BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:EO centre seminar with Professor Clement Atzberger - Clement Atzbe
 rger
DTSTART:20230727T090000Z
DTEND:20230727T100000Z
UID:TALK203419@talks.cam.ac.uk
CONTACT:Yi Zhang
DESCRIPTION:The mapping and monitoring of forest canopy height over large 
 areas at a fine\, deca-metric spatial resolution is important for the quan
 tification of carbon budgets in forests\, the assessment of tree growth ra
 tes and for the detection of deforestation events. The height information 
 can in principle be derived from photogrammetric analysis of stereo imager
 y (SfM) as well as from airborne laser scanning (ALS) data. Both approache
 s\, however\, do not scale well and entail relatively large costs for data
  acquisition. We assess the suitability of open and free Sentinel-2 time s
 eries for mapping and monitoring of tree height using sparsely sampled\, s
 pace-borne GEDI laser data as reference. Instead of relying on (monthly) c
 ompositie images as predictor variables\, we encode the noisy and irregula
 rly sampled Sentinel-2 times series into a few orthogonal\, gap-free and i
 nformation-rich representations. The representation learning is fully self
  supervised and based on an approach known as Barlow Twins. A simple neura
 l net is used to map the derived representations to the target heights fro
 m GEDI. The trained network is afterwards applied to the entire region of 
 interest to create gap-free forest height maps at 10m spatial resolution.
LOCATION:Main Seminar Room (1.25)\, David Attenborough Building
END:VEVENT
END:VCALENDAR