BEGIN:VCALENDAR
VERSION:2.0
PRODID:IEEE vTools.Events//EN
CALSCALE:GREGORIAN
BEGIN:VTIMEZONE
TZID:US/Eastern
BEGIN:DAYLIGHT
DTSTART:20190310T030000
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
RRULE:FREQ=YEARLY;BYDAY=2SU;BYMONTH=3
TZNAME:EDT
END:DAYLIGHT
BEGIN:STANDARD
DTSTART:20191103T010000
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
RRULE:FREQ=YEARLY;BYDAY=1SU;BYMONTH=11
TZNAME:EST
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20191022T181506Z
UID:A9BFC4FD-FFD5-4BF1-BBBB-56B27A807872
DTSTART;TZID=US/Eastern:20191022T120000
DTEND;TZID=US/Eastern:20191022T130000
DESCRIPTION:Classifying space-borne measurements of reflected sunlight into
  complex land surface processes like deforestation are bound to generate i
 mperfect results. The magnitude and impact of classification errors in map
 s of land cover and land cover change remain unknown until the accuracy of
  the map has been estimated. Further\, the presence of errors prohibits 
 “pixel-counting” to obtain the areas of classes of land cover and land
  change classes. “Pixel-counting” refers to methods that produce areas
  as sums of map units assigned to map classes\, and generally make no prov
 ision for accommodating the effects of map classification errors. What is 
 required are observations of the reference conditions on the land surface 
 at locations selected by probability sampling of the study area. This samp
 le data (or reference data or accuracy assessment data) enable adjustment 
 for classification errors by the application of an unbiased estimator. The
  use of a variance estimator corresponding to the sampling design allows f
 or uncertainty quantification in the form of a confidence intervals. The f
 ree data policies of Landsat and Sentinel-2 combined with the emergence of
  powerful computing platforms such as Google Earth Engine\, have facilitat
 ed the sampling of study areas and the collection of reference observation
 s. In this presentation\, approaches to accuracy and area estimation are i
 llustrated and new cloud-based applications to facilitate sample-based est
 imation are discussed. Obstacles and issues to sample-based estimation and
  potential solutions are also discussed.\n\nSpeaker(s): Pontus Olofsson\, 
 \n\nGreenbelt\, Maryland\, United States
LOCATION:Greenbelt\, Maryland\, United States
ORGANIZER:zhuosen.wang@nasa.gov
SEQUENCE:1
SUMMARY:Why and how: sample-based estimation of area and map accuracy
URL;VALUE=URI:https://events.vtools.ieee.org/m/206583
X-ALT-DESC:Description: <br /><p>Classifying space-borne measurements of re
 flected sunlight into complex land surface processes like deforestation ar
 e bound to generate imperfect results. The magnitude and impact of classif
 ication errors in maps of land cover and land cover change remain unknown 
 until the accuracy of the map has been estimated. Further\, the presence o
 f errors prohibits &ldquo\;pixel-counting&rdquo\; to obtain the areas of c
 lasses of land cover and land change classes. &ldquo\;Pixel-counting&rdquo
 \; refers to methods that produce areas as sums of map units assigned to m
 ap classes\, and generally make no provision for accommodating the effects
  of map classification errors. What is required are observations of the re
 ference conditions on the land surface at locations selected by probabilit
 y sampling of the study area. This sample data (or reference data or accur
 acy assessment data) enable adjustment for classification errors by the ap
 plication of an unbiased estimator. The use of a variance estimator corres
 ponding to the sampling design allows for uncertainty quantification in th
 e form of a confidence intervals. The free data policies of Landsat and Se
 ntinel-2 combined with the emergence of powerful computing platforms such 
 as Google Earth Engine\, have facilitated the sampling of study areas and 
 the collection of reference observations. In this presentation\, approache
 s to accuracy and area estimation are illustrated and new cloud-based appl
 ications to facilitate sample-based estimation are discussed. Obstacles an
 d issues to sample-based estimation and potential solutions are also discu
 ssed.</p>
END:VEVENT
END:VCALENDAR