A5036376825- Land Use and Ecosystem Services
- Remote Sensing and Land Use
- Remote-Sensing Image Classification
- Remote Sensing in Agriculture
- Impact of Light on Environment and Health
- Cellular Automata and Applications
- Modular Robots and Swarm Intelligence
- Remote Sensing and LiDAR Applications
- Systems Engineering Methodologies and Applications
- Water Quality Monitoring Technologies
- Urban Heat Island Mitigation
- Oil Palm Production and Sustainability
- Automated Road and Building Extraction
- Urban Transport and Accessibility
- Wildlife Ecology and Conservation
- Flood Risk Assessment and Management
- Precipitation Measurement and Analysis
- Distributed and Parallel Computing Systems
- Human Mobility and Location-Based Analysis
- Conservation, Biodiversity, and Resource Management
- Hydrological Forecasting Using AI
- Image Processing and 3D Reconstruction
University of California, Berkeley
2024
World Resources Institute
2018-2022
Radiant Earth
2021
National Geographic Society
2021
Unlike satellite images, which are typically acquired and processed in near-real-time, global land cover products have historically been produced on an annual basis, often with substantial lag times between image processing dataset release. We developed a new automated approach for globally consistent, high resolution, near real-time (NRT) use (LULC) classification leveraging deep learning 10 m Sentinel-2 imagery. utilize highly scalable cloud-based system to apply this provide open,...
Abstract This paper reports on recent improvements made to the FORMA ( Hammer et al., 2014a ) data product. The resulting system, FORMA250, is a 250-m alerting system updated daily. FORMA250 alerts are available through Global Forest Watch. These can empower law enforcement officials, government agencies responsible for protecting forests, nongovernmental organizations, companies committed sustainable forest management practices and supply chains, indigenous groups forest-dependent...
This technical note describes the data sources and methodology underpinning a computer system for automated generation of land use/land cover (LULC) maps urban areas based on medium-resolution (10–30m/pixel) satellite imagery. The deploy LULC taxonomy Atlas Urban Expansion—2016 Edition: open, nonresidential, roads, four types residential space. We used supervised machine learning techniques to apply this at scale. Distinguishing between recognizable, clearly defined use within built-up area,...
An issue of concern in landscape and urban planning, articulated the United Nation’s (UN’s) Sustainable Development Goals (SDGs), is increase land consumption over time. Indicator 11.3.1 SDGs dedicated to measuring it, underlining importance decreasing per person, a strategy that understood contribute positively climate mitigation host other social, economic, environmental objectives. This article aims explore practical implications official methods for 11.3.1, as well two alternatives,...
Synthetic Aperture Radar (SAR) data from Sentinel-1 mission is a valuable all-weather observation for various monitoring applications on the land surface. In particular, SAR observations have unique signatures over surface water which makes them appropriate to develop global system. Existing products rely multispectral significant shortcomings in cloudy regions. this study, we present novel Convolutional Neural Network (CNN) model applied at scale detect water. We used an existing product as...