AI-Driven Approach for Crop Type and Crop Parcel Mapping
Date & Time: May 13, 2025 Tuesday 9.30 EST / 15:30 CET / 16:30 EET
Presenters: Dr. Tao Liu and PhD Candidate Judy Long, Michigan Technological University
Tao Liu is an Assistant Professor at Michigan Technological University, whose research expertise includes developing data-driven models for land cover classification, forest health assessment, and agricultural crop mapping. Judy Long, a PhD candidate under Liu’s supervision, collaborates closely on these topics. In their presentation, they will discuss recent advancements in crop mapping, specifically highlighting best practices for crop type classification with a comparative analysis of pixel-based versus image-patch-based classification methods. In addition, they will also discuss the fine-tuning of the Segment Anything Model (SAM) for accurate delineation of crop parcels.
- - - - - - - - - - - - - - -
Evaluating vegetation patterns at the species level using UAV optical and thermal data
Date & Time: April 29,2025 Tuesday 9.30 EST / 15:30 CET / 16:30 EET
Presenters: Dr. Lucie Kupkova and Bc. Daniela Dančejová
Assoc. Prof. Lucie Kupková is the Head of the Department of Applied Geoinformatics and Cartography at the Faculty of Science, Charles University in Prague, and the leader of the ‘Team of Image and Laboratory Spectroscopy’ (https://www.tilspec.cz/). Her expertise lies in remote sensing for vegetation studies, image and laboratory spectroscopy, landscape change analysis, and environmental monitoring, with a particular focus on conservation applications. She serves as a regional coordinator of SCERIN, a member of the Steering Committee of the IGU Commission on Land Use and Land Cover Change, and a member of the Eurosite Remote Sensing Support Group.
Bc. Daniela Dančejová is a Master student at the same department. Her thesis focuses on modeling heat flux from UAV thermal data over the Úpské rašeliniště peat bog (a Ramsar site) in the Krkonoše Mountains. She received the Dean’s Award for an excellent final thesis in the Bachelor's program in Geography for her thesis titled ‘Remote sensing for classification of new wilderness vegetation in the hinterland of Kutná Hora.’
Abstract: The webinar will present case studies mainly from the Krkonoše Mountains National Park, Czechia focusing on vegetation dynamics evaluation and latent heat flux estimation from multitemporal UAV data, supporting the conservation of this unique relict arctic-alpine tundra in Central Europe. Aspects related to the correct acquisition and processing of UAV thermal data will also be discussed.
- - - - - - - - - - - - - - -
Retrieving plant traits from satellite observations using machine learning methods
December 18, 2024 Wednesday 9.30 EST / 15:30 CET / 16:30 EET
Presenter: Marian Švik, Global Change Research Institute of the Czech
Academy of Science (CzechGlobe), Brno, Czechia
Marian Švik studied Geoinformatics and Remote Sensing at Masaryk University Brno. Now he works at the Global Change Research Institute in the Remote Sensing department on topics like plant traits retrieval or land surface phenology.
- - - - - - - - - - - - - - -
Exploring alternative characterizations of cropland land surface phenologies in Southern Romania
November 26 Tuesday 9.30 AM ET
Presenter: Dr. Monika Anna Tomaszewska,
Center for Global Change and Earth Observations (CGCEO),
Michigan State University, East Lansing, United States
Dr. Tomaszewska is a post-doctoral research associate at the Center for Global Change and Earth Observation at Michigan State University. Together with Prof. Geoffrey Henebry, she is involved in several research projects funded by NASA, including: 1) Understanding the impacts of environmental changes, migration, and remittances on agro-pastoral communities in montane Central Asia; 2) Investigating the influence of changing agricultural policies on land cover and land use in Romania; 3) Analyzing the ecological traps for avian migrants in urban centers; and 4) Evaluating the potential of fine spatial resolution X-band SAR for crop phenology monitoring.
- - - - - - - - - - - - - - -
The impacts of the 2023 Türkiye–Syria earthquakes on agricultural land use
Date & Time: April 1 Tuesday 9.30 EST / 15:30 CET / 16:30 EET
Presenter: Dr. He Yin, Kent State University, USA
Dr. Yin is an Assistant Professor of Geography at Kent State University, where he heads the Remote Sensing and Land Science Lab (SensLand Lab). His research focuses on understanding the dynamic changes in land systems using remote sensing techniques and interdisciplinary approaches. His research spans a diverse range of topics, including armed conflicts, cropland abandonment, grassland management, and forest mapping. His recent research examining the impacts of armed conflicts on agricultural lands has been reported by media outlets such as CNN and BBC.
Abstract: In February 2023, earthquakes measuring 7.8 and 7.7 struck Türkiye and Syria. Described as the “disaster of the century,” these events rank among the most devastating in Türkiye’s history. This study provides a comprehensive assessment of their impact on agricultural land in Hatay, Türkiye’s hardest-hit province, using Sentinel-1 and Sentinel-2 imagery, supplemented by field interviews.
- - - - - - - - - - - - - - -
Cloud-gap-filled global maps of essential biophysical variables processed from the TOA Sentinel-3 catalogue in Google Earth Engine
December 13th, 2023: 15:30 – 16:30 CET (16:30-17:30 EET)
Presenter: David Kovacs- (University of Valencia)
I am David Kovacs from Hungary. I have done my bachelors in the UK at the University of Southampton in Aerospace Engineering. Because I love satellites and everything geo-astrophysics, I wanted to specialize in this field therefore I came to Valencia to do my Masters in Remote Sensing. I have joined the research groups of Jochem Verrelst, where I conduct research with him amongst many others with Katja Berger as well. I am currently working with Sentinel-3 OLCI data thus I will be presenting our global biophysical products at the Sentinel-3 Validation Meeting.
Abstract: We explored Sentinel-3 (S3) top-of-atmosphere OLCI data with hybrid retrieval models produced with ARTMO to inter four essential key variables: LCC, LAI, FAPAR, FVC. Models were trained based on SCOPE-6SV and Gaussian Process Regression algorithms were used to retrieve global vegetation maps for the year 2019. Results were correlated against MODIS LAI/FPAR products over different land covers. Uncertainty and correlation maps were created as well and will be discussed during the seminar.
- - - - - - - - - - - - - - -
Applications of thermal infrared imagery in the study of Earth’s ecosystems
When: March 15th, 2023: 10:00 – 11:00 AM
Presenter: Dr. Kerry Cawse-Nicholson
Dr. Kerry Cawse-Nicholson is a scientist at the NASA Jet Propulsion Laboratory. She serves as the assistant Section Manager for Earth Science at JPL, the ECOSTRESS science lead, and the co-lead of the SBG algorithms working group. Her research interests include the study of biodiversity, wildfires, and drought, as well as uncertainty quantification and the mathematical optimization of spectroscopic mission design.
Abstract: ECOSTRESS is a thermal radiometer mounted on the International Space Station and designed to study plant response to heat and water stress. With 5 thermal infrared channels, it also serves as a pathfinder for future thermal missions such as the Surface Biology and Geology (SBG) thermal instrument. Here we show several examples of ECOSTRESS science and applications, ranging from the original goal of studying plant water stress, to other enabled applications such as urban heat, volcanic eruptions, and wildfire studies. We also discuss how this work has led to the design of the upcoming SBG mission.
View recording