, B: Chiquitano Dry Forest (Bolivia)

Some of the most intense fires are concentrated in the dry Chiquitano of southern Bolivia. The Chiquitano is part of the largest tropical dry forest in the world and is a unique, high biodiversity, and poorly explored Amazonian ecosystem. Zooms A-C illustrate fires in the Chiquitano between August 18-21 of this year, likely burning a mixture of dry forest, scrubland, and grassland.

Zoom A. Recent fires in the dry Chiquitano of southern Bolivia. Data: Planet.
Zoom B. Recent fires in the dry Chiquitano of southern Bolivia. Data: Planet.

Zoom D: Beni Grasslands (Bolivia)

Zoom D shows recent fires and burned areas in Bolivia’s Beni grasslands.

Zoom D. Recent fires and burned areas in Bolivia’s Beni grasslands. Data: ESA.

Zooms E,F,G,H: Brazilian Amazon (Amazonas, Rondônia, Pará, Mato Grosso)

Zoom E-H take us to moist forest forests of the Brazilian Amazon, where much of the media and social media attention has been focused. All fires we have seen in this area are in agricultural fields or at the agriculture-forest boundary. Note Zoom E is just outside a national park in Amazonas state; Zoom F shows fires at the agriculture-forest boundary in Rondônia state; Zoom G shows fires at the agriculture-forest boundary within a protected area in Pará state; and Zoom H shows fires at the agriculture-forest boundary in Mato Grosso state.

Zoom E. Fires at the agriculture-forest boundary outside a national park in Amazonas state. Data: Planet.
Zoom F. Fires at the agriculture-forest boundary in Rondônia state. Data: ESA.
Zoom G. Fires at the agriculture-forest boundary within a protected area in Pará state.
Zoom H. Fires at the agriculture-forest boundary in Mato Grosso. Data: ESA.

Zooms I, J: Southern Mato Grosso (Brazil)

Zooms I and J shows fires in grassland/scrubland at the drier southern edge of the Amazon Basin. Note both of these fires are within Indigenous Territories.

Zoom I. Fires within an Indigenous Territory at the drier southern edge of the Amazon Basin. Data: Planet.
Zoom J. Fires within an Indigenous Territory at the drier southern edge of the Amazon Basin. Data: Planet.

Zooms C, K: Bolivia/Brazil/Paraguay Border

Zooms C and K show large fires burning in the drier ecosytems at the Bolivia-Brazil-Paraguay border. This area is outside the Amazon Basin, but we include it due it’s magnitude.

Zoom C. Recent fires in the dry Chiquitano of southern Bolivia. Data: Planet.
Zoom K. Large fires burning around the Gran Chaco Biosphere Reserve. Data: NASA/USGS.

Acknowledgements

We thank  J. Beavers (ACA), A. Folhadella (ACA), M. Silman (WFU), S. Novoa (ACCA), M. Terán (ACEAA), and D. Larrea (ACEAA) for helpful comments to earlier versions of this report.

This work was supported by the following major funders: MacArthur Foundation, International Conservation Fund of Canada (ICFC), Metabolic Studio, and Global Forest Watch Small Grants Fund (WRI).

Citation

Finer M, Mamani N (2019) Seeing the Amazon Fires with Satellites. MAAP: 108.

MAAP #107: Seeing the Amazon Fires with Satellites

Posted on by dcadmin
Recent fire (late July 2019) in the Brazilian Amazon. Data: Maxar.

Fires now burning in the Amazon, particularly Brazil and Bolivia, have become headline news and a viral topic on social media.

Yet little information exists on the impact on the Amazon rainforest itself, as many of the detected fires originate in or near agricultural lands.

Here, we advance the discussion on the impact of the fires by presenting the first Base Map of current “fire hotspots” across three countries (Bolivia, Brazil, and Peru). We also present a striking series of satellite images that show what the fires look like in each hotspot and how they are impacting Amazonian forests. Our focus is on the most recent fires in August 2019.

Our key findings include:

  • Fires are burning Amazonian forest in Bolivia, Brazil, and Peru.
    .
  • The fires in Bolivia are concentrated in the dry Chiquitano forests in the southern Amazon.
    .
  • The fires in Brazil are much more scattered and widespread, often associated with agricultural lands. Thus, we warn against using fire detection data alone as a measure of impact as many are clearing fields. However, many of the fires are at the agriculture-forest boundary and maybe expanding plantations or escaping into forest.
    .
  • Although not as severe, we also detected fires burning forest in southern Peru, in an area that has become a deforestation hotspot along the Interoceanic Highway.

Given the nature of the fires in Bolivia and Brazil, estimates of total burned forest area are still difficult to determine. We will continue monitoring and reporting on the situation over the coming days.


Base Map

The Base Map shows “fire hotspots” for the Amazonian regions of Bolivia, Brazil, and Peru in August 2019. The data comes from a NASA satellite that detects fires at 375 meter resolution. The letters (A-G) correlate to the satellite image zooms below.

Base Map. Fire Hotspots in the Amazon during August 2019. Data: VIIRS/NASA.

Zoom A: Southern Bolivian Amazon

Fires are concentrated in the dry Chiquitano of southern Bolivia. It is part of the largest tropical dry forest in the world. The fires coincide with areas that have been part of cattle ranching expansion in recent decades (References 1 and 2), suggesting that poor burning practices could be the cause of the fires. Ranching using sown pastures has previously been referred to as a direct cause of forest loss in Bolivia (References 2 and 3). The Bolivian National Service of Meteorology and Hydrology (SENAMHI) issued high wind alerts in July and August for southern Bolivia, which could have led to the expansion of poorly managed fires. Also, August is usually the driest month of the year in this region. These conditions could explain the origin (poor fire practice) and expansion (little rain and strong winds) of the current fires.

Zoom A1. Fire in southern Bolivian Amazon. Data: ESA
Zoom A2. Fire in southern Bolivian Amazon. Data: ESA
Zoom A3. Fire in southern Bolivian Amazon. Data: Planet

Zooms B, C, E, F, G: Western Brazilian Amazon

The major fires in western Brazil seem to be at the agriculture-forest boundary. Note that Zoom B shows fire in a protected area in Amazonas state; Zoom C seems to show fire escaping (or deliberately set) in the primary forests in Rondonia state; and Zooms F and G seems to show fire expanding plantation into forest in Amazonas and Mato Grosso states, respectively.

Zoom B. Fire in a protected area in Amazonas state. Data: ESA
Zoom C. Fires at agriculture-forest boundary in Rondonia state. Data: Sentinel.
Zoom E. Fire escaping (or deliberately set) in the primary forests in Rondonia state. Data: Planet
Zoom F. Fire that seems to be expanding plantation into forest in Amazonas state. Data: Planet.
Zoom G. Fire that seems to be expanding plantation into forest in Mato Grosso state. Data: Planet.
Bonus Zoom. Recent fire in Brazilan Amazon. Data: Planet.

 

Zoom D: Southern Peruvian Amazon

Fires burning forest near the town of Iberia, an area along the Interoceanic Highway that has become a deforestation hotspot in the region of Madre de Dios (see MAAP #28 and MAAP #47).

Zoom D. Fire in southern Peruvian Amazon (near Iberia, Madre de Dios). Data: ESA

Additonal References

We have these to be some of the most informative additional references:

New York Times, Aug 24

Global Forest Watch, Aug 23

Technical References

1 Müller, R., T. Pistorius, S. Rohde, G. Gerold & P. Pacheco. 2013. Policy options to reduce deforestation based on a systematic analysis of drivers and agents in lowland Bolivia. Land Use Policy 30(1): 895-907. http://dx.doi.org/10.1016/j. landusepol.2012.06.019

2 Muller, R., Larrea-Alcázar, D.M., Cuéllar, S., Espinoza, S. 2014.  Causas directas de la deforestación reciente (2000-2010) y modelado de dos escenarios futuros  en las tierras bajas de Bolivia. Ecología en Bolivia 49: 20-34.

3 Müller, R., P. Pacheco & J. C. Montero. 2014. El contexto de la deforestación y degradación de los bosques en Bolivia: Causas, actores e instituciones. Documentos Ocasionales CIFOR 100, Bogor. 89 p.

Acknowledgements

We thank  J. Beavers, D. Larrea, T. Souto, M. Silman, A. Condor, and G. Palacios for helpful comments to earlier versions of this report.

This work was supported by the following major funders: MacArthur Foundation, International Conservation Fund of Canada (ICFC), Metabolic Studio, and Global Forest Watch Small Grants Fund (WRI).

Citation

Novoa S, Finer M (2019) Seeing the Amazon Fires with Satellites. MAAP: 107.

MAAP #100: Western Amazon – Deforestation Hotspots 2018 (a regional perspective)

Posted on by dcadmin
Base Map. Deforestation Hotspots in the western Amazon. Data: Hansen/UMD/Google/USGS/NASA, GFW, SERNANP, SNAP, SINAP, SERNAP, RAISG

For the 100th MAAP report, we present our first large-scale western Amazon analysis: Colombia, Peru, Ecuador, Bolivia, and western Brazil (see Base Map).

We use the new 2018 data for forest cover loss, generated by the  University of Maryland (Hansen et al 2013) and presented by Global Forest Watch.

These data indicate 2.5 million acres of forest cover loss in the western Amazon in 2018.*

We conducted an additional analysis that indicates, of this total, 1.9 million acres were primary forest.*

To identify deforestation hotspots consistently across this vast landscape, we conducted a kernel density analysis (see Methodology).

The Base Map shows the hotspots in yellow, orange and red, indicating areas with medium, high, and very high forest loss concentrations, respectively.

Next, we focus on five zones of interest (Zooms A-E) in Colombia, Brazil, Bolivia, and Peru. For all images, please click to enlarge.

*Forest Cover Loss: 5 acres per minute. Almost half (49%) occurred in Brazil, followed by Peru (20%), Colombia (20%), Bolivia (8%), and Ecuador (3%). see Annex.

**Primary Forest Loss: 3.5 acres per minute. Over half (53%) occurred in Brazil, followed by Peru (20%), Colombia (18%), Bolivia (7%), and Ecuador (2%). see Annex.

Colombia

The largest concentration of 2018 forest loss is in the northeast Colombian Amazon (494,000 acres). Out of this total, 11% (56,800 acres) occurred in national parks. National experts indicate that land grabbing has emerged as a leading direct driver of deforestation (Arenas 2018). See MAAP #97 for more information.

Zoom A shows the forest loss expanding towards western Chiribiquete National Park, including distinct deforestation in this protected area during 2018.

Zoom B shows the extensive 2018 deforestation (30,000 acres) within Tinigua National Park. A recent news report indicates that cattle ranching is one of the factors related to this deforestation.

Zoom A. Colombia-Chiribiquete. Data: Hansen/UMD/Google/USGS/NASA, SINAP, Planet, ESA
Zoom B. Colombia – Tinigua. Data: Hansen/UMD/Google/USGS/NASA, SINAP, Planet, ESA

Brazil (border with Bolivia)

Another important result is the contrast between northern Bolivia (Pando department) and adjacent side Brazil (states of Acre, Amazonas, and Rondônia). Zoom C shows several deforestation hotspots on the Brazilian side, while the Bolivian side is much more intact.

Zoom C. Brazil, Bolivia border. Data: Hansen/UMD/Google/USGS/NASA, ESA, RAISG

Bolivia

In Bolivia, the major forest loss hotspots are further south. Zoom D shows the recent deforestation (5,000 acres in 2018) due to agricultural activity associated with one of the first major Mennonite settlements in Beni department (Kopp 2015). The other Mennonite settlements are located further south.

Zoom D. Bolivia, Black River Mennonite settlement. Data: Hansen/UMD/Google/USGS/NASA, SERNAP, Planet

Peru

The Hansen data indicates over 200,000 acres of forest loss during 2018 in the Peruvian Amazon. One of the most important deforestation drivers, especially in southern Peru, is gold mining. We estimate 23,000 acres of gold mining deforestation during 2018 in the southern Peruvian Amazon (see MAAP #96).

Zoom E shows the most emblematic case of gold mining deforestation: the area known as La Pampa.

It is important to emphasize, however, that in February 2019 the Peruvian government launched “Operation Mercury 2019” (Operación Mercurio 2019), a multi-sectoral and comprehensive mega-operation aimed at eradicating illegal mining and associated crime in La Pampa, as well as promote development in the region.

Zoom D. Peru – La Pampa. Data: Hansen/UMD/Google/USGS/NASA, SERNAP, Planet

Annex

Annex. Forest cover and primary forest loss in the western Amazon.  Data: Hansen/UMD/Google/USGS/NASA, Global Forest Watch.

Methods

The 2018 forest loss data presented in this report were generated by the Global Land Analysis and Discovery (GLAD) laboratory at the University of Maryland (Hansen et al 2013) and presented by Global Forest Watch. Our study area is strictly what is presented in the Base Map: the areas within the Amazonian biogeographic boundary of the western Amazon.

Specifically, for our estimate of forest cover loss, we multiplied the annual “forest cover loss” data by the density percentage of the “tree cover” from the year 2000 (values >30%).

For our estimate of primary forest loss, we intersected the forest cover loss data with the additional dataset “primary humid tropical forests” as of 2001 (Turubanova et al 2018). For more details on this part of the methodology, see the Technical Blog from Global Forest Watch (Goldman and Weisse 2019).

All data were processed under the geographical coordinate system WGS 1984. To calculate the areas in metric units the UTM (Universal Transversal Mercator) projection was used: Peru and Ecuador 18 South, Colombia 18 North, Western Brazil 19 South and Bolivia 20 South.

Lastly, to identify the deforestation hotspots, we conducted a kernel density estimate. This type of analysis calculates the magnitude per unit area of a particular phenomenon, in this case forest cover loss. We conducted this analysis using the Kernel Density tool from Spatial Analyst Tool Box of ArcGIS. We used the following parameters:

Search Radius: 15000 layer units (meters)
Kernel Density Function: Quartic kernel function
Cell Size in the map: 200 x 200 meters (4 hectares)
Everything else was left to the default setting.

For the Base Map, we used the following concentration percentages: Medium: 10%-20%; High: 21%-35%; Very High: >35%.

References

Arenas M (2018) Acaparamiento de tierras: la herencia que recibe el nuevo gobierno de Colombia. Mongabay, 2 AGOSTO 2018. https://es.mongabay.com/2018/08/acaparamiento-de-tierras-colombia-estrategias-gobierno/

Goldman L, Weisse M (2019) Technical Blog: Global Forest Watch’s 2018 Data Update Explained. https://blog.globalforestwatch.org/data-and-research/blog-tecnico-explicacion-de-la-actualizacion-de-datos-de-2018-de-global-forest-watch

Hansen, M. C., P. V. Potapov, R. Moore, M. Hancher, S. A. Turubanova, A. Tyukavina, D. Thau, S. V. Stehman, S. J. Goetz, T. R. Loveland, A. Kommareddy, A. Egorov, L. Chini, C. O. Justice, and J. R. G. Townshend. 2013. “High-Resolution Global Maps of 21st-Century Forest Cover Change.” Science 342 (15 November): 850–53. Data available on-line from: http://earthenginepartners.appspot.com/science-2013-global-forest.

Kopp Ad (2015) Las colonias menonitas en Bolivia. Tierra. http://www.ftierra.org/index.php/publicacion/libro/147-las-colonias-menonitas-en-bolivia

Planet Team (2017). Planet Application Program Interface: In Space for Life on Earth. San Francisco, CA. https://api.planet.com

Turubanova S., Potapov P., Tyukavina, A., and Hansen M. (2018) Ongoing primary forest loss in Brazil, Democratic Republic of the Congo, and Indonesia. Environmental Research Letters  https://doi.org/10.1088/1748-9326/aacd1c 

Acknowledgements

We thank M. Terán (ACEAA), M. Weisse (GFW/WRI), A. Thieme (UMD), R. Catpo (ACCA) and A. Cóndor (ACCA) for helpful comments to this report.

Citation

Finer M, Mamani N (2019) Western Amazon – Deforestation Hotspots 2018 (a regional perspective). MAAP: 100.

MAAP Synthesis #3: Deforestation in the Andean Amazon (Trends, Hotspots, Drivers)

Posted on by dcadmin
Satellite image of the deforestation produced by United Cacao. Source: DigitalGlobe (Nextview)

MAAP, an initiative of the organization Amazon Conservation, uses cutting-edge satellite technology to monitor deforestation in near real-time in the megadiverse Andean Amazon (Peru, Colombia, Ecuador, and Bolivia).

The monitoring is based on 5 satellite systems: Landsat (NASA/USGS), Sentinel (European Space Agency), PeruSAT-1, and the companies Planet and DigitalGlobe. For more information about our innovative methodology, see this recent paper in Science Magazine.

Launched in 2015, MAAP has published nearly 100 high-impact reports on the major Amazonian deforestation issues of the day.

Here, we present our third annual synthesis report with the objective to concisely describe the bigger picture: Deforestation trends, patterns, hotspots and drivers across the Andean Amazon.

Our principal findings include:

Trends: Deforestation across the Andean Amazon has reached 4.2 million hectares (10.4 million acres) since 2001. Annual deforestation has been increasing in recent years, with a peak in 2017 (426,000 hectares). Peru has had the highest annual deforestation, followed by surging Colombia (in fact, Colombia surpassed Peru in 2017). The vast majority of the deforestation events are small-scale (‹5 hectares).

Hotspots: We present the first regional-scale deforestation hotspots map for the Andean Amazon, allowing for spatial comparisons between Peru, Colombia, and Ecuador.  We discuss six of the most important hotspots.

Drivers: We present MAAP Interactive, a dynamic map with detailed information on the major deforestation drivers: gold mining, agriculture (oil palm and cacao), cattle ranching, logging, and dams. Agriculture and ranching cause the most widespread impact across the region, while gold mining is most intense southern Peru.

Climate Change. We estimated the loss of 59 million metric tons of carbon in the Peruvian Amazon during the last five years (2013-17) due to forest loss. In contrast, we also show that protected areas and indigenous lands have safeguarded 3.17 billion metric tons of carbon.

I. Deforestation Trends

Image 1 shows forest loss trends in the Andean Amazon between 2001 and 2017.*  The left graph shows data by country, while the right graph shows data by forest loss event size.

Image 1. Annual forest loss by country and size. Data: Hansen/UMD/Google/USGS/NASA, UMD/GLAD, Global Forest Watch, MINAM/PNCB, RAISG.

Trends by Country

Over the past 17 years (2001-2017), deforestation has surpassed 4.2 million hectares (10.4 million acres) in the Andean Amazon (see green line). Of this total, 50% is Peru (2.1 million hectares/5.2 million acres), 41% Colombia (1.7 million hectares/4.27 million acres), and 9% Ecuador (887,000 acres/359,000 hectares). This analysis did not include Bolivia.

Since 2007, there has been an increasing deforestation trend, peaking during the past two years (2016-17). In fact, 2017 has the highest annual forest loss on record with 426,000 hectares (over one million acres), more than double the total forest loss in 2006.

Peru had the highest average annual Amazonian deforestation between 2009 and 2016. The past four years have the highest annual deforestation totals on record in the country, with peaks in 2014 (177,566 hectares/439,000 acres) and 2016 (164,662 hectares/406,888 acres). According to new data from the Peruvian Environment Ministry, there was an important decline in 2017 (155,914 hectares/385,272 acres), but it is still the fourth highest annual total on record.

There has been a surge of deforestation in Colombia during the past two years. Note that in 2017, Colombia surpassed Peru with a record high of 214,700 hectares (530,400 acres) deforested.

Deforestation is also increasing in Ecuador, with highs of 32,000 hectares (79,000 acres) in 2016 and 55,500 hectares (137,000) acres in 2017.

For context, Brazil has had an average deforestation loss rate of 639,403 hectares (1.58 million acres) over the past several years.

* Data: Colombia & Ecuador: Hansen/UMD/Google/USGS/NASA; Peru: MINAM/PNCB, UMD/GLAD. While this information includes natural forest loss events, it serves as our best estimate of deforestation resulting from anthropogenic causes.  It is estimated that the non-anthropic loss comprises approximately 3.5% of the total loss. Note that the analysis does not include Bolivia.

Trends by Size

The pattern related to the size of deforestation events in the Andean Amazon remained relatively consistent over the last 17 years. Most noteworthy: the vast majority (74%) of the deforestation events are small-scale (‹5 hectares). Only 2% of deforestation events are large-scale (>100 hectares). The remaining 24% are medium-scale (5-100 hectares).

These results are important for conservation efforts.  Addressing this complex situation – in which most of the deforestation events are small-scale – requires significantly more attention and resources.  In addition, while large-scale deforestation (usually associated with agro-industrial practices) is not that common, it nonetheless represents a serious latent threat, due to the fact that only a small number of agro-industrial projects (for example, oil palm) are able to rapidly destroy thousands of acres of primary forest.

II. Deforestation Hotspots

Image 2: Deforestation hotspots 2015-2017. Data: Hansen/UMD/Google/USGS/NASA.

We present the first regional-scale deforestation hotspots map across the Andean Amazon (Colombia, Ecuador, Peru).  Image 2 shows the results for the past three, 2015 – 2017.

The most critical zones (“high” deforestation density) are indicated in red. They include:

A. Central Peruvian Amazon: Over the last 10 years, this zone, located in the Ucayali and Huánuco regions, has consistently had one of the largest concentrations of deforestation in Peru (Inset A).  Its principal drivers include oil palm and cattle grazing.

B. Southern Peruvian Amazon: This zone, located in the Madre de Dios region, is impacted by gold mining (Inset B1), and increasingly by small- and medium-scale agriculture along the Interoceanic Highway (Inset B2).

C. Central Peruvian Amazon: A new oil palm plantation located in the San Martín region has been identified as a recent large-scale deforestation event in this zone (Inset C).

D. Southwestern Colombian Amazon: Cattle grazing is the principal deforestation driver documented in this zone, located in the departments of Caquetá and Putumayo (Inset D).

E. Northern Colombian Amazon: There is expanding deforestation along a new road in this zone, located in the department of Guaviare (Inset E).

F. Northern Ecuadoran Amazon: This zone is located in the Orellana province, where small- and medium-scale agriculture, including oil palm, is the principal driver of deforestation (Inset F).

 

 

III. Drivers of Deforestation     

MAAP Interactive (screenshot)

One of the main objectives of MAAP is to improve the availability of precise and up-to-date information regarding the current drivers (causes) of deforestation in the Andean Amazon.  Indeed, one of our most important advances has been the use of high-resolution imagery to identify current deforestation drivers.

In order to improve the analysis and understanding of the identified drivers, we have created an Interactive Map that displays the spatial location of each driver associated with every MAAP report.  An important characteristic of this map is the ability to filter the data by driver, by selecting the boxes of interest.

Image 3 shows a screenshot of the Interactive Map.  Note that it contains detailed information on these principal drivers: gold mining, oil palm, cacao, small-scale agriculture, cattle pasture, logging roads, and dams.  It also includes natural causes such as floods, forest fires, and blowdowns.  In addition, it highlights deforestation events in protected areas.

Below, we discuss the principal drivers of deforestation and degradation in greater detail.

 

 

 

 

Agriculture – oil palm, cacao, and other crops

Image 4: Interactive Map, agriculture. Data: MAAP.

Image 4 shows the results of the interactive map when applying the agriculture-related filters.

Legend:
Oil palm (bright green)
Cacao (brown)
Other crops (dark green)

Agricultural activity is one of the principal causes of deforestation in the Andean Amazon.

The majority of agriculture-related deforestation is caused by small- and medium-scale plantations (‹50 hectares).

Deforestation for large-scale, agro-industrial plantations is much less common, but represents a critical latent threat.

 

 

 

 

 

Large-scale Agriculture

We have documented five major deforestation events produced by large-scale plantations since 2007:  four of these occurred in Peru (three of which are related to oil palm and one to cacao) and one in Bolivia (resulting from sugar cane plantations).

First, between 2007 and 2011, two large-scale oil palm plantations caused the deforestation of 7,000 hectares on the border between Loreto and San Martín (MAAP #16).  Subsequent plantations in the surrounding area caused the additional deforestation of 9,800 hectares.

It is importnat to note that the Peruvian company Grupo Palmas is now working towards a zero deforestation value chain and has a new sustainability policy (see Case C of MAAP #64).

Next, between 2012 and 2015, two other large-scale oil palm plantations deforested 12,000 hectares in Ucayali  (MAAP #4, MAAP #41).

Between 2013 and 2015, the company United Cacao deforested 2,380 hectares for cacao plantations in Loreto (MAAP #9, MAAP #13, MAAP #27, MAAP #35).

Deforestation from large-scale agriculture decreased in Peru between 2016 and 2017, but there was one notable event: an oil palm plantation of 740 hectares in San Martín (MAAP #78).

Another notable case of deforestation related to large-scale agriculture has been occurring in Bolivia, where a new sugarcane plantation has caused the deforestation of more than 2,500 hectares in the department of La Paz.

Additionally, we found three new zones in Peru characterized by the deforestation pattern produced by the construction of organized access roads which have the potential of becoming large-scale agriculture areas (MAAP #69).

Small and Medium-scale Agriculture

Deforestation caused by small- and medium-scale agriculture is much more widespread, but it is often difficult to identify the driver from satellite imagery.

We have identified some specific cases of oil palm in Huánuco, Ucayali, Loreto, and San Martín (MAAP #48, MAAP #26, MAAP #16).

Cacao and papaya are emerging drivers in Madre de Dios.  We have documented cacao deforestation along the Las Piedras River (MAAP #23, MAAP #40) and papaya along the Interoceanic Highway (MAAP #42).

Corn and rice cultivation appear to be turning the area around the town of Iberia into a deforestation hotspot (MAAP #28).  In other cases, we have documented deforestation resulting from small- and medium-scale agriculture, though it has not been possible to identify the type of crop (MAAP #75, MAAP #78).

Additionally, small-scale agriculture is possibly a determining factor in the forest fires that degrade the Amazon during the dry season (MAAP #45, MAAP #47).

The cultivation of illicit coca is a cause of deforestation in some areas of Peru and Colombia.  For example, in southern Peru, the cultivation of coca is generating deforestation within the Bahuaja Sonene National Park and its surrounding areas.

Cattle Ranching

Image 5: Interactive Map, cattle ranching. Data: MAAP.

By analyzing high-resolution satellite imagery, we have developed a methodology for identifying areas deforestated by cattle ranching.*

Image 5 shows the results of the Interactive Map when applying the “Cattle pasture” filter, indicating the documented examples in Peru and Colombia.

Legend:
Cattle ranching (orange)

Cattle ranching is the principal driver of deforestation in the central Peruvian Amazon (MAAP #26, MAAP #37, MAAP #45, MAAP #78). We also identified recent deforestation from cattle ranching in northeastern Peru (MAAP #78).

In the Colombian Amazon, cattle ranching is one the primary direct drivers in the country’s most intense deforestation hotspots (MAAP #63, MAAP #77).

* Immediately following a major deforestation event, the landscape of felled trees is similar for both agriculture and cattle pasture.  However, by studying an archive of images and going back in time to analyze older deforestation cases, it is possible to distinguish between the drivers.  For example, after one or two years, agriculture and cattle pasture appear very different in the images. Ther former tends to have organized rows of new plantings, while the latter is mostly grassland.

 

 

 

Gold Mining

Image 6: Interactive Map, gold mining. Data: MAAP.

Image 6 shows the results of the Interactive Map when applying the “Gold mining” filter.

Legend:
Gold Mining (yellow)
*With dot indicates within protected area

The area that has been most impacted by gold mining is clearly the southern Peruvian Amazon, where we estimate the total deforestation of more than 63,800 hectares. Of this, at least 7,000 hectares have been lost since 2013.  The two most critical zones are La Pampa and Alto Malinowski in Madre de Dios (MAAP #87, MAAP #75, MAAP #79).  Another critical area exists in Cusco in the buffer zone of the Amarakaeri Communal Reserve, where mining deforestation is now less than one kilometer from the boundary of the protected area (MAAP #71).

It is important to highlight two important cases in which the Peruvian government has taken effective actions to halt illegal mining within protected areas (MAAP #64).  In September 2015, illegal miners invaded Tambopata National Reserve and deforested 550 hectares over the course of a two-year period.  At the end of 2016, the government intensified its interventions and the invasion was halted in 2017. In regards to Amarakaeri Communal Reserve, in June 2015 we revealed the mining invasion deforestation of 11 hectares.  Over the course of the following weeks, SERNANP and ECA Amarakaeri implemented measures and rapidly halted the illegal activity.

Other small gold-mining fronts are emerging in the northern and central Peruvian Amazon (MAAP #45, MAAP #49).

In addition, we have also documented deforestation linked to illegal gold-mining activities in the Puinawai National Park in the Colombian Amazon.

Logging

Image 7: Interactive Map, logging roads. Data: MAAP.

In MAAP #85 we proposed a new tool to address illegal logging in the Peruvian Amazon: utilize satellite imagery to monitor construction of logging roads in near real-time.

Image 7 shows the results of the Interactive Map when applying the “Logging roads” filter.

Legend:
Logging Road (purple)

We estimate that 2,200 kilometers of forest roads have been constructed in the Peruvian Amazon during the last three years (2015-2017).  The roads are concentrated in southern Loreto, Ucayali, and northwestern Madre de Dios.

 

 

 

 

 

 

Roads

Image 8: Interactive map, roads. Data: MAAP.

It has been well-documented that roads are one of the most important drivers of deforestation in the Amazon, particularly due to the fact that they facilitate human access and activities related to agriculture, cattle ranching, mining, and logging.

Image 8 shows the results of the Interactive Map when applying the “Roads” filter.

Legend:
Road (gray)

We have analyzed two controversial proposed roads in Madre de Dios, Peru.

The Nuevo Edén – Boca Manu – Boca Colorado road would traverse the buffer zone of two protected areas: Amarakaeri Communal Reserve and Manu National Park (MAAP #29).

The other, the Puerto Esperanza-Iñapari road, would traverse the Purús National Park and threaten the territory of the indigenous peoples in voluntary isolation who live in this remote area (MAAP #76).

 

 

 

 

Hydroelectric dams

Image 9 shows the results of the Interactive Map when applying the “Dams” filter.

Legend:
Hydroelectric Dam (light blue)

To date, we have analyzed three hydroelectric dams located in Brazil.  We have documented the loss of 36,100 hectares of forest associated with flooding produced by two dams (San Antonio and Jirau) on the Madeira River near the border with Bolivia (MAAP #34).  We also analyzed the controversial Belo Monte hydroelectrical complex located on the Xingú River, adn estimate that 19,880 hectares of land have been flooded. According to the imagery, this land is a combination of forested areas and agricultural areas (MAAP #66).

Additionally, we show a very high-resolution image of the exact location of the proposed Chadín-2 hydroelectric dam on the Marañón River in Peru (MAAP #80).

Hydrocarbon (oil and gas)

Image 10: Interactive map, hidrocarbon. Data: MAAP.

Image 10 shows the results of the Interactive Map when applying the “Hydrocarbon“ filter.

Legend:
Hydrocarbon (black)

Our first report on this sector focused on Yasuní National Park in the Ecuadorian Amazon.  We documented the direct and indirect deforestation amounts of 417 hectares (MAAP #82).

We also show the location of recent deforestation in two hydrocarbon block in Peru: Block 67 in the north and Blocks 57 in the south.

 

 

 

 

 

 

 

Climate Change

Tropical forests, especially the Amazon, sequester huge amounts of carbon, one of the main greenhouse gases driving climate change.

In MAAP #81, we estimated the loss of 59 million metric tons of carbon in the Peruian Amazon during the last five years (2013-17) due to forest loss, especially deforestation from mining and agricultural activities. This finding reveals that forest loss represents nearly half (47%) of Peru’s annual carbon emissions, including from burning fossil fuels.

In contrast, in MAAP #83 we show that protected areas and indigenous lands have safeguarded 3.17 billion metric tons of carbon, as of 2017. That is the equivalent to 2.5 years of carbon emissions from the United States.

The breakdown of results are:
1.85 billion tons safeguarded in the Peruvian national protected areas system;
1.15 billion tons safeguarded in titled native community lands; and
309.7 million tons safeguarded in Territorial Reserves for indigenous peoples in voluntary isolation.

Citation

Finer M, Mamani N (2018) Deforestation in the Andean Amazon (Trends, Hotspots, Drivers). MAAP Synthesis #3.

MAAP #76: Proposed Road would cross Primary Forest along Peru-Brazil Border

Posted on by dcadmin
Image 76a. Base Map. Data: Mosaic of 16 images from Sentinel-2/ESA, July 2017

In December 2017, the Peruvian Congress approved a bill that declared it in the national interest to construct new roads in the border zone of Ucayali region, which shares a remote border with Brazil.

The main proposed road in this border area would cover 172 miles and connect the towns of Puerto Esperanza and Iñapari, in the Ucayali and Madre de Dios regions, respectively. Image 76a, a mosaic of satellite images from July 2017, illustrates just how remote and intact is the area surrounding the proposed road route.

Indigenous organizations and the Ministry of Culture have warned that the road would have major impacts on the indigenous peoples in voluntary isolation that are documented to inhabit parts of this remote area.

In this report, we add new information that complements the evaluation of possible impacts by calculating how much primary forest would be threatened as a result of road construction. We found that around 680,000 acres (275,00 hectares) of primary forest are at risk. Much of this area is within protected areas and a reserve for isolated indigenous groups.

 

Primary Forest

Image 76b. Data: GLCF/GSFC 2014, Hansen/UMD/Google/USGS/NASA, UMD/GLAD, PNCB/MINAM, UAC/ProPurús, SERNANP

We generated a primary forest layer based on existing satellite-based forest cover and forest loss data (see Methodology section for more details). We define primary forest as areas with intact forest cover dating back to the earliest available satellite-based data, 1990 in this case.

Image 76b shows the major results:

  • Virtually the entire route (172 miles; 277 km) crosses primary forest (dark green). Note the proliferation of forest roads in recent years around Iñapari (red lines).
  • The road would cross 3 critical protected areas and indigenous reserves: Madre de Dios Territorial Reserve, Alto Purús National Park, and Purús Communal Reserve.

 

 

 

 

 

 

Primary Forest at Risk

Imagen 76c. GLCF/GSFC 2014, Hansen/UMD/Google/USGS/NASA, UMD/GLAD, PNCB/MINAM, UAC/ProPurús

The Interoceanic Highway, the main existing road in the area, has experienced substantial deforestation within 5 km* along the length of its route (Image 76c).

Using this estimate of impact range (10 km), we calculated that at least 274,727 hectares of primary forest would be at risk if this road is constructed.

*We estimate that approximately 80% of forest loss has occurred in a 5 km radius on both sides of the Interoceanic highway.

 

 

 

 

 

 

 

 

Methodology

To generate our primary forest layer, we combined three satellite-based data sources. As baseline, we used data from the Global Land Cover Facility (2014), which identifies forest cover as of 1990. We also used this dataset to remove areas with detected forest cover change between 1990 and 2000. Next, we removed areas with detected forest loss between 2001 – 2017 identified by Hansen/UMD/Google/USGS/NASA (Hansen et al 2013) and early warning data from GLAD alerts and the National Program of Forest Conservation of the Peruvian Environment Ministry (PNBC-MINAM). As a result, combining all datasets, this methodology defines primary forest as area with intact forest from the first available satellite-based data, 1990, until 2017.

Global Land Cover Facility (GLCF) and Goddard Space Flight Center (GSFC). 2014. GLCF Forest Cover Change 2000, 2005, Global Land Cover Facility,University of Maryland, College Park.

Hansen MC et al. 2013. High-Resolution Global Maps of 21st-Century Forest Cover Change. Science 342: 850–53.

Citation

Finer M, Novoa S (2018) Proposed Road would cross Primary Forest along Peru-Brazil Border. MAAP: 76.

MAAP Interactive: Deforestation Drivers in the Andean Amazon

Posted on by dcadmin

Since its launch in April 2015, MAAP has published over 70 reports related to deforestation (and natural forest loss) in the Andean Amazon. We have thus far focused on Peru, with several reports in Colombia and Brazil as well.

These reports are meant to be case studies of the most important and urgent deforestation events. We often use forest loss alerts (known as GLAD) to guide us, and satellite imagery (from Planet and DigitalGlobe) to identify the deforestation driver.

Here we present an interactive map highlighting the drivers identified in all published MAAP reports. These drivers include gold mining, agriculture (e.g. oil palm and cacao), cattle pasture, roads, and dams (see icon legend below map). We also include natural causes such as floods and blowdowns (fire included under agriculture since most human caused). Furthermore, we highlight deforestation events within protected areas. Note that you can filter by driver by checking boxes of interest.

We hope the result is one of the most detailed and up-todate resources on patterns and drivers of deforestation in the Andean Amazon. Over the coming year we will continue to focus on Peru and Colombia, and begin to include Ecuador and Bolivia as well.

To view the interactive map, please visit:

MAAP Interactive: Deforestation Drivers in the Andean Amazon
https://www.maapprogram.org/interactive/

For more information on patterns and drivers of deforestation in the Peruvian Amazon, see our latest Synthesis report 

MAAP #66: Satellite Images of Belo Monte Dam Project (Brazil)

Posted on by dcadmin
Image 66a. Red circle indicates dam project area.

The Belo Monte hydroelectric dam complex, located on the Xingu River in the state of Para in the eastern Brazilian Amazon (see Image 66a), has been controversial since its inception over 15 years ago, due to both environmental and social concerns related to building and operating one of the largest dams in the world in a sensitive environment.

The dam has recently become operational, providing an opportunity to evaluate initial impacts.

The objective of this article is to present satellite imagery, including a time series from 2011 to 2017, that provides insight into major ecological impacts of the hydroelectric dam project.

 

 

 


Despite legal challenges and strong opposition from impacted indigenous groups, construction of Belo Monte began in 2011 and the first turbines became operational in early 2017. Image 66b shows a direct comparison of before (left panel, July 2011) and after (right panel, July 2017) dam construction.

Image 66b. NASA/USGS

The dam is in fact a complex: the main dam (red circle) on the Xingu River creates a main reservoir (blue circle); a canal diverts much (up to 80%) of the river’s flow from the main reservoir to the canal reservoir (yellow circle), which feeds the turbines generating the electricity. As a result, downstream of the main dam is left with a much reduced flow (20%) for a stretch of 100 km. This reduced flow stretch, known as the Xingu River’s “Big Bend,” is home to two indigenous peoples (Arara and Juruna). The reference points in the images show these four areas of the complex across time, including before construction.

Satellite Image Time Series

Image 66c. Data: NASA/USGS

Image 66c is a GIF showing a satellite (Landsat) imagery time series of the project impact area from July 2011 through May 2017. July 2011 serves as the pre-project baseline before the start of construction. By July 2015, construction of the main dam and canal is well under way. By January 2016, the main dam has closed, forming both the main and canal reservoirs. August 2016 provides a nearly cloudless view of the dam complex, including how dry the downstream section becomes. July 2017 represents the most recent cloud-free Landsat image.

In the most recent images, note the negative impact on local fisheries: flooding of river islands, rock outcrops, and seasonally flooded forests in the Main Reservoir that were important fish habitat; and reduced water flow along the “Big Bend” below the Main Dam, also important fish habitat.

Flooding Estimate

Based on an analysis of the Landsat images, we estimate the flooding of 48,960 acres (19,880 hectares) of land that, according to the imagery, appeared to be a mix of forest and agriculture (Image 66d). In other words, some of the flooded area was previously degraded.

Image 66d. Data: NASA/USGS, MAAP

Damming of the Xingu River

Image 66e shows, in high resolution (50 cm), the drastic change at the dam site between July 2010 (left panel) and June 2017 (right panel). The July 2010 image, which serves as the pre-construction baseline, shows the free-flowing Xingu River, whereas the June 2017 image shows the impact of the main dam and main reservoir. Image 66f is a GIF showing, in striking detail, the construction of the main dam and formation of the main reservoir between 2010 and 2017.

Image 66e. Data: DigitalGlobe (via ACT), Airbus (via Apollo Mapping)
Image 66f. Data: DigitalGlobe (Nextview), DigitalGlobe (via ACT), Airbus (via Apollo Mapping)

We thank International Rivers and Amazon Watch for reviewing earlier drafts of this article and providing crucial comments.

Finer M, Olexy T, Scott A (2017) Satellite Images of Controversial Belo Monte Dam Project. MAAP: 66.

MAAP #34: New Dams on the Madeira River in Brazil Cause Forest Flooding

Posted on by dcadmin

The Amazon lowlands have been connected to the Andes Mountains for millions of years by only six major rivers: the Caqueta, Madeira, Maranon, Napo, Putumayo, and Ucayali* (see Image 34a). This intimate connection allows rich Andean nutrients to fuel the Amazon floodplain and enables long-distance catfish migration between feeding grounds in the lowlands and spawning grounds in the highlands.

Image 34a. Data: Esri, DigitalGlobe, GeoEye, Earthstar Geographics, CNES/Airbus DS, USDA, AEX, Getmapping, Aerogrid, IGN, IGP, swisstopo
Image 34a. Data: Esri, DigitalGlobe, GeoEye, Earthstar Geographics, CNES/Airbus DS, USDA, AEX, Getmapping, Aerogrid, IGN, IGP, swisstopo

However, one of these six major Andean tributaries has recently been dammed on its main channel: the Madeira River in western Brazil (See Inset A). The Santo Antônio dam was completed in 2011, followed by the upstream Jirau dam in 2013.

Note in Image 34a that these dams are are located downstream of the Madre de Dios River in southern Peru. Thus, major ecological impacts — such as blocking the route of migratory catfish**— are also very relevant to Peru.

Here in MAAP #34, we describe the forest loss—over 36,100 hectares—associated with the flooding caused by these two dams (with a focus on the Jirau dam).

Zoom A: Forest Loss due to Flooding

Image 34b shows the forest loss due to flooding immediately upstream of the Jirau dam. As of 2015, the total flooded area for both dams is 36,139 hectares (89,301 acres). Major flooding was first detected in 2010, rose substantially in 2011-12, and peaked in 2014.

According to Fearnside 2014, although much of the forest along the Madeira is seasonally flooded, it dies when permanently flooded.*** Therefore, the flooded area is an appropriate measure of forest loss.

Further below, we show a series of satellite images of the areas indicated by Inset B (see Images 34c-e) and Inset C (see Image 34f).

Image 34b. Flooding-related forest loss along the Upper Madeira River. Data: USGS, CLASlite, Hansen/UMD/Google/USGS/NASA.
Image 34b. Flooding-related forest loss along the Upper Madeira River. Data: USGS, CLASlite, Hansen/UMD/Google/USGS/NASA.

Zoom B: Flooding Immediately Upstream Jirau Dam

Image 34c shows the flooding immediately upstream of the Jirau dam between 2011 (left panel) and 2015 (right panel). The red dot is a point of reference that indicates the same place in both images. Below, we show high-resolution images of the areas indicated by Insets B1 and B2.

zoomB_rnd2
Image 34c shows the flooding immediately upstream of the Jirau dam between 2011(left panel) and 2015 (right panel).

Zooms B1 and B2: Jirau Dam and Flooding

Image 34d shows a high-resolution view of the Jirau dam in July 2015. Image 34e shows a high-resolution view of a portion of the flooded area immediately upstream of the Jirau dam in August 2015. The red dot is a point of reference that indicates the same place in both panels.

b1_rnd2
Image 34d. High-resolution view of the Jirau dam. Data: WorldView-2 from Digital Globe (NextView).
zoomb2_rnd2
Image 34e: High-resolution view of flooded area immediately upstream of the Jirau dam. Data: WorldView-2 from Digital Globe (NextView).

Zoom C: Flooding Further Upstream of Jirau Dam

Image 34f shows the flooding further upstream of the Jirau dam between 2011 (left panel) and 2015 (right panel). The red dot is a point of reference that indicates the same point in both images.

zoomC_rnd2
Image 34f. Forest flooding further upstream of the Jirau dam between 2011 (left panel) and 2015 (right panel). Data: USGS

References

*Finer M, Jenkins CN (2012) Proliferation of Hydroelectric Dams in the Andean Amazon and Implications for Andes-Amazon Connectivity. PLOS ONE: 7(4): e35126. Link: http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0035126

**Duponchelle F et al (2016) Trans-Amazonian natal homing in giant catfish. J. Appl. Ecol. http://doi.org/bd45

***Fearnside PM (2014) Impacts of Brazil’s Madeira River dams: Unlearned lessons for hydroelectric development in Amazonia. Environmental Science & Policy 38: 164-172.

Citation

Finer M, Olexy T (2015) New Dams on the Madeira River (Brazil) Cause Forest Flooding. MAAP: 34.