Plenary Talks

Satellite Technologies to Sustain the Earth’s Forest Biodiversity, Carbon, and Communities

Dr. Chad Oliver, School of Forestry and Environmental Studies, Yale University

Forests and rangelands each occupy about one third of the earth’s terrestrial area. They have the potential to provide much greater biodiversity, social livelihoods, and carbon sequestration than at present. Problems of habitat loss, illegal logging and poaching, forest and rangeland degradation, and community poverty stem largely from difficulties in planning, communicating, and monitoring sustainable practices. Technical tools that can be adapted/developed for Google Earth Engine can become the core mechanism for overcoming these problems. For forests, planning entails: 1) obtaining an inventory; 2) dividing the forest into “stands,” 3) projecting over several decades (e.g., six) the changes in inventory, habitats, carbon sequestered, harvested products, fire risk, and other values under various management scenarios (including “no management”); and 4) finding a suitable combination of scenarios for the different stands that provides a desirable (e.g., sustainable) flow of values over the decades. Several technologies have been developed for each step of the planning process, but they have not been well integrated or updated. Similarly, communicating the sustainable plan entails user-friendly, accessible tools for displaying the plan so that it is transparent and everyone trusts it. Here, too, many tools exist that have not been well integrated or updated. Monitoring the plan entails comparing the planned condition of each stand with the observed condition, a task that is uniquely suited for remote sensing. Here, satellite observation technologies can quickly detect forest activities where there should be none (e.g., infrared detection of heavy machinery and equipment; or the detection of roads), as well as unplanned forest changes (e.g., loss or decline of forest cover). Innovative, publicly accessible platforms such as Google Earth Engine have the opportunity to enable forests—as well as rangeland--to provide their environmental and social values in a rapid, user-friendly, and cost-effective way.

About the speaker: Dr. Chad Oliver, is the Pinchot Professor of Forestry and Environmental Studies and Director, Global Institute of Sustainable Forestry, at the School of Forestry and Environmental Studies, Yale University. Dr. Oliver has a wide geographic range of expertise. Much of his early work is incorporated in a book he wrote entitled Forest Stand Dynamics (1990, and update edition in 1996) with a former student as co-author. He is currently working on landscape approaches to management and global issues of landscape management and is involved in the technical tools, the policies, and the management approaches. Technical tools include Landscape Management System and “toggle system” for landscape sustainability analyses (, SilviaTerra for forest satellite inventory (a private company with Dr. Oliver as minor owner:, and two Google Earth Engine awards for identifying invasive species and detecting land condition change ( ). Dr. Oliver has consulted with timber companies, private landowners, industry associations, state and national forests, and public and elected officials in most major timber regions of the United States and some areas abroad. He has worked with these groups convening regional, national, and international symposia, forums, and workshops. He has testified at United States Senate and House of Representatives Committee Hearings. He was a member of the Science Panel at President Clinton’s Forest Conference in 1993, and has served on or chaired various scientific panels for the United States and Washington State executive and legislative branches of government, NAFTA, and the Society of American Foresters. Dr. Oliver is the author of more than 150 scientific and technical papers. His work has taken him to all parts of the United States and to Canada, Mexico, Turkey, Nepal, Japan, Thailand, Sweden, Finland, Russia, Ecuador, India, China, Germany, France, Austria, South Korea, Brazil, Chile, Ukraine, Bhutan, and Australia.

Urban Infrastructure Growth and its Unintended Environmental Impacts

Dr. Mikhail Chester, Arizona State University

The sustainability of cities is affected by the design and use of infrastructure, and modern infrastructure is the result of decades of land use decisions. Using the Google Earth and Maps Engines, we assess how infrastructure has grown in Los Angeles since 1900 and the resulting greenhouse gas emissions and urban heat island that it has enabled over the past century.

About the speaker: Dr. Mikhail Chester is an Assistant Professor in Civil, Environmental, and Sustainable Engineering at Arizona State University. His research focuses on the i) energy and environmental life cycle impacts of infrastructure systems, with an emphasis on transportation, and ii) resilience of infrastructure systems to climate change. Chester is leading several studies to investigate how the design of infrastructure is vulnerable to heat waves, how the design of infrastructure contributes to social vulnerability to heat, and how we can prepare infrastructure to be resilient to extreme weather events. Prior to ASU, Chester was a researcher at the Lawrence Berkeley National Lab and a post-doctoral researcher at UC Berkeley. He received his Civil and Environmental Engineering Ph.D. from UC Berkeley in 2008 and M.S. from Carnegie Mellon University in 2003. He received his B.S. from Carnegie Mellon in 2002 with a double major in Civil & Environmental Engineering and Engineering & Public Policy.

A scalable satellite-based crop yield mapper: Integrating satellites and crop models for field-scale estimation

Dr. Meha JainStanford University

Food security will be challenged over the upcoming decades due to increased fooddemand, natural resource degradation, and climate change. In order to identify potential solutions to increase food security in the face of these changes, tools that can rapidly and accurately assess farm productivity are needed. With this aim, we have developed generalizable methods to map crop yields at the field scale using a combination of satellite imagery and crop models, and implement this approach within Google Earth Engine. We test these methods in multiple regions across the globe for multiple crops, ranging from large-scale corn fields in the United States to smallholder wheat farming systems in India. We then use these data to understand yield trends through time, identify factors that may be associated with reductions in yield, and examine the efficacy of potential solutions to enhance yield in the face of environmental change.

About the speaker: Dr. Meha Jain is a postdoctoral research fellow at Stanford University. Her research broadly examines the impacts of environmental change on agricultural production, and strategies that farmers may adopt to reduce negative impacts. She combines remote sensing and geospatial analyses with household-level and census datasets to examine farmer decision-making and behavior across large spatial and temporal scales. To date her work has focused on the impacts of weather variability and groundwater depletion on agricultural production in India, and whether farmers are able to adapt their cropping practices to mitigate these impacts. She received her Ph.D. in Environmental Biology from Columbia University and her B.A. in Ecology from Princeton University. 

Automated malaria risk mapping with Google Earth Engine

Dr. Hugh Sturrock, UCSF Global Health Group

In elimination settings, malaria transmission becomes highly focal. Targeting interventions such as indoor residual spraying with insecticide, distributing insecticide treated bednets and active case detection becomes necessary to optimally allocate limited resources. Risk maps, based on malaria data and climatic/environmental variables, are valuable tools to allow such targeting to be done in an evidenced based way. However, these often require input from experts and are not updated regularly. Using Swaziland as an example, this presentation describes the development of an online mapping platform, using Google Earth Engine, which provides malaria control and elimination programs with automated real-time risk maps. 

About the speaker: Dr. Hugh Sturrock is an Assistant Professor of Epidemiology and Biostatistics working in the Malaria Elimination Initiative (MEI) at the UCSF Global Health Group. Prior to joining the Global Health Group, he completed a MSc in zoology at the University of Otago and a PhD at the London School of Hygiene and Tropical Medicine, investigating optimal survey methods for Neglected Tropical Diseases and Malaria. He has a broad interest in optimizing surveillance methods for tropical infectious diseases, with his research to date involving a combination of field work, spatial analyses, geostatistics and computerized simulations. With the Global Health Group, Hugh is focusing on the use and optimization of active surveillance to find and target asymptomatic infections as well as using routine surveillance data to generate risk maps to guide interventions.

Where the Trees go Ping and the Cows go Bong

Dr. Peter ScarthJoint Remote Sensing Research Centre

In Australia, comprehensive field data collection, time series image processing, and data integration approaches are used for reporting on the structure, state, and change of forest, woodland and arid ecosystems and for assessing and supporting grazing management in response to policy and landholder requirements. This work has historically been supported by high performance government owned computing infrastructure and extensive archives of field data, satellite, aerial and ground-based imagery, including optical, radar and lidar data.  Recent improvements in the curation and delivery of open data has enabled public access to these data, and with that, the ability to calibrate, validate and process remote sensing algorithms that work across Australia’s unique ecosystems. By democratizing access to compute resources, image data and delivery tools, the Earth Engine platform has allowed local NGOs to detect and report on land clearing, measure ground cover to assess grazing pressure and check on water use in intensively cropped and coal seam gas regions. The Earth Engine has also opened up opportunities to integrate hyperlocal spatial and satellite data into everyday activities, potentially offering rich, relevant information across a range of disciplines. I’ll be talking about some of these things.

About the speaker: Dr. Peter Scarth is a principal scientist with a cross disciplinary team at the Joint Remote Sensing Research Centre, a collaborative research program across several state governments and universities, that develops operational products based on clear policy and landholder needs. With extensive field data and imagery, they monitor changes in ground cover, tree structure, cropping areas, wetland ecosystems and biomass to map land use and change, vegetation clearing and thickening and detect changes due to fire, floods and overgrazing at a national scale. He also works for Australia's Terrestrial Ecosystem Research Network (TERN) to democratise access to spatial data, promote uptake of products by researchers, policy and the public, and advise a number of PhD students in the fields of lidar, rangeland science, big data, ecology and vegetation structure.

Mapping three decades of global surface water occurrence

Dr. Alan Belward, European Commission’s Joint Research Centre

Beyond the generalisation that 29% of Earth’s surface is ‘land’ and 71% ‘ocean’ the surface area of our planet covered by water is largely unknown. ‘Permanent’ features such as coasts, lakes and rivers are far from permanent (coastlines move, rivers meander, rivers dry up, new lakes form, others disappear), some seasonal water bodies appear regularly whilst others are highly episodic. Our knowledge of these dynamics at global scale is inadequate. Yet documenting changes in the world’s permanent and seasonal surface water is needed for climate, biodiversity, and desertification studies, for health, for disaster reduction and recovery and to manage agricultural, domestic and industrial consumption (among many other uses). Access to the entire Landsat archive combined with access to massive processing power, such as that from Google’s Earth Engine opens up opportunities for mapping water surfaces globally at 30m intervals over the last 30 years. Established methods for seasonal and permanent water mapping from high temporal revisit/low spatial resolution Earth Observing satellite systems have been adapted to Landsat. This paper reports on a series of Global Surface Water Occurrence maps derived from the analysis of 2,804,590 orthorectified Landsat 5 TM, 7 ETM+ and 8 OLI scenes acquired between 1st March 1985 to 10th March 2015 held in the Google Earth Engine (GEE). These new maps consistently resolve the global land/water boundary; they document changes in the world’s wetland, riverine and lacustrine ecotones, they provide a record of the establishment and disappearance of lakes and provide a unique picture of coastline advance and retreat.

About the speaker: Dr. Alan Belward works at the European Commission’s Joint Research Centre in Italy where he is head of the Land Resource Management Unit in the Institute for Environment and Sustainability. His research provides information for European policies linked to the Copernicus Earth Observation programme, the Rio Conventions, natural resource management and development-aid. He is currently lead author for the revision of the Global Climate Observing System’s Implementation Plan serving the UN Framework Convention on Climate Change and he is a member of the NASA USGS Landsat Science Team. He has a BSc in Plant Biology from Newcastle University, a PhD in satellite remote sensing of the global environment from Cranfield University and is a Fellow of the Society of Biology. 


Landsat Advisory Group Survey

Kass Green, Kass Green & Associates

About the speaker:  Kass Green’s experience spans thirty years of managing and supervising GIS and remote sensing professionals, as well as leadership in GIS and remote sensing research and policy. Ms. Green chairs NASA’s Earth Science Applications Committee, co-chairs the Department of the Interior’s Landsat Advisory Group, and is a member of the Landsat Science Team. She is a fellow in the American Society of Photogrammetry and Remote Sensing (ASPRS), and a past president of both MAPPS and ASPRS. Deemed a “rock star of remote sensing” by Directions Magazine, her research and accomplishments in mapping and GIS are world renowned.