Document: SKO-3-514-201

Measurement accuracies for continental scale and multi-date remote sensing of land use and cover change.

SKOLE, D.L.* ¹, W.H.CHOMENTOWSKI ¹ and W.A.SALAS ²

Michigan State University, Lansing, MI 48933 USA ¹
University of New Hampshire, Durham, NH 03824-3525 USA ²

Abstract:
Global Change research relies heavily on the use of remote sensing information for large area mapping of land cover change, and for monitoring change over time using change detection methods. The use of historical satellite information is also an important component. Meaurement error can be an important factor in these analyses, and development of accuracy assessment methods and determination of measurement error is an often overlooked aspect of this work, but also extremely difficult. For instance statistically valid sampling strategies for whole continents, or for historical data is often not possible in any classical way. Moreover temporal gaps in time series are common due to difficulties associated with cloud free data, costs, and temporal discontinuities between satellite senors. This paper reviews the current issues and approaches, focusing particular attention on issues associated with landscape scale measurements over large regions.

Keywords: landscape scale measurements

This abstract is being presented at: 11:15 AM in session:
Symposium # 15: Measurement Error in Ecological Data.