Egorov, Alexey

Geospatial Analyst
Office Phone Number: 

2015 — Ph.D. in Biology, Faculty of Biology, Moscow State University. Thesis title: “Structure of biodiversity of plant communities of  the Teberda reserve" (North-West Caucasus, Russia).

2008 — Visiting Scholar at the Geographic Information Science Center of Excellence, South Dakota State University. Project: “Establishing a global forest monitoring capability using multi-resolution and multi-temporal remotely sensed datasets”.

1995 — M.Sc. in Biology. Faculty of Biology, Dept. of Botany and microbiology, Yaroslavl State University.  Thesis title: "Influence of livestock grazing on highland vegetation of Karachayevo-Cherkessia".


Alexey Egorov joined the GSCE as a Geospatial Analyst in May 2009 to work on a NASA-funded research project entitled “Web-enabled Landsat data (WELD) — a consistent seamless near real time MODIS-Landsat data fusion for the terrestrial user community” with Dr. David Roy and Dr. Matthew Hansen. In the first stage of the project his duties included developing methods for mass-processing Landsat, Ikonos and QuickBird images for land cover and change characterizations. His results are available for science community from USGS and GIBS servers. Read more about WELD...






Currently he is working on continuation of WELD project - “Global Long-Term Multi-Sensor Web-Enabled Landsat Data Record – A Continuation Request” (proposal 17‑MEASURES-0086). He is developing an automated processing system of Landsat Analysis Ready Data (ARD) to generate medium resolution (30 m) maps, comprehensively characterizing land surface condition and dynamic for the conterminous United States qualitatively and quantitatively in space and timeline. The list of thematic maps that will be delivered includes: % tree cover, forest loss (location, date, magnitude), surface water, snow and ice, and % bare ground for all available Landsat mission medium resolution (30 m, TM/ETM+) history, i. e. 1983-2015. Read more about GWELD...





Alexey's research is based on a novel active learning algorithm, adapted to multispectral image classification problem. Well known in machine learning, this technique is undeservedly deprived of attention in remote sensing field. Active learning assumes querying expert by learning machine (i. e., bidirectional interaction between human and classifier) and allows more efficient use of computer and human resources in compare with traditional passive classification techniques. Read more...







Alexey is developing a forest monitoring algorithm, allowing near to real-time change‑detection of tree cover (deforestation and forest degradation), including location, date and magnitude. The algorithm is based on on nested segmentation active learning classifier and time series analysis.







Alexey is originally from Pushchino (Russia). He has a Ph.D. degree from Moscow State University in Biology which is devoted to study of highland vegetation of the Teberda State Biosphere Reserve (Northwestrn Caucasus, Russia), namely biodiversity of alpine communities and individual species distribution along 12 environmental gradients More (in Russian)...