Service Areas

SERVIR-Amazonia develops a diverse collection of user-tailored geospatial services that use Earth observations and NASA data to inform resilient development in the Amazon in four thematic areas

Ecosystem Management

Earth observation and geospatial information and analysis are critical for ecosystem monitoring and management. In Amazonia, geospatial information is regularly used to monitor terrestrial ecosystems, such as deforestation early alert systems. However the use of geospatial information to monitor water resources and aquatic/wetland ecosystems is much less developed. Geospatial and Earth observation services can provide information for planning and management at multiple scales. The relative importance of certain areas for biodiversity and ecosystem integrity, analysis of potential impacts of human activities, and options for sound management and conservation decisions are examples of information needs for key aquatic ecosystems affected by seasonal and longer-term changes in hydro-meteorological cycles, land use change, and development activities such as hydropower, roads, mining and others.
Contact: Karis Tenneson ktenneson@sig-gis.com

Activities

Naiara Pinto / Jet Propulsion Laboratory

Active remote sensing is critical for characterizing high cloud cover / high biomass tropical ecosystems such as the Amazon. This project will help Latin American governments incorporate free radar and lidar datasets into existing monitoring programs. Pinto’s team proposes to co-develop and implement an open-source land cover mapping system with the SERVIR-Amazonia Hub in support of the Land Cover and Land Use Change and Ecosystems thematic service area. The collaborative framework is focused on land cover mapping at the forest/agriculture interface through characterization of disturbance gradients and identification of plantation sites. At the same time, the project will prepare Latin American researchers and decision makers for leveraging the high volume of free datasets produced by NASA’s NISAR and GEDI missions.

The stakeholders are farmers enrolled in sustainable landscape initiatives as well as decision makers in NGOs and government that support sustainable agriculture practices. The project team plans to leverage subnational initiatives in Ucayali – Peru, Caquetá – Colombia, and Pará – Brazil building on collaborations with local communities and monitoring agencies such as MINAM and IMAFLORA, producing lessons and workflows that can be adapted to other Amazonian sites.

Photo: Site where Pinto’s team will be implementing the project

Stephanie Spera / University of Richmond

The Amazon rainforest provides ecosystem services to 33 million people, including 1.5 million Indigenous people from 385 different groups, living within the biome boundary. The goals of this project are to understand how forest degradation, deforestation, and road building affect the ecosystem services provided by the hydrologic cycle in the Southwestern Amazon and to develop data and tools to improve water management in the region.

Spera’s team will work in the areas of Ucayali, Peru and Acre, Brazil to characterize changes in forest cover using remotely-sensed data and fieldwork, and attribute these changes in forest cover to localized changes in evapotranspiration (ET), soil moisture, humidity, and surface temperature. The effect-radius of these changes in forest cover (how far away these changes are felt) will be determined, and maps generated that highlight areas that have undergone changes in microclimatology and land-use. The team will also create a statistics-based tool that will allow users to analyze tradeoffs between land cover change and ecosystem services through scenarios modelling. This information will be shared through trainings, fieldwork, and workshops with partner communities and universities.

The project will emphasize capacity strengthening of local indigenous and non-indigenous groups, as well as colleagues at CIAT and SERVIR Amazonia-Hub partner institutions such as ACCA-MAAP, and colleagues and students with current partners in the region. Updated maps of transportation infrastructure, forest cover change, and the disturbance of ecosystem services (temperature, ET, precipitation, and soil moisture) will be useful to locals and planners. These maps can also provide a baseline for future climate modelling analyses on feedbacks between forest disturbance and regional climate. Moreover, involving locals will allow indigenous and non-indigenous groups to participate in the creation of knowledge in the region.

Figure: South America, with the Amazon basin outlined in red, Ucayali and Acre highlighted in grey. B: Blue hatched lines designate indigenous reserves; green areas are conservation areas. The Purus River (P), Juruá River (J) and Ucayali River (U) are also highlighted. C: Tributary of the Ucayali River (pink box, B). Colors represent forest loss between 2001-2005 (purple); 2005-2010 (green); 2011-2015 (yellow) and 2016-2017 (red) (Hansen et al. 2013). 

Integrate mapped deforestation areas from multiple algorithms and generate updated area estimates of forest loss

Problem: Ecuador is home to over 12 million hectares of native forests and ecosystems. However, these are experiencing losses due to pressures from land-use changes. Ecuador implemented an incentive policy, the Socio Bosque Program, to support the dual objectives of simultaneously balancing development and forest conservation. The country needs timely, robust monitoring products regarding forest cover, deforestation, and degradation to allocate incentives for partners’ properties. This information is also required to meet their commitments to international agreements and to take advantage of opportunities from international donors and other agencies providing payments for climate change mitigation.

Photo credit: Ministry of Environment of Ecuador

Solution: Currently the Forest Conservation Program of the Ministry of the Environment (MAE) and REDD+ Program have been testing deforestation monitoring algorithms that integrate information from synthetic aperture radar and optical imagery. The service will integrate mapped deforestation areas from multiple algorithms and generate updated area estimates of forest loss. The Ministry of Environment’s (MAE) team will work with the SERVIR-Amazonia team, led the Spatial Informatics Group (SIG). The service team will facilitate a workshop to assess the production of deforestation layers and produce an integrated deforestation product along with a document outlining the integration methodology. The team will use Collect Earth Online to build a reference data set of locations and their corresponding land use, to assess maps.  The Program will build on the System for Earth Observations, Data Access, Processing & Analysis for Land Monitoring (SEPAL) platform.

Optimize the workflow through automation

Problem: The Peruvian Amazon is facing one of the biggest challenges to reduce deforestation rates, which in recent years are over 150,000 hectares  (ha) per year. Gold mining deforestation in Peru caused the forest loss of more than 96,000 ha of primary forest in the last 30 years, hitting historic highs in both 2017 and 2018 when the activity reached several protected areas including the Tambopata National Reserve (RNTB) and the Amarakaeri Communal Reserve (RCA).

In February 2019, the Peruvian government started an unprecedented operation aimed at eradicating illegal gold mining in the most impacted area of the country (La Pampa), located in the buffer zone of the RNTB. The Ministry of Environment (MINAM) has been developing information on potential locations of illegal gold mining. However, the information arrives often too late and lacks precision resulting in continued illegal mining activities and failure to prosecute offenders. In addition, extensive cloud cover during the wet season limits the capacity for optical remote-sensing products to provide appropriate data for near real-time detection.

Credit: ACCA, proposed conservation corridors, from Manu National Park to Tambopata National Reserve

 

Solution: SERVIR-Amazonia partner Conservación Amazónica (ACCA) has extensive experience working with the Peruvian government to provide actionable insights on illegal gold mining, supported by funding from NORAD. ACCA produces a brief report including geolocation of potential mining activities in priority regions and high-resolution satellite data. ACCA is currently working with MINAM to compare the performance of various algorithms to detect mining. ACCA’s current workflow is well advanced combining multiple remote sensing products (optical and SAR), using Google Earth Engine and validating through high-resolution data. The intervention from SERVIR-Amazonia will support enhancements to optimize the workflow through automation, investigate opportunities for using machine learning to identify reservoirs associated with mining activities and improve the timeliness and content of the final products delivered to government agencies. The Hub service team will be led by ACCA with support from SIG, and the service will be co-designed and co-developed with MINAM.

Read our blog post: Reducing illegal gold mining in the tropical forests of Ghana and Peru: A forthcoming collaboration across the Atlantic

 

Build on methods described in the SAR Handbook to provide additional tools to MAAP and Terra-I

Problem: In Peru, Colombia and other places in the tropics, the lack of detailed information on forest change is largely due to high cloud coverage and limited detection capacity with optical imagery. Existing efforts such as the MAAP (produced by ACCA) and Terra-I (produced by CIAT) are working across sectors to improve forest change monitoring. They work with government agencies and civil society to provide much-needed information on forest changes. These efforts can be improved by incorporating new methods into their existing toolsets.

Solution: Recent advances in radar remote sensing would add an important data source to the MAAP and Terra-I initiatives. The service will build on methods described in the SAR Handbook to provide additional tools to MAAP and Terra-I. A SCO-sponsored Subject Matter Expert and SERVIR-Amazonia staff will establish the service through three training sessions, two in Peru and one in Colombia, directed to SERVIR-Amazonia Consortium partners. The intended outcome of the service is to improve technical capacity predominantly with Hub partners.

Map the extent and structure of mangrove forests along the coast of Guyana automating the analysis of radar and optical imagery

Problem: Mangrove forests make up much of the coast of Guyana. Because many areas of coastal Guyana are is below sea level, the mangroves serve as a natural barrier protecting inland ecosystems from flooding and saltwater intrusion. However, these forests are under threat from rising sea levels, growing population and land-use change. Accurate mapping of the extent of mangroves and their structure would allow the nation to develop plans and programs to protect them. Specifically the Guyana Forestry Commission (GFC) needs remote-sensing analysis to measure and monitor the spatial and temporal trends in mangrove growth, deforestation and degradation.

Solution: This service will bring remote-sensing resources, primarily Synthetic Aperture Radar (SAR) but also others, to bear on mapping the extent and structure of mangrove forests along the coast of Guyana. Professionals in Guyana together with SERVIR-Amazonia staff will build a platform for automating the analysis of radar and optical imagery going back several years and setting a year-2020 baseline for future analysis. The service will make mangrove-related land-use change transparent and the resulting analysis publicly available for use by government and civil society. The key input for starting this service and bringing it to fruition will be a subject matter expert funded by the NASA SERVIR Science Coordination Office over a six-month period at the end of 2019 and beginning of 2020. Two five-day in-person training sessions will be held at the training facilities of the University of Guyana in Georgetown.

Drought and Fire Risk

Extended dry seasons and drought are a threat to agricultural food systems and create a different kind of disaster, including that of fire, one of the greatest threats to tropical forests. The Amazon has a long history of fire detection work, including Brazil’s fire monitoring system, which provides fire hotspot data and risk analysis for all of South America. However, drought/fire forecasting tools and their impacts on reducing food insecurity in the region are produced at coarse scales not always applicable for subnational management needs, and capacity at these levels is also sometimes lacking. Geospatial data and modeling, and associated capacity building, can create decision support services related to drought and fire risk that are relevant on regional, national, and subnational levels and that allow stakeholders to allocate resources more efficiently and effectively.                                                                                                                Contact: Kátia Fernandes kdfernan@uark.edu

Activities

Douglas Morton / Goddard Space Flight Center

Rainfall timing and distribution drives changes in ecosystems and how people adapt to nature in the Amazon. Water scarcity or overabundance may affect river navigation, agricultural food security or sustainable development problems at multiple scales. Droughts are particularly harmful to human and natural systems especially when they increase risks for destructive fires. Amazon droughts can be predicted in the future by analyzing sea surface temperatures in both the Atlantic and Pacific oceans. Droughts can drive destructive wildfires. The capacity to predict droughts can help us better understand vulnerability to fires and fire risk across the basin.

This project seeks to provide information for mitigating negative impacts of drought and fire on forest and agriculture in the Amazon basin. The project will evaluate drought conditions at temporal and spatial resolution, which is needed for management by government agencies, the private sector and civil society. The project will evaluate drought conditions at temporal and spatial resolution is needed for management by government agencies, the private sector and civil society. The effort includes three applied science tasks that will be developed in partnership with SERVIR-Amazonia scientist and stakeholders. First, the project will build on existing work to improve seasonal fire forecasting with a focus on better temporal and spatial resolution in the prediction of fire vulnerability. Second, the project will detect small fires and understory fires using several satellite platforms, with a focus on information for improved management. These products will fill a current gap by detecting fires that current global algorithms often miss. The second component of work will help stakeholders characterize fires and quantify their negative impacts on ecosystem services. Finally the project will use fire forecasting models for developing dry season metrics and for predicting how these metrics change.

Water Resource Management and Hydro-climatic Disasters

Both flooding and drought in Amazonia can cause multiple severe impacts on human and natural systems, and both can be exacerbated by climate variability, land use change, and poor land management practices. In the Amazon, many areas flood seasonally. As such, variation in river levels and accompanying floods or lack of flooding can dramatically affect population centers, agriculture, food systems and transportation, as well as natural ecosystems. Improvements in the timeliness of flood and low flow warnings, using local river- and stream-flow observations combined with remotely sensed precipitation, can improve disaster response and management/planning tools.
Contact:  Glenn Hyman ghyman@sig-gis.com 

Activities

Jim Nelson / Brigham Young University

The Amazon Basin is a critical component of the global climate system. In recent years, extreme hydrologic events such as floods and droughts have intensified and wreaked havoc on the communities and environments of the world’s largest river basin. This project will bring together well-established and newly developed free and open sources web tools to empower key stakeholders in the region to co-develop actionable water resources information. Ultimately, the project aims to support water security that provides adequate water quantity for livelihoods, ecosystems, and production while maintaining acceptable water-related risks to people, environments, and economies. This information enables decision-making that supports flood early warning and disaster risk reduction, as well as enhanced management for drinking water supply, improved sanitation and wastewater, reliable energy security, drought planning and response, climate change resilience, environmental restoration and more.1

Nelson’s team at BYU’s Hydroinformatics Lab has several years of experience working in the region, with several important stakeholders including Colombia’s Instituto de Hidrología, Meteorología y Estudios Ambientales (IDEAM), Peru’s Servicio Nacional de Meteorología e Hidrología (SENAMHI), and Brazil’s disaster management agency Centro Nacional de Monitoramento e Alertas de Desastres Naturais (CEMADEN). The team intends to contribute not only our own scientific expertise in flood impact analysis, but are also to the extended Hub network and SERVIR thematic teams by developing geospatial decision-making web apps using the free and open-source software Tethys Platform.

Weather and Climate

Weather and climate data are key to effectively inform water and ecosystem management, disaster preparedness, food security, energy provision and management, and land use planning. Successful risk management in these sectors requires climate and weather information to be provided at the right time and in useful formats. In the Amazon region, such data are not always collected, managed, shared, and analyzed to produce information needed for decision making at appropriate scales. Tailoring climate information appropriately for national to local scales is critical, as well as aligning the timescales of information with the timescales of decision-making.
Contact: Steve Prager s.prager@cgiar.org